Lake Mareotis: Reconstructing the Past: Proceedings of the International Conference on the Archaeology of the Mareotic Region held at Alexandria University, Egypt, 5th-6th April 2008 9781407306544, 9781407336527

Citation preview

BAR S2113 2010

Lake Mareotis: Reconstructing the Past

BLUE (Ed)

Proceedings of the International Conference on the Archaeology of the Mareotic Region held at Alexandria University, Egypt, 5th-6th April 2008 Edited by

Lucy Blue assisted by

Emad Khalil LAKE MAREOTIS

B A R Blue 2113 cover.indd 1

BAR International Series 2113 2010

21/05/2010 12:02:01

Lake Mareotis: Reconstructing the Past Proceedings of the International Conference on the Archaeology of the Mareotic Region held at Alexandria University, Egypt, 5th-6th April 2008 Edited by

Lucy Blue assisted by

Emad Khalil

BAR International Series 2113 2010

Lake Mareotis: Reconstructing the Past Proceedings of the International Conference on the Archaeology of the Mareotic Region held at Alexandria University, Egypt, 5th-6th April 2008 Edited by

Lucy Blue assisted by

Emad Khalil

BAR International Series 2113 2010

Lake Mareotis: Reconstructing the Past Proceedings of the International Conference on the Archaeology of the Mareotic Region held at Alexandria University, Egypt, 5th-6th April 2008 Edited by

Lucy Blue assisted by

Emad Khalil

BAR International Series 2113 2010

ISBN 9781407306544 paperback ISBN 9781407336527 e-format DOI https://doi.org/10.30861/9781407306544 A catalogue record for this book is available from the British Library

BAR

PUBLISHING

Contents CONTENTS LIST OF FIGURES, TABLES AND MAPS INTRODUCTION CONTRIBUTORS

i ii vii viii

1. FAWZI EL-FAKHARANI: PIONEER EXCAVATOR AT MAREOTIS. Mona Haggag

1

2. THE MAREOTIC REGION IN ANCIENT SOURCES. Mohamed S. Abd-el-Ghani

3

3. A NOTE ON LAKE MAREOTIS IN BYZANTINE TIMES. Mostafa El Abbadi

7

4. A STUDY OF THE EVOLUTION OF THE MARYUT LAKE THROUGH MAPS. Ismaeel Awad

11

5. LAKE MAREOTIS RESEARCH PROJECT. Lucy Blue

25

6. THE RESULTS OF A PRELIMINARY SURVEY AT MAREOTIS ISLAND. Dylan Hopkinson

35

7. THE CITY OF MAREA/PHILOXENITÉ. REFLECTIONS ON THE ALEXANDRIA

47

UNIVERSITY EXCAVATIONS, 1977-1981. Mona Haggag 8. MAREA PENINSULA: OCCUPATION AND WORKSHOP ACTIVITIES ON THE

57

SHORES OF LAKE MARIOUT IN THE WORK OF THE CENTER D’ÉTUDES ALEXANDRINES (CEAlex, CNRS USR 3134). Valérie Pichot 9. ON INTERPRETATIONS OF ARCHAEOLOGICAL EVIDENCE CONCERNING

67

MAREA AND PHILOXENITE. Mieczyslaw D. Rodziewicz 10. MAREA OR PHILOXENITE? POLISH EXCAVATIONS IN THE MAREOTIC REGION 2000-2007.

75

Krzysztof Babraj & Hanna Szymańska

11. THE LAKE STRUCTURES AT TAPOSIRIS. Marie-Françoise Boussac & Mourad El Amouri

87

12. SCHEDIA, ALEXANDRIA’S HARBOUR ON THE CANOPIC NILE. INTERIM REPORT ON

107

THE GERMAN MISSION AT KOM EL GIZA/BEHEIRA (2003-2008).

Marianne Bergmann, Michael Heinzelmann & Archer Martin

13. RECENT SURVEY WORK IN THE SOUTHERN MAREOTIS AREA. Penelope Wilson

119

14. WINERIES OF THE MAREOTIC REGION. Dorota Dzierzbicka

127

15. WATERFRONT INSTALLATIONS AND MARITIME ACTIVITIES IN THE

135

MAREOTIC REGION. Emad Khalil 16. LAKE MAREOTIS RESEARCH PROJECT. PHASES OF OUTRAGE AND DESTRUCTION.

147

Sameh Ramses & Ahmed Omar

SELECT INDEX

155 i

LAKE MAREOTIS: RECONSTRUCTING THE PAST

List of Figures, Tables and Maps Figures Chapter 4. Ismaeel Awad 1: Maryut Lake region. 2: 1548, Vray portraict de la ville d’Alexandrie en Egypte, Pierre Belon du Mans. 3: 1570, detail from Aegyptus antiqua, including a focused map of the region of Alexandria, Abraham Ortelius. 4: 1588, detail from Africae Tabula VII, Livio Sanuto. 5: 1655, detail from Royaume et désert de Barca et l’Egypte, Nicolas Sanson D’Abbeville. 6: 1717, detail from Carte de la Basse Egypte et du cours du Nil, Paul Lucas. 7: 1722, detail from Carte de l’Egypte Ancienne, Claude Sicard. 8: 1743, detail from Carte d’Egypte, Richard Pococke. 9: 1753, detail from Carte de l’Egypte ancienne et moderne, Robert de Vaugondy. 10: 1764, detail from Carte des Embouchures du Nil, et partie de son cours, Jacques Nicolas Bellin. 11: 1801, Carte d’Alexandrie, French campaign. 12: 1801, Carte d’Alexandrie, French campaign, original scale 1:100,000. 13: 1866, Carte des Environs d’Alexandrie, Mahmoud Bey El-Falaki. 14: 1866, Carte des Environs d’Alexandrie, Mahmoud Bey El-Falaki, original scale 1:200,000. 15: 1911 (1914), Atlas of Egypt, Vol. I, Lower Egypt, Survey of Egypt (G. Daressy), original scale 1:50,000. 16: 1949, map series 1:100.000. Survey of Egypt, original scale 1:100,000. 17: 1970s, map series 1:25.000, Egyptian General Survey Authority, original scale 1:25,000. 18: 1991, map series 1:50.000, Egyptian General Survey Authority. 19: 2001, Landsat satellite image, resolution 15m/pxl. 20: Evolution diagram of Maryut Lake’s surface areas (1801-2001). 21: Comparison between Maryut Lake 1801, French campaign map series 1:100,000 and 0 m level in the Mareotic region. 22: Comparison between Maryut Lake 1801-2001. Chapter 5. Lucy Blue 1: Location of Alexandria and Lake Mareotis. 2: An approximation of the ancient and present limits of Lake Mareotis. 3: Sites located along the northern shoreline, the southern shoreline and on Mareotis Island. 4: Topographic, ceramic and auger survey plan of Site 44 on Mareotis Island. 5: An example of a ‘lake wall’ feature (Site 44). 6: The Kibotos site (Site 09). 7: The ‘complex building’ (Site 13). Chapter 6. Dylan Hopkinson 1: The location of Mareotis Island and the sites mentioned in text, modified from De Cosson 1935. 2: Schematic of the main features of Mareotis Island. 3: The littoral features mentioned in this paper and topographic survey of the eastern settlement of the island. 4: The littoral features mentioned in this paper and topographic survey of the western settlement of the island. 5: Longitudinal and transverse schematic sections through the ‘water tower’ on Mareotis Island. Chapter 7. Mona Haggag 1: Kibotos harbour structure on the south shore of Lake Mareotis (Site 9 – Lake Mareotis Research Project). 2: Kibotos harbour structure, mooring ring. 3: Rock-cut Kibotos harbour structure at Ikingi Mariut. 4: Byzantine harbour of Marea. 5: Plan of the eastern harbour of Marea, after El-Fakharani 1977. 6: Decumanus between the eastern and middle ports of Marea. 7: Plan of shops off the main decumanus. 8: Basilica, Marea. ii

11 12 12 13 13 14 14 14 14 15 16 16 17 17 18 18 19 20 20 21 22 23 26 27 29 30 30 31 31 35 36 38 40 42 47 48 48 50 51 51 51 52

9: Enigma Buliding, Marea. 10: Winery excavated by El-Fakharani 1983.

52 53

Chapter 8. Valérie Pichot 1: Map of Egypt: location of Lake Mariout and the site of Marea. 2: General plan of the site identified as Marea. 3: The peninsula of Marea. 4: A) pier to the north of Marea peninsula, view from the north-west. B) ancient causeway connecting the peninsula to the mainland, view from the west. 5: Ancient causeway under excavation of Sector 1, view from the east. 6: Ancient causeway under excavation of Sector 2, view from the north. 7: Landmark and jetty to the south of the peninsula, view from the south-west. 8: Results of the geophysical survey undertaken on the Marea peninsula. 9: Sector 4 - Hearths in FS4024, under excavation. 10: Overall plan of the western zone of Sector 3. 11: Overall plan of the eastern zone of Sector 3. 12: Eastern zone of Sector 3 under excavation. 13: Sector 3 - One of the two limestone sphinxes discovered in the tower building ST300. 14: Sector 3 - Bronze candelabra after restoration in the CEAlex laboratory. 15: Sector 3 - Hoe chalk burner. Chapter 9. Mieczyslaw D. Rodziewicz 1: Central part of the plan of Mareotis drawn by Müller-Wiener in 1966, with the pilgrimage road from the lake harbour to Abu Mina and settlements, among them settlement “M” in Huwaryia village. 2: Abu Mina with surrounding settlements/karms, and pilgrimage road towards the lake harbour. 3: Byzantine House/Hospitium at the pilgrimage road from the lake harbour to Abu Mina. 4: Byzantine pilgrimage harbour/Philoxenite (Nos. 5-17) and western/coastal rural remains (Nos. 1-4). On the eastern side, structured causeway (No. 18). Chapter 10. Krzysztof Babraj & Hanna Szymańska 1. Remains of structures along the southern shore of Lake Mareotis. 2. Marea. Byzantine baths. 3. Marea. Baths and western courtyard. View from the south-west. 4. Marea. Well operated with sāqiyah. View from the south. 5. Marea. Funerary chapel. View from the west. 6. Marea. Plan of the basilica. 7. Marea. Apse of the basilica. View from the west. ‘A’ and ‘b’ locate the two burials. 8. Marea. Basilica. Amphora kiln. View from the west. Chapter 11. Marie-Françoise Boussac & Mourad El Amouri 1. General map of the site and location of areas under study at Taposiris Magna and Plinthine. 2. Drawing by Pascal Coste, made in 1820. 3. Map of the remains of the ancient city of Taposiris by A. De Cosson, 1935. 4. Location of the different sectors of the site of Taposiris that were studied along the lake shore. 5. View of the Taposiris bridge from north. 6. Plan and elevation of the north pile of the bridge, November 2004. 7. Model and reconstruction of the bridge in Taposiris Magna. 8. Overall plan and detailed outline of the eastern jetty. 9. The southern end of the eastern jetty. 10. Jetty at Gamal. View from north-west. 11. Jetty at Quseir. View from south-east. 12. Plan and elevation of the inner south face of the north flush and buttress. 13. View of the north flush during excavation, from east. 14. Lay out, elevations and sections of north and south walls of the water supply channel through the eastern jetty. 15. Water supply channel through the eastern jetty opening into the hoop-shaped basin. 16. Grooves and notches in the water supply chanel. 17. View of the hoop-shaped basin, the water supply channel and the fish-tank. iii

57 58 59 60 60 60 60 61 62 63 64 65 65 65 66 68 68 70 71

76 78 79 80 80 81 82 82 88 89 89 90 92 92 92 93 93 94 94 94 95 97 98 98 99

LAKE MAREOTIS: RECONSTRUCTING THE PAST 18. 19. 20. 21.

The pavement in the hoop-shaped basin. The opening between the hoop-shaped basin and the fish-tank. Main types of ceramics found in Taposiris Magna (Sector 9) during the Late Roman era. Ceramics from the filling of the hoop-shaped basin.

Chapter 12. Marianne Bergmann, Michael Heinzelmann & Archer Martin 1: Alexandria and its hinterland in 1925, Survey of Egypt, sheet 4. 2: Schedia. Topographical map based on quickbird satellite image (2005) indicating excavated areas (A) and corings (H). 3: Schedia. Map with reconstruction of the ancient topography. 4. Schedia. Plan of the southeastern settlement (Kom el Giza, Kom el Hamam) with excavation areas 1980-92 and 2003-06. 5: Kom el Giza, Area 3. Plan of Hellenistic baths. 6: Kom el Giza, Area 1. Plan of Roman villa, tombs and later basins. 7: Kom el Hamam. Plan with excavations areas 2003-2006. German Mission at Kom el Giza/Beheira. 8: Kom el Hamam. Plan of Area 6 with Roman storage building and later phases. 9: Kom el Hamam. Reconstruction of the Roman storage building. 10: Kom el Hamam, Area 9. Roman enclosure wall. 11: Kom el Hamam. Plan of Area 8 with basins for wine-production.

99 99 101 102 108 109 109 110 111 111 112 112 113 114 115

Chapter 13. Penelope Wilson 1: Ancient sites in the Delta, highlighting the Mareotis region. 2: Map of the sites surveyed, with 0 metre sea-level contour area highlighted in grey. 3: The main mound at Kom el-Mahar. 4: Late Ptolemaic to Early Roman pottery sample from Kom el-Mahar. 5: Red brick structures at Kom Trugi, excavated below foundation level. 6: Section through the cemetery mound at Kom el-Qadi.

119 120 121 122 123 123

Chapter 14. Dorota Dzierzbicka 1: A map of the Mareotic region with marked locations of wineries. 2: A schematic drawing of a Type 4 winery with names of its elements used in Greek papyri from Egypt.

127 130

Chapter 15. Emad Khalil 1: The approximate ancient and present limits of Lake Mareotis. 2: The Lake Mareotis Research Project survey area along the shores of the western Mareotic Arm. 3: Taposiris Magna tower. 4: The middle quay of the harbour of Marea. 5: The quay of Marea harbour was made of large regular limestone blocks without evidence of the use of mortar. 6: The box-shaped harbour on the southern shore of Lake Mareotis. 7: One of the quays extending into the lake at the northern shore of Mareotis western arm. 8: A lake wall extending parallel to the southern shore of the lake. Evidence for red mortar (opus signinum) can still be seen between the blocks. 9: Some of the multi-room square buildings located at Site 118 along the northern shore of the Mareotis Island. The buildings could have been used for storage purposes. Chapter 16. Sameh Ramses & Ahmed Omar 1: General plan from Alexandria to Mareotis. 2: Map of sites investigated during the Lake Mareotis Research Project. 3: Two examples of fish farm deformation along the southern shore of the lake. 4: Lime storing and collecting at Site 44. 5: Lake wall affected at Site 44. 6: Site 214: the partially excavated winery. 7: Site 214: part of the fragmented paved floor of a Roman period waterfront building. 8: Kibotos harbour extending from the southern shore of the lake. 9: One of the university properties on the land surrounding the Kibotos. 10: Archeological tell divided for private possession. 11: Private property sign and wall established on Site 201. 12: Flattened Hellenistic tomb at Site 202 during bulldozer work in 2007. 13: Two soccer fields erected at the tell boundaries around Site 204. iv

135 136 137 138 138 139 140 140 141

147 148 149 149 149 150 150 151 151 151 151 151 152

Tables Chapter 4. Ismaeel Awad 1: Comparison table between different periods of Maryut Lake.

21

Chapter 13. Penelope Wilson 1: Summary table of sites surveyed by the ‘Sais and its Hinterland Project’ in the region of southern Lake Mareotis.

124

Chapter 14. Dorota Dzierzbicka 1: Installations forming part of a wine-making complex mentioned in Greek papyri containing references to immovable property found in vineyards or to the appurtenances of a winery.

131

Maps Chapter 7. Mona Haggag 1: Marea in Lake Mareotis. 2: Marea’s public buildings on the lake shore.

48 50

v

LAKE MAREOTIS: RECONSTRUCTING THE PAST

vi

Introduction The following collection of papers represents the final synthesis of a conference entitled The International Conference on the Archaeology of the Mareotic Region. Lake Mareotis: Reconstructing the Past hosted by the University of Alexandria, Egypt between 5th & 6th April 2008. The objective of the conference was to bring together scholars that work in the Lake Mareotis region of Alexandria to a forum where they could present and discuss their ongoing research and field projects. The two-day conference was a great success as it provided an opportunity to share data and tackle themes and issues of common concern. It was also an opportunity for the Lake Mareotis Research Project, a collaboration between the Centre for Maritime Archaeology, at the University of Southampton, and the Department of Underwater Antiquities (DUA) of the Supreme Council of Antiquities (SCA), Alexandria, in conjunction with the Centre for Maritime Archaeology and Underwater Cultural Heritage, University of Alexandria, to show case the Lake Mareotis survey project and present recent results to colleagues. Lake Mareotis is a shallow body of brackish water currently about 90 km2 in area that extends south of Alexandria for about 12 km and to the west in the form of an arm some 40 km and is 3 km wide. Lake Mareotis was much larger in antiquity having been subject to siltation and reclamation over the last two thousand years. Due to its location and characteristics, Lake Mareotis it is known to have contributed significantly to the ancient economy of Alexandria and to the internal and international transport system of Greco-Roman Egypt. The significance of this region has been acknowledged since at least the 1930’s (De Cosson 1935), and subsequently highlighted by the pioneering work of scholars such as ElFakharani (1983). This volume is published in honour of these early pioneers, in particular El-Fakharani who not only initiated scholarly investigation of a number of sites in the Lake Mareotis region but who also engaged for the first time, future Egyptian scholars by including and encouraging students of Archaeology from the University of Alexandria to participate in excavations at Marea (see Haggag this volume). Thus, the conference and this publication attempts to follow in the footsteps of El-Fakharani by seeking to encourage an inclusive dialogue between all scholars currently active in the region. The following papers represent the most recent attempt to contextualise and interpret Lake Mareotis. Current research, not only that generated as a product of the Lake Mareotis Research Project (Blue & Khalil, forthcoming), but also exciting new discoveries from other sites around the shores of the lake, are beginning to provide a more comprehensive understanding of the breadth and scale of activities conducted around the shores of Lake Mareotis and highlight the substantial contribution the region made to the ancient economy of Alexandria. There is still much to be discovered but with continued dialogue we can begin to tackle unanswered questions and resolve sometimes contradictory, interpretations. This volume therefore presents a collection of thoughts and perspectives that are often variable in their focus and approach. A degree of editorial consistency has been employed but in order to retain the particular and diverse perspectives of the individual authors, specific nomenclature has been maintained. To that end, a variety of spellings for Lake Mareotis (Maryut, Mariout) and sites in the Mareotic region such as Marea, Mareia etc., are supported. It is hoped that this volume will encourage future discussion and publications that continue to highlight the importance of the Lake Mareotis region in antiquity. Lucy Blue Acknowledgements The editor would like to acknowledge both the British Academy and the Leverhulme Trust who provided financial support to undertake the Lake Mareotis Research Project of which this volume is just one product. In addition, the support of Dr Emad Khalil and the University of Alexandria is acknowledged for hosting and organising the conference. I would also like to thank the anonymous reviewer of this volume for their helpful comments. Bibliography El-Fakharani, F.A., 1983, Recent Excavation at Marea in Egypt. In G. Grimm, H. Heinen & E. Winter (eds.), Aegyptiaca Trevernsia II, Das Römisch-Byzantinische Ägypten. Akten des internationalen Symposions 26.-30. September 1978 in Trier, 175-186. Mainz. Blue, L., & Khalil, E. (eds.), forthcoming, A multidisciplinary approach to Alexandria’s economic past: The Mareotis case study. Southampton Monograph Series. Oxford De Cosson, A., 1935, Mareotis: Being an Account of the History and Ancient Monuments of the North-West Desert of Egypt and Lake Mareotis. London vii

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Contributors Marie-Françoise Boussac Professor in Greek History Paris Ouest Nanterre University 92000 Nanterre, France [email protected]

Mostafa El Abbadi Professor Emeritus of Greco-Roman Studies Faculty of Arts University of Alexandria Hussein Sobhi St. El-Shatbi 21526 Alexandria, Egypt [email protected]

Delphine Dixneuf Institut Français d’Archéologie Orientale 37, rue al-Cheikh Aly Youssef B.P. Qasr al-Ayni 11562 11441 Cairo, Egypt [email protected]

Mohamed S. Abd-el-Ghani Professor of Ancient History Dept. of Archaeology and Greco-Roman Stiudies Faculty of Arts University of Alexandria Hussein Sobhi St. El-Shatbi 21526 Alexandria, Egypt

Dorota Dzierzbicka Department of Papyrology Institute of Archaeology University of Warsaw Krakowskie Przedmieście 26/28 00-927 Warsaw 64 Poland [email protected]

Mourad El Amouri Ipso Facto Bureau d’Étude et de Recherche, Archéologie et Océanographie 4 rue de Tilsit 13006 Marseille, France [email protected]

Thomas Faucher ANR-Nomisma Université Paris-Sorbonne-Paris IV 1, rue Chevalier de la Barre 75018 Paris, France [email protected]

Ismaeel Awad Topography Department Centre d’Études Alexandrines 50, rue Soliman Yousri 21131 Alexandria, Egpyt [email protected]

Mona Haggag Bibliotheca Alexandrina P.O. Box 138 El-Shatbi 21526 Alexandria, Egypt [email protected]

Krzysztof Babraj Archaeological Museum in Kraków ul. Senacka 3 31-002 Kraków, Poland [email protected]

Michael Heinzelmann University of Cologne Institute of Archaeology Albertus Magnus Platz D-50923 Cologne [email protected]

Marianne Bergmann Georg-August-University Goettingen Archaeological Institute Nikolausberger Weg 15 D-37073 Goettingen [email protected]

Dylan Hopkinson Centre for Maritime Archaeology University of Southampton Highfield Southampton, England SO17 1BJ [email protected]

Lucy Blue Director, Centre for Maritime Archaeology University of Southampton Highfield Southampton, England SO17 1BJ [email protected]

viii

Sameh Ramses Supervisor of Egyptian Excavations & Surveying Team Department of Underwater Archaeology Supreme Council of Antiquities (SCA) 19 Abdel Moety el Gazaly, Stanley Alexandria, Egypt [email protected]

Emad Khalil Centre for Maritime Archaeology & Underwater Cultural Heritage Faculty of Arts University of Alexandria Hussein Sobhi St. El-Shatbi 21526 Alexandria, Egypt [email protected]

Mieczyslaw D. Rodziewicz Ul. Literacka 25 m.10 01-864 Warsaw, Poland [email protected]

Archer Martin American Academy in Rome Via Angelo Masina, 5 I-00153 Roma [email protected]

Hanna Szymańska Research Centre for Mediterranean Archaeology University of Warsaw 11, Sharia Mahalla, Heliopolis Cairo, Egypt [email protected]

Ahmed Omar Inspector, Department of Underwater Archaeology Supreme Council of Antiquities (SCA) 19 Abdel Moety el Gazaly, Stanley Alexandria, Egypt

Penelope Wilson Department of Archaeology Durham University South Road Durham, England DH1 3LE [email protected]

Valérie Pichot Centre d’Études Alexandrines 50, rue Soliman Yousri 21131 Alexandria, Egpyt [email protected]

ix

LAKE MAREOTIS: RECONSTRUCTING THE PAST

x

Fawzi El-Fakharani: Pioneer Excavator at Mareotis Mona Haggag

This volume is dedicated to the memory of Fawzi El-Fakharani who was, in many ways, a pioneer investigator of the area of Mareotis.

certain or clear that it was a temple dedicated to Hercules (Kanellopoulos 1994: 60-65). In 1970 Fakharani was delegated to Libya, where he established a department of Classical Archaeology. In Libya, Fakharani excavated the area of the Byzantine palace of the City of Tocra, ancient Taucheira. The patriarch, who presented the city at the council of Nicea in 325 CE, must have resided in this palace. Fakharani uncovered parts of the Eastern Church to which the palace was attached. He also unearthed the church’s confirmatorium and baptistery, as well as two beautiful mosaic panels (pers. comm.).

Fawzi El-Fakharani was born in Alexandria in 1921. He graduated from the Faculty of Arts, University of Alexandria in 1946 and directly after his graduation he went to the United Kingdom for higher studies. In 1953 he obtained a Masters degree from the University of Liverpool. His undergraduate dissertation was entitled “The origins of stucco reliefs in Classical Art”. Afterwards he moved to University College London where he acquired a doctoral degree in 1957 on “Stucco Reliefs in Roman Art”.

In 1973, back in Alexandria, Fakharani began his studies of the city of Marea, thus realising his life’s dream, which was to discover and document the ruins of this city. After a long and exhaustive period of preparation, he finally obtained the permission to excavate the site identified previously as the city of Marea, to the south of the western extremity of Lake Mareotis, some 45 km distance from Alexandria. With the exception of previous identifications of the site as that of Marea made by Mahmoud Pasha ElFalaki (1872: 85, 86, 91, 96, 97, 101), and the valuable monograph on Mareotis by De Cosson (1935: 131-135), the academic community of archaeologists and researchers had paid no particular interest in the area of Mareotis. Fakharani’s determined enthusiasm and efforts to identify the city’s area proved successful, as the results of his excavations shed light on the various stages of life within the city and on the importance of the area as a whole (El-Fakharani 1983: 175-204). Some of the articles of this volume are devoted to Marea (see Haggag, Pichot, Rodziewicz, Babraj and Szymańska) through which the reader can figure out that the discovery of Marea was and is, of great importance to researchers, as well as to the community at large.

Fakharani returned to Egypt and was appointed lecturer in the University of Alexandria in 1958. In 1964 he was promoted to an assistant professor, and won the chair of Professor of Classical Archaeology in 1974. He was appointed Chair of the Department of Classical Civilization between 1976 and 1980, and the director of the Institute for Mediterranean Studies between 1986 and 1989. In 1990 he became Chair of the Department of Archaeology at the University of Tanta, Egypt. In 1966 Prof. Fakharani was delegated to the University of Amman in Jordan. There, he began his first endeavours in archaeological fieldwork. His excavation target was to continue uncovering the grand Roman theatre of Amman. He published the results of his investigations in the Jahrbuch Des Deutschen Archäologischen Instituts Archäologischen Anzeiger, in 1975 (El-Fakharani 1975a: 377-403). In 1967, his interests shifted towards the site of Qal’at AlJabal on which stands the so-called Temple of Hercules (El-Fakharani 1975b: 533-554). There he reinvestigated the remains and uncovered most of the building’s foundations and annexed colonnades. Scholars differed with regard to which deity the temple was dedicated, yet Fakharani’s hypothesis was that the building had characteristic features unknown to temples of the East Roman Provinces in the time of Marcus Aurelius, although common in libraries of the same epoch in the neighbouring countries (El-Fakharani 1975b: 554). Fakharani reached the conclusion that the building was not a temple but the library built by Marcus Aurelius at Philadelphia, the city that he described as Alexandria’s daughter city. Despite his reasonable arguments and study published in the periodical of the University of Rostock, the building was still termed as the “Temple of Hercules”. This remained the case until a study conducted in 1994 of an inscription carved on the architrave of the building showed that it was not at all

With only ten thousand Egyptian pounds as a budget, Fakharani was given the daunting task of locating a suitable area to undertake a bold and pioneering step in Egyptian archaeology, his goal being to accommodate both male and female Egyptian students in the field. This was something that had never been undertaken before and had to be managed with great care and respect for social traditions. Not only did he have to secure separate sleeping and study areas but he also had to supply electricity to the camp, food and sustenance for the students, and transportation to the desert area of Mareotis. All of this he accomplished on a meagre budget. Besides the difficulties of obtaining permission from the university to take the students to the field, he had also to convince their parents, especially those of the female students, that their daughters would be safe and 1

LAKE MAREOTIS: RECONSTRUCTING THE PAST secure. It was not easy for Egyptian families during the 1970s to send their daughters to the desert for field training, but Fakharani managed to allay any of the families fears and worries.

German team. They all emphasised that the remains from the Ptolemaic era lay only in the area of the nursery which then belonged to the Faculty of Agriculture, the University of Alexandria. Again, it took determination and patience to acquire the permission to excavate this site. During this time, Fakharani excavated a few trial trenches in the area around the remains of an alabaster tomb. Although the trenches yielded nothing of conspicuous importance, a Byzantine well just next to the tomb shed new light on the use of this area during this period. As the level of the monolithic pavement of the early Ptolemaic tomb was higher than that of the Byzantine rim of the uncovered well, it became evident that the tomb was not in situ as Adriani had assumed (Adriani 1940: 15-23), and that it had been moved to its present location during the Byzantine or early Islamic periods (author’s interpretation).

Soon after the ancient buildings emerged from the desert sands, articles and interviews in local and international media started to appear. These articles helped win the support of the families of the students who had been allowed to camp in the desert with Fakharani on this pioneering expedition. The student families soon realised from these articles and media reports just how important the work was that their young daughters and sons were accomplishing. Thus, Prof. Fakharani overcame all obstacles and all the students, including myself, who participated in the excavations, were, for the first time in the history of the department, to become the lucky generation of students who had the opportunity to be trained in the field. We are extremely grateful to Prof. Fakharani for that experience. The directors of the Lake Mareotis Research Project, Lucy Blue, Sameh Ramses and Emad Khalil (to whom this volume owes its inspiration), are now reminding us of those wonderful days by involving students in their current fieldwork surveying the shores around Lake Mareotis, the first time this has happened since the pioneering days of Fakharani.

At the same time that Prof. Fakharani received the permission for his excavation of the nursery, Parkinson’s disease had already begun digging its deadly roots into his aging body. Accordingly, he entrusted the project to another great scholar who has also made considerable contributions to the archaeology of Alexandria and its environs, namely Jean-Yves Empereur. In March 2004, Fakharani passed away but he left behind a generation of empowered, highly motivated, committed and devoted students who strive to continue their work to the standards and example set by Fawzi El-Fakharani.

Prof. Fakharani never once lost sight of the idea that Marea was the “Pompey of the East”, to use his own expression (El-Fakharani 1977: 5). It deserves more care and diligence as it represents a unique archaeological site that gives us a complete picture of the everyday life of ordinary people, in contrast to other archaeological projects in Egypt which have a tendency to focus on discovering how our once noble Pharaonic families lived and prepared for the afterlife. Marea, together with the results of survey and excavations from other settlements around the shores of the Lake (see Blue, Hopkinson, Boussac & El Amouri, Bergmann, et al., Wilson, Dzierzbicka, Khalil, and Ramses & Omar this volume), illustrate how people in a postPharaonic era lived, what their beliefs were, what their baths looked like and how they were used, the shape and layout of their shops, cisterns, and how they buried their dead. Marea necessitates more preservation, conservation, restoration and preparation efforts to emphasise the site’s potential, in order to amongst other things, place it firmly on the tourist map of Egypt.

Bibliography Adriani, A., 1940, Fouilles et decouvertes. Alexandrie. I. Tombeau en alabaster du cimetiere latin. Annuaire 3, 1935-1939: 15-23. Alexandria. De Cosson, A., 1935, Mareotis: Being an Account of the History and Ancient Monuments of the North-West Desert of Egypt and Lake Mareotis. London El-Fakharani, F.A., 1975a, Das Theater von Amman in Jordanien. Jahrbuch des Deutschen Archäologischen Instituts Archaologischer Anzeiger 3: 377-403. El-Fakharani, F.A., 1975b, The Library of Philadelphia (?) Or The So-Called Temple on the Citadel Hill in Amman. Wissenchaftlische Zeitschrift der Universität Rostok XXIV.6: 533-554. El-Fakharani, F. A., 1977, A Preliminary Report on the 1st Season of Excavation at Marea. Excavation Report presented to the Antiquities Department, Egypt. El-Fakharani, F.A., 1983, Recent Excavation at Marea in Egypt. In G. Grimm, H. Heinen & E. Winter (eds.), Aegyptiaca Trevernsia II, Das Römisch-Byzantinische Ägypten. Akten des internationalen Symposions 26.-30. September 1978 in Trier, 175-186. Mainz. El-Falaki, Mahmoud Bey, 1872, Memoire sur l’antique Alexandrie, ses faubourgs, ses environs. Copenhagen. Kanellopoulos, C., 1994, The Great Temple of Amman, Vol. I: Architecture. American Center of Oriental Research – Amman: i-xii.

Finally, in 1999, Fakharani again took us to the field. This time Fakharani’s field was in Alexandria itself. His investigations were of the Latin cemetery which he believed to have been the site of the royal Ptolemaic burial ground. A geophysical survey was conducted in three successive phases: firstly, by colleagues from Alexandria University, secondly, in collaboration with Mrs. Calliope Lemniou Pappacosta, a researcher from Greece in consultation with experts from the University of Patras, and thirdly, by a

2

M. ABD-EL-GHANI: ANCIENT SOURCES

The Mareotic Region in Ancient Sources Mohamed S. Abd-el-Ghani

Before addressing the Mareotic region in antiquity it would be appropriate to explore how Egyptian frontiers are discussed in the classical sources. There is a crucial and decisive paragraph in Strabo (17.1.5) to this effect – one that makes particular reference to the western frontiers of Egypt – the primary focus of this paper – and is quoted “verbatim” as follows:

and on the south Lake Marea, also called Mareotis: “This is filled by many canals from the Nile, both from above and on the sides …” (Strabo 17.1.7). What is particularly interesting to note about the inhabitants of the Mareotic region, and the Western Desert of Egypt in general, is the ethnic distinction that characterises them apart from the rest of the Egyptians, even after their identity as Egyptians was settled from an administrative viewpoint. It is a phenomenon that is observed throughout the ancient texts through to the present day. This contrast between the Mareotes and the Egyptians is explicitly expressed in a document dated from the middle of the 4th century AD about the smuggling of natron.1

“Now the early writers gave the name Egypt to only the part of the country that was inhabited and watered by the Nile, beginning at the region of Syene and extending to the sea; but the later writers down to the present time have added on the eastern side approximately all the parts between the Nile and the Arabian Gulf (i.e. the present Red Sea), and on the western side the parts extending as far as the oases, and on the sea-coast the parts extending from the Canobic mouth to Catabathmus (the present el-Salloum) and the domain of the Cyrenaeans. For the kings after Ptolemy became so powerful that they took possession of Cyrenaea itself and even united Cyprus with Egypt. The Romans, who succeeded the Ptolemies, separated their three dominions and have kept Egypt within its formers limits.”

Marea is stated by Herodotus as being located at the key position with regard to the Egyptian frontiers to the west. He relates that king Psammetichus I of the 26th Dynasty posted garrisons at Elephantine on the side of Ethiopia, at Daphnae of Pelusium on the side of Arabia and Assyria, and at Marea on the side of Libya (Herodotus II.30). Marea was also mentioned by other classical authors as the battlefield of some pitched battles in antiquity before the conquest of Egypt by Alexander the Great. In his reference to the second Egyptian revolt against the Persian rule under King Artaxerxes (c. 460 BC), Thucydides pointed out that Inaros son of Psammctichus, “setting out from Maria, the city just north of Pharos, caused the greater part of Egypt to revolt from king Artaxerxe” (Thucydides I.104).

Among the early writers referred to above by Strabo, it was Herodotus (II.17) who adopted the ancient theory that Egypt was the land inhabited by Egyptians and was watered by the Nile. Nevertheless, he considered that the Egyptians were not only the inhabitants of the Nile Valley and the Delta proper, particularly the area that was directly watered by the Nile River, but also those places to which the Nile waters extended. In order to illustrate this, he set out to write a significant account about the inhabitants of the cities of Marea and Apis. His account, in this concern, goes as follows:

Diodorus of Sicily reported that the decisive pitched battle between King Apries, the legitimate king (588-566 BC) of the Saite Dynasty, and Amasis II, one of the king’s generals and claimant to the throne who became a renowned king, took place near the village of Maria (Diodorus Siculus I.68),2 after the former’s defeat by the Greeks of Cyrenê and Barcê.

“The men of the cities of Marea and Apis, in the part of Egypt bordering on Libya, thinking themselves to be not Egyptians but Libyans, and misliking the observance of the religious law which forbade them to eat cow’s flesh, sent to Ammon saying that they had no part or lot with Egypt: for they dwelt outside the Delta and did not consent to the ways of its people, and they wished to be allowed to eat of all food. But the god forbade them: all the land, he said, watered be the Nile in its course was Egypt and all who dwelt lower than the city Elephantine and drank of that river’s water were Egyptians.” (Herodotus II.18)

The first specific and detailed account about Lake Marea and the Mareotic region was that of Strabo. After his statement that Lake Mareotis was filled by many canals from

1. P. Lond. II.231, p. 285 - Hunt and Edgar, Select papyri II. 428 = P. Abinn. I.9, ll. 4-5: ὥστε ὅσα νίτρα καταλαμβάνεις εἴτε δια Μαρεωτῶν εἴτε δια Αἰγυπτείων κατερχόμενα. 2. Cf. Herodotus II.169 where he identified the site of that battle to be at the town of Momemphis “ἐς Μώμεμφιν πόλιν” which is roughly identified as Kom Abu Billu by Ball 1942: 18; Yoyotte & Charvet 1997: 118, n. 249 identified it as “aujourd’hui Kôm el-Hisn” in their commentary on Strabo 17.1.22

That they drank of that river’s water is well attested in Strabo’s account of Alexander’s first visit to the place of Alexandria, the advantages of the site, and his determination to found a city bearing his name on the site. Among the advantages of the (new) city’s site was its location between two (seas), on the north the Egyptian sea, as it is called, 3

LAKE MAREOTIS: RECONSTRUCTING THE PAST the Nile, both from above and on the sides, he goes on to complete the picture of the whole region as follows:

sites of the “Libyan coastal nome” from immediately before Catabathmus (es-Salloum) to Glaucus Promontory (El-Imâyid) (V. 5. 4 - 7), he goes on to report the sites of the “coastal Mareotic Nome” or “Νομοῦ Μαρεώτου παράλιος”, i.e., namely, the village of “Cheimw” (El-Bordân) and Plinthine (ruins 6 km NE of Sidi Kireir) (Ptolemaios IV.5.8). Finally, he comes further east to Chersonesos the Smaller (El-Dekheila) and then to Alexandria, the capital of the whole of Egypt, to quote Ptolemy (Ptolemaios IV.5.9).

“...and through these canals the imports are much larger than those from the sea, so that the harbour on the lake was in fact richer than that on the sea, and here the exports from Alexandria also are larger than the imports; and anyone might judge, if he were at either Alexandria or Dicaearchia (Puteoli in Italy) and saw the merchant vessels both at their arrival and at their departure, how much heavier or lighter they sailed thither or therefrom. And in addition to the great value of the things brought down from both directions, both into the harbour on the sea and into that on the lake, the salubrity of the air is also worthy of remark. And this likewise results from the fact that the land is washed by water on both sides and because of the timeliness of the Nile’s risings … at Alexandria, at the beginning of the summer, the Nile, being full, fills the lake also and leaves no marshy matter to corrupt the rising vapours. At that time, also, the Etesian winds blow from the north and from a vast sea, so that the Alexandrians pass their time most pleasantly in summer.” (Strabo 17.1.7.)

In another passage of Ptolemy (13-32) he gives a list of the “villages (inland) of the Lybian Nome” from west to east; the last village to the east being the one called “Mareotis”. In another passage he provides another list of the “cities and villages of the Mareotic Nome” (the part inland), further to the east on the confines of the previous Lybian Nome: the farthest village to the east being that called the “village of ancient Marea” or Παλαιμάρεια κώμη (Ptolemaios IV.5.34). According to Ptolemy’s calculations the village of Mareotis, the farthest point to the east of the Libyan Nome, was situated on longitude 58 (27º.25’) and latitude 28º.20’, however, the village of ancient Marea is located at longitude 60 (29º.25’) and latitude 30º.10’. Thus, the village of “Mareotis” was situated to the south-west of the village of “Ancient Marea”, at a location yet to be identified with a specific ancient site. Ball (1942: 114) could not identify the present location of either site.

From the above paragraph of Strabo the advantages of Lake Mareotis with its vivid harbour, as well as that of the Mareotic region in general, is quite obvious: intensive commercial activities of imports and exports from the interior of Egypt as well as Mediterranean cargoes; an excellent and healthy climate which was also a tourist attraction in itself. Another passage Strabo (17.1.14) sheds more light on the size of the lake and the other activities of the inhabitants of its surrounding shores. He describes the situation as follows:

Some of these aspects referred to in the works of the classical authors concerning the Mareotic region are reflected in papyrological documents, scarce though they are. In the light of such documents the region of Mareotis constituted an administrative “nome” during the 1st (OGIS 669 = S.B. V.8444 (AD 68), l. 48; P. Iand. IV.53, col. 3, l. 1 (AD 88), 2nd (P. Ryl. II.78, l.8 (AD 157); P. Oslo III.84 (AD 138161) ll. 12-14; P. Haun. II.22, l. 3.) and 3rd (P. Flor. III.338, l.6; BGU I.13, AD 289, l.2.) centuries AD.

“Lake Mareia, which extends even as far as this [the Chersonesus/ ‘Marabit’ according to Yoyotte & Charvet 1997 or el-’Dekheila’ according to Ball 1942], has a breadth of more than one hundred and fifty stadia and a length of less than three hundred. It contains eight islands; and all the shores round it are well inhabited; and the vintages in this region are so good that the Mareotic wine is racked off with a view to ageing it.”

About a century after Strabo’s visit to Egypt, Pliny the Elder in his Historia Naturalis reported about Lake Mareotis as follows:

The second important aspect highlighted in these few documents is the close relationship between the Mareotic Nome and Alexandria. Some individuals in these documents would come to Alexandria, through the Mareotic Nome, for business or official or judicial purposes. This is clear in some private or business letters (P. Haun. II.22; P. Flor. III.338). Sometimes they hint at some degree of co-ordination between the high officials of Mareotis and Alexandria when necessity requires. In a poorly preserved document, the essence of which could be understood in its general lines owing to the existence of several key words, two persons domiciled in the Mareotis region are mentioned in a petition to the strategos of Alexandria regarding the theft of two camels belonging to them. According to the publisher’s reading and filling of the lacunae, which seems almost plausible, they ask the Alexandrian strategos to write to his Mareotic counterpart to summon some suspected person to present himself before him for investigation (P. Oslo III.84). Why was the Alexandrian strategos asked to interfere instead of resorting directly to the Mareotic strategos? Does this further illustrate a degree of collaboration between the two strategoi?

“Lake Mareotis, which lies on the south side of the city, carries traffic from the interior by means of a canal from the Canopic mouth of the Nile; it also includes a considerable number of islands; being 30 miles across and 250 miles in circumference, according to Claudius Caesar.” (Pliny V.XI.63)

Strabo and Pliny also note the variations in measurement of the lake particularly during the inundation season in the summer when the Nile flooded. From their accounts it is clear that the lake was thus much more extensive in antiquity than today (Yoyotte & Charvet 1997: 104, n. 179; Strabo 17.1.14). Last but not least among the classical works concerning the description of the Mareotic region, is the Geography of Claudius Ptolemaios in the 2nd century AD. In his fourth book (Chapter 5 about Egypt) Ptolemy begins his description of the geography of Egypt from west to east stating the longitude and latitude of each given site. After covering the 4

M. ABD-EL-GHANI: ANCIENT SOURCES Bibliography

The theft of two camels in this document also leads us to comment on the repeated reference to camels in the Mareotic documents. This is not unexpected from an area situated on the edge of the Western Desert where the camel was the chief means of traffic and transport. In one of the documents relating to business affairs, a sum of 900 drachmae is reported as a rent for some camels (P. Haun. II.22, ll. 9-10 and note on l.9.). In another document dated AD 289, two Mareotes sell an Arabian camel belonging to them to one of the chief officers (primpilaris) of the Prefect of Egypt for 16½ talents of silver (BGU I.13, II. 1-6).

Papyri BGU = Aegyptische Urkunden aus Königlichen (later Staatlichen) Museen zu Berlin, Griechische Urkunden (13). H. Satzinger (ed.), 1904. Berlin. P. Abinn. = The Abinnaeus Archive: Papers of a Roman Officer in the reign of Constantius II (1-9). H.I. Bell, V. Martin, E.G. Turner & D. Van Berchem (eds.), 1962. Oxford. P. Fay. = Fayum Towns and their Papyri (134). B.P. Grenfell, A.S. Hunt, D.G. Hogarth & J.G. Milne (eds.), 1900. London. P. Flor. = Papiri greco-egizii, Papiri Fiorentini III. Documenti e testi letterari dell‘età romana e bizantina (III, 338). G. Vitelli (ed.), 1915. Milan. P. Fouad = Les Papyrus Fouad I (11). A. Bataille (ed.), 1939. Cairo. P. Hamb. = Griechische Papyrusurkunden der Hamburger staats- und Universitätsbibliothek (I, 39). P.M. Meyer (ed.), 1924. Leipzig. P. Haun. = Papyri Graecae Haunienses (II, 53). Letters and mummy labels from Roman Egypt. A. Bülow-Jacobsen. (transl. & ed.), 1981. Bonn. P. Iand. = Papyri Iandanae (IV, 53). K. Kalbfleisch, E.H. Schaefer, G. Rosenberger, D. Curschmann, L. Eisner, J. Sprey, G. Spiess, L. Spohr & J. Hummel (eds.), 1912. Leipzig. P. Lond. = Greek Papyri in the British Museum (II, 231). F. Kenyon & H. Bell (eds.), 1973. Milan. P. Oslo = Papyri Osloenses (III, 84). S. Eitrem & L. Amundsen (eds.), 1925. Oslo. P. Ryl. = Catalogue of the Greek and Latin Papyri in the John Rylands Library. Manchester (II, 98, 227). C.H. Roberts & E.G. Turner (eds.), 1952. Manchester. S.B. = Sammelbuch griechischer Urkunden aus Aegypten (V, 8444). F. Bilabel & E. Kiessling (eds.), 1952. Wiesbaden.

Some of the Mareotic products are also mentioned in the few available documents. In spite of the excellence and distinction of the Mareotic vineyards and wine, there appears to be limited documentary sources that comment on their merits. A single document, a private letter dating to the 2nd century AD, in which the sender tells his friend and addressee that he dispatched him a variety of gifts, among which we find: “... and a number of bunches of grapes, - 6 small clusters: 2 of them white, 2 of the ‘royal’ variety, the Mareotic as it is called, and likewise 2 smoke coloured, - in a small basket through Harklides” (P. Fouad, 77, ll. 15-20, l. 17). In other documents dating to the 3rd century AD, there are occasional references to “strips of linen”, ταινίδιον: some of them white, some “Mareotic” (P. Ryl. II. 227, ll. 25-33). If we notice that the use of ταινίδιον is a diminutive of ταινεία which means “a headband” worn especially as a sign of victory; and if we bear in mind that a strip of land near Lake Mareotis was named ταινεία, mentioned in Athenaeus (I.33e) and C. Ptolemaios (4.5.14), could one infer some connection between this place and the production of “linen headbands” whence “it might have derived its name? There is another mention of “a glassware ... of a fine Mareotic quality” from the 4th century AD (P. Fay. 134, ll. 4, 6). Another 4th century document makes a clear reference to the illegal smuggling of natron by Mareotes or other Egyptians across the desert. In this correspondence between two high officials in charge of the natron monopoly (AD 342-51), strict orders are issued to detain the camel drivers and their beasts, whether in Arsinoe or elsewhere, in order to safeguard the natron of the Fiscus (P. Abinn. I.9 = P. Lond. II.231).

Other Sources Athenaeus, Deipnosophists, C. Burton (transl.), 1927. London. Ball, J., 1942, Egypt in the Classical Geographers. Cairo Diodorus Siculus, Diodorus Siculus. C.H. Oldfather (transl.), 1935. Cambridge, MA. Herodotus, The Histories. A. de Sélincourt (transl.), 1971. Middlesex. Pliny the Elder, Natural Histories. J. Bostock & H.T. Riley (eds.), Perseus Digital Library: http://old.perseus.tufts.edu/cgi-bin/ptext?doc=Perseus:text:1999.02.0138:toc (accessed January, 2010). Ptolemaios (Claudius Ptolemy), The Geography. E.L. Ste venson (transl. & ed.), 1932. New York. Strabo, Geography. H.L. Jones (transl.), 2001. London. Thucydides, The Peloponnesian War. J.M. Dent (transl.), 1910. New York, London. Yoyotte, J., & Charvet, P., 1997, Strabon, Le voyage en Egypt. Paris.

Finally, there are some references to army troops stationed in the Mareotic region, together with the payments and extra payments to them, as well as the divisions of the army to which they were assigned (P. Iand. IV. 53 (AD 88), col. 3; P. Hamb. I.39 (AD 179) (63=BB). Although the image of the Mareotic region, especially in the papyri, is somewhat fragmentary, mosaic-like, and far from complete, it does shed, however, some light on the region. It is clear that the material evidence of future archaeological excavations will add a great deal to our knowledge of this region.

5

LAKE MAREOTIS: RECONSTRUCTING THE PAST

6

M. EL ABBADI: LAKE MAREOTIS IN BYZANTINE TIMES

A Note on Lake Mareotis in Byzantine Times Mostafa El Abbadi

This paper will present readings of selected passages in the ancient sources that have a bearing on Lake Mareotis and its region. As will be seen, the name Mareotis was often exchanged for Mareia, the name of its capital city and harbour, which in Pharaonic times served also as a garrison city protecting the north-western approaches to Egypt, as briefly stated in Herodotus (II.18 & 30) and Thucydides (I.104.1).

Emperor Justinian built many constructions and in particular the residences of the magistrates and baths.”

Of special interest is his use of the Greek form Taphosiris, and not the Latin corruption Taposiris, which was used by Strabo. Also the construction work of Justinian reveals the continued importance of the town more than five centuries after the time of Strabo, despite the total change of religion from paganism to Christianity.

The first detailed description of the Lake Mareotis region is found in Strabo (17.1.71) in his well known description of Alexandria where reference is also made to the southern region of Lake Mareotis. In a further significant passage, Strabo (17.1.15) comments on an important aspect of the economic value of the lake:

As regards the southern confines of Alexandria, Procopius reports (VI.1. 1-5): “The Nile River does not flow all the way to Alexandria, but after flowing to the town which is named Chaereou, it then turns to the left, leaving aside the confines of Alexandria. Consequently the men of former times, in order that the city might not be entirely cut off from the river, dug a very deep canal from Chaereou and thus by means of a short branch made the river accessible to it. There also, as it chances, are the mouth of certain streams flowing in from Lake Maria.

“The byblus (i.e. papyrus plant) grows in the Egyptian marshes and lakes, as also the Egyptian cyamus believed to be the lotus plant (Germer 1985: 39 ff.) from which was made the ciborium (a kind of vessel for drinking); it has stalks approximately equal in height, about 10 feet. But though the byblus is a bare stalk with a tuft on top, the cyamus produces leaves and flowers in many parts, and also a fruit like our cyamus, differing only in size and taste. Accordingly the cyamus-fields afford a pleasing sight, and also enjoyment to those who wish to hold feasts therein. They hold feasts in cabin-boats, in which they enter the thick of the cyami and the shade of the leaves; for the leaves are so very large that they are used both for drinking-cups and for bowls, for these even have a kind of concavity suited to this purpose; and in fact Alexandria is full of these in the work-shops, where they are used as vessels; and the farms have also this as one source of their revenues, I mean the revenue from the leaves.”

In this canal, it is by no means possible for large vessels to sail, so at Chaereou they transfer the Egyptian grain to boats (lemboi) which they are wont to call diaremata, and thus convey it to the city, which they are enabled to reach by way of the canal-route and they deposit it in the quarter of the city, which the Alexandrians call Phialè. But since it often came about that the grain was destroyed in that place by the people rising in sedition, the Emperor Justinian surrounded this district with a wall ….”

In this passage, the following points are of special interest: a) The freshwater canal from Chaereou received additional supplies from “certain streams flowing in from Lake Maria”. It is worthy of note that Procopius states that the freshwater canal to Alexandria branches off at a location he called Chaereou, whereas Strabo uses the name of Schedia. Surprisingly, other sources of the 6th century and later, use the name of Chaereou only. Significantly we can mention the Mosaic Map of Madaba which is also believed to date from the 6th century (Avi-Yonah 1954: 16-18). Another source is Stephanus Byzantius (677:18) also in the 6th century, who speaks of Chaereou as a city in Egypt. To all appearances, Schedia and Chaereou were two different locations close to one another, Schedia north of the canal and located on the Canopic Branch, whereas Chaereou is situated immediately south of the canal (Haas 1997: 25, top of map 2) and seems to have gained in importance at the time of Justinian when Stephanus called it a polis.

Furthermore, in connection with feasting, Strabo makes mention of Taposiris at the western end of Lake Mareotis and adds (17.1.14) that there they also: “hold a great public festival …. and near it there is a rocky place on the sea where likewise crowds of people in the prime of life assemble during every season of the year.” With this classical description in mind, we move on to some Byzantine sources in order to identify any similarities or differences in the descriptions attributed to the lake in later periods. An interesting passage is to be found in Procopius’ work on the “Buildings” of Justinian in the 6th century (VI.1.12). It deals with the regions to the west and south of Alexandria. He states that to the west: “In that territory, is a city one day’s journey distant from Alexandria, Taphosiris by name, where they say the god of the Egyptians, Osiris, was buried. In this city the

7

LAKE MAREOTIS: RECONSTRUCTING THE PAST b) In that canal the large Nile vessels could not sail so their cargo of grain had to be transferred to “diaremata” (light transport boats). This word was known only in Egypt; hence Procopius explained it by the Greek word “lembos”. The same word with the change of a vowel “dieremata” has been found in papyri (cf. Oertel 1917: 130). These dieremata were probably also used to navigate across Lake Mareotis.

The above discussed passages provide evidence that with the founding of Alexandria, Lake Mareotis acquired a lively and vital significance that continued into late antiquity. With its variety of flowering plants, the lake provided a pleasurable recreation resort for the Alexandrians who were, according to Strabo, in the habit of holding feasts in cabin boats. Even as far west as Taposiris, Strabo adds, a great public festival was held, obviously in connection with the cult of Osiris. More than five centuries later, Procopius reports that Justinian showed interest in that town “Taphosiris” by adding new construction work which is an indication of its continued importance.

c) The grain was eventually carried by light transport boats and deposited in the quarter of Alexandria called Phialè. This Phialè is not attested anywhere else and we have the impression that it was close to a landing-place near the canal. However, Procopius speaks of it as already existing and how it used to be plundered in times of trouble; hence Justinian surrounded it with a wall for protection.

It is of interest to know that the freshwater canal that links to Alexandria received additional supplies from “certain streams flowing in from Lake Maria”. This passing mention by Procopius provides additional evidence regarding the intricate hydraulic system of the freshwater canal. Furthermore, this canal provided the final facility for transferring grain from large Nile vessels – which could not sail in the canal - to light transport boats (dieremata).

Finally, we come to a text by Sophronius (early 7th century) in his account of the Miracles of Saints Cyrus and Johannes. The miracle happened to a certain Christodorus, deacon of the church of the two saints at the time of John the Almsgiver, then Bishop of Alexandria (610-619), whom Sophronius describes as “the beloved of the poor”. In his account, Sophronius has this to say (miracle no.8):

Finally, the lake and the regions surrounding it were of vital economic value to Alexandria, since the waters of the lake teemed with fish and water-fowl and the surrounding fertile lands flourished with large plantations until the early 7th century. Accordingly, Alexandria depended upon this neighbouring hinterland for the regular supply of its daily needs of meat and vegetables.

“Lake Mareia: You Alexandrians beloved of Christ, you all without exception know it well (i.e. the miracle), so also do a good number of Egyptians and most of the Libyans and as well as you, who are curious to know about Alexandria. Christodorus once sailed across Mareotis with the intention of inspecting the domains belonging to the church of the saints. A storm suddenly blew violently (the time was winter), agitating the waters of the lake and raising great waves. Christodorus was in great danger and thought he was going to perish. Nevertheless, he prayed to the saints who listened to his prayer …. And the wind at once dropped and he was saved.”

Bibliography Avi-Yona, M., 1954, The Madaba Mosaic Map. Jerusalem. Calderini, A., 1935-1987, Dizionario. Cairo, Madrid. C.Theod. = Codex Theodosianus, T. Mommsen & P.M. Meyer (eds.), 1904-5, 2 Vols. Berlin. Donner, H., 1992, The Mosaic Map of Madaba. Kampen. (http://198.62.75.1/www1/ofm/mad/legends/legends156.html. Accessed October 2009. Courtesy Prof. M. Haggag). Gascou, J. (transl. & ed.), 2006, Sophrone de Jérusalem, Miracles de Saints Cyr et Jean. Paris. Germer, R., 1985, Flora des Pharaoischen Aegypten, DAI. Sonderschrift 14. Kairo (with thanks to Prof. Mirvat Seif). Haas, C., 1997, Alexandria in Late Antiquity. Baltimore, London. Herodotus, Historiae, C. Hude (transl.), 1966. Oxford. Leontius Neapoleos, Life of John the Almsgiver. In E. Dawes & N.R. Baynes (transl. & eds.), Three Byzantine Saints: Contemporary Biographies of St. Daniel the Stylite, St. Theodore of Sykeon, and St. John the Almsgiver 1948. Oxford. Oertel, F., 1917, Liturgie. Leipzig. Palladius, The Lausiac History. C. Butler (transl.), 1904. Cambridge. Palladius, Historia Lausiaca. Migne, Patrologia Graecia 34. 1860, reprint 1998. Gottingen.

For our purpose however, the main point of interest in this passage is the mention of “the domains of the church”. In the Byzantine period, just as important temples of earlier times disposed of great wealth and property, so now the Church of Alexandria possessed considerable property. In 415 Emperor Theodosius, in a law on patronage, guaranteed the property of the Church of Alexandria which had been acquired before 397 (C.Theod. XI.23.6). The material wealth of the Church of Alexandria is also asserted by the fact that it owned a merchant fleet in the Mediterranean. Early in the 7th century the same John the Almsgiver, used that fleet to import corn from Sicily during a famine at Alexandria (Leontius Neapoleos, 13). However, throughout antiquity, the lake’s waters teemed with fish and water-fowl (P.Tebt. 3.867; Leontius Neapoleos, 8; Rufinus, 27.10). Alexandria depended on its neighbourhood for the supply of its daily needs of meat and vegetables. We may also recall that there were some 2,000 monks near Alexandria at Nitria, south of Lake Mareotis and that however frugal their diet might have been, they were amply sustained by products of the region (Palladius VII, 1-2; Sozomen, VI, 29). 8

M. EL ABBADI: LAKE MAREOTIS IN BYZANTINE TIMES Procopius, Buildings. H.B. Dewing (transl.), 1961. London. P.Tebt. = Tebtunis Papyri, vol. III.2, A.S. Hunt & J.G. Smyly (transl.), 1933. London Rufinus, Historia Monachorum. Migne, Patrologia Latina 21, 1849. Gottingen. Sophrone de Jérusalem, Miracles de Saints Cyr et Jean, J. Gascou (transl.), 2006. Paris.

Sozomen, Historia Ecclesiastica. Migne, Patrologia Graecia 67. 1864, reprint 2002. Gottingen. Stephanus Byzantius, Opera, 4 vols., K.W. Dindorf (ed.), 1825. Leipzig. Strabo, Geography, H.L. Jones (transl.), 1917. London. Thucydides, Historiae, H. Stuart Jones (transl.), 1897. Oxford.

9

LAKE MAREOTIS: RECONSTRUCTING THE PAST

10

I. AWAD: MARYUT LAKE THROUGH MAPS

A Study of the Evolution of the Maryut Lake through Maps Ismaeel Awad

This study concerns the evolution of the shape of Maryut Lake and its connections with the Nile River and the Mediterranean Sea from antiquity until the modern day by studying and observing maps of the Maryut Lake region, classified chronologically from the oldest to the most recent.

The aim of this study is to use maps to measure the evolution of the Maryut Lake, to test the value of using the map as a tool in the study, to evaluate the type of results obtained, and lastly, to try to forecast the shape and state of the Maryut Lake in the future. The Maryut Lake region consists geographically of five different parts (Fig. 1). In the north, a sandy strip extends along the Mediterranean coastline. To the south there is a calcareous ridge known as Abusir-Mexx that has an average altitude of 30 m high. The center of the Maryut region is divided into two sub-divisions, the first is the Maryut Basin which is situated towards the east with an average altitude attaining sea level; the second sub-division is the Maryut Valley situated towards the west of the region. And finally, the Maryut Ridge lies towards the south-west and has an average altitude of 50 m high.

In the following pages, two different types of available maps are considered. The first type of map represents edited illustrations of Maryut Lake by different historians and ancient cartographers who built up their illustrations through visual observation or by drawing. The study of these illustrations using the superimposition method, essentially overlapping one map on another, cannot be applied, as any attempt to do so would result in the production of inaccurate results based on the fact that these illustrations display approximate distances and inexact locations. The second type of map utilised topographic maps, which when edited using accurate topographic methods, can produce more accurate results. The application of topographic methods began during the French campaign at the end of the 18th century in which levelling and goniograph tools are used to construct maps in Egypt. Topographic methods have developed over time through to the current use of satellite images in mapping. The data extracted from these maps are accurate enough to be simply compared and overlapped.

About 20 maps and illustrations are used in this study, of which the oldest map was edited at the end of the 16th century and the most recent was extracted from a Landsat satellite image produced in 2001.There are two different ways of studying these different types of maps, as will be revealed during the course of this paper. Fig. 1: Maryut Lake region. Digitised by I. Awad. Archives CEAlex.

11

LAKE MAREOTIS: RECONSTRUCTING THE PAST Illustrations The illustrations utilised in this study date from the end of the 16th century until the end of the 18th century. Some of them were constructed ‘indoors’ without ever stepping foot in Egypt. They were done by using other maps edited by other publishers, used as base maps for adding new information. For this reason it is difficult to use these illustrations to derive any form of accurate reference or in application to an overlapping study.

• the evolution of names and locations of settlements such as towns or villages existing all around the Mareotic region. Illustrations edited before the end of the 16th century used to ignore the Maryut Lake as it was of little importance at that time, certainly when compared with the town of Alexandria. Ancient map makers rarely mentioned Maryut Lake, locating it at the margin of the map and illustrating its location in small letters. They portrayed an ignorance of the lake’s shape, such as for example the illustration edited in 1548 by Pierre Belon du Mans (Fig. 2).

However, some useful information could be extracted from examining these illustrations, such as: • the evolution of the connections between Maryut Lake and the Nile River, Nile branches and the Mediterranean Sea,

The illustrations used in this study are classified chronologically from the oldest to the most recent: Fig. 2: 1548, Vray portraict de la ville d’Alexandrie en Egypte, Pierre Belon du Mans. Jondet G., 1921, pl. II. Archives CEAlex.

Fig. 3: 1570, detail from Aegyptus antiqua, including a focused map of the region of Alexandria, Abraham Ortelius. Jondet G., 1921. Archives CEAlex.

12

I. AWAD: MARYUT LAKE THROUGH MAPS 1.01: 1570,1 Abraham Ortelius (Fig. 3)2 It can be observed from the general map that the Maria, et Mareotis lacus qui et Ara potes (Maryut Lake’s) shape took a direction from north-east to south-west incorporating eight islands as shown in the figure.

River), which was also connected to the Nile River at its south-eastern end. The boats depicted in the Moeris Swamp show that there was extensive navigation during this period between Moeris Swamp, Maryut Lake and the Nile River.

Maryut Lake had two connections, the first was the connection with the Myris quod et Moeris stagnum (Moeris Swamp) to the south through the Lycus flu(vius) (Lycus

The second connection related directly to the Nile River in three places:3 the first one to the north related to the Canopica fossa4 (Canopic Branch), the second connected the lake with the Canopic Branch at the north-east near the town of Schedia, and finally, the third connected the lake directly with the Nile River at its eastern side. 1.02: 1588, Livio Sanuto (Fig. 4)5 In this map, the Bycheira lacus (Maryut Lake) has deformed in its shape forming a triangle enclosing a large number of islands (21 islands). Just one connection linked the lake to the Canopic Branch at its eastern end while no connection was illustrated with the Moeri dis lacus (Moeris Lake). 1.03: 1655, Nicolas Sanson D’Abbeville (Fig. 5)6 The map of D’Abbeville has similar limitations to the previous map. The shape of Maryut Lake is similar although more rectangular and adopting more of a northwest to southeast orientation. However, most of the islands that had been previously observed have disappeared. 1.04: 1717, Paul Lucas (Fig. 6)7 No connections are shown with the Lac de la Mareote (Maryut Lake) in this map while some islands reappeared (nine in total). The Moeris Lake has totally disappeared, and at the same time the Lac du Madie (Madieh Lake/Abuqir Lake) is illustrated for the first time.

Fig. 4: 1588, detail from Africae Tabula VII, Livio Sanuto. Archives Biblioteca Nazionale Centrale Di Roma.

1.05: 1722, Claude Sicard (Fig. 7)8 In Sicard’s illustration the Mareotis Palus (Maryut Swamp’s) form has changed again to resemble the shape of a bird or a duck. Some connections are apparent between Maryut Lake and the Nile to the east that joined together before flowing into Maryut Swamp. A small lake named Nitraria Palus9 (Nitraria Swamp) was identified to

Fig. 5: 1655, detail from Royaume et désert de Barca et l’Egypte, Nicolas Sanson D’Abbeville. Commercial reproduction.

1. There is another illustration of Ortelius dated from 1595, found in a late commercial reproduction from the beginning of the 21th century. 2. Aegyptus antiqua, map of the Antique Egypt. 3. There is a fourth connection shown in the focused map of Alexandria and not in the general map. It is situated between the first and the second connections at the north passing near the village of Chabrie pagus. 4. Mentioned in the focused map of Alexandria. 5. Africae Tabula VII, map of Africa, sheet no. 7. 6. Royaume et désert de Barca et l’Egypte, Kingdom of the desert of Barca and Egypt. 7. Carte de la Basse Egypte et du cours du Nil, map of Lower Egypt and the Nile River. 8. Carte de l’Egypte Ancienne, map of Ancient Egypt. The original map edited by Claude Sicard was not found. This is a synthesis of two copies extracted from the original map done by D’Anville and Danlisle and the place name’s list of Sicard. Sauneron & Martin 1982. 9. There are two lakes named Nitraria in the map: one near the Maryut Lake which is mentioned in the study and the other located in Wadi Natrun between Alexandria and Cairo.

13

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig. 6: 1717, detail from Carte de la Basse Egypte et du cours du Nil, Paul Lucas. Archives IFAO.

Fig. 7: 1722, detail from Carte de l’Egypte Ancienne, Claude Sicard. Sauneron & Martin 1982, Archives IFAO.

Fig. 8: 1743, detail from Carte d’Egypte, Richard Pococke. Archives IFAO.

Fig. 9: 1753, detail from Carte de l’Egypte ancienne et moderne, Robert de Vaugondy. Archives J.-Y. Empereur.

the south-east of Maryut Lake, connected with it through Acheron Canalis (Acheron Canal). The town of Marea is illustrated, but strangely to the north-west of Maryut Swamp and not on the southern shores of the lake as located nowadays.

that used to connect Moeris Lake with Maryut Lake. The towns of Plinthine.S. and Taposiris.P.11 were indicated but very far from the lake towards the west. 1.07: ca 1753, Robert de Vaugondy (Fig. 9)12 This 18th century map indicates yet a further change to the shape of Birk Mariou (Maryut Lake) now depicted as being aligned from north-east to south-west. There is still

1.06: 1743, Richard Pococke (Fig. 8)10 The shape of Birk Mariou.s./Lacus Mareia.f./Mariotis.S. (Maryut Lake) changed again in this map taking a triangular form. A lake is mentioned at the place of the Nitraria Lake but located further south than before, and the Moeris Lake appeared as Birk El Karoon/Lacus Moeris feu Moerius (Karoun Lake).

10. Carte d’Egypte, map of Egypt. 11. Known as ancient navigation ports situated actually at the western arm of the Maryut Lake. 12. Carte de l’Egypte ancienne et moderne, map of Ancient and Modern Egypt. The base map is inspired from Claude Sicard’s map.

Bahr Bellomah i.e Mare seu Lacus Sine Aqua (lake or sea without water) was mentioned at the place of the old canal 14

I. AWAD: MARYUT LAKE THROUGH MAPS one easterly connection between Maryut Lake and the Nile River. A lake is mentioned and located at the same place of the Nitraria Lake as in the previous map. 1.08: 1762, Rigobert Bonne13 The Bonne map has many similarities with the previous map, with just one change which is the appearance of the town of Abusir located at the middle of the northern shoreline of the Maryut Lake. 1.09: 1764, Jacques Nicolas Bellin (Fig. 10)14 A connection between Lac Mareotis (Maryut Lake) and the Mediterranean Sea is shown for the first time to the west of Alexandria, with the disappearance of Lake Nitraria. Two connections appeared between Maryut Lake and the Nile (both meet at the same point, southeast of the lake). One of them approached from the east and the other from the Bahr-el-Helame ou Fleuve seché (canal or sea without water) from the south-east, and passed through a Lac de Sel (Salty Lake)15. 1.10: 1799, Edme Mentelle & Pierre-Gregoire Chanlaire16 In this map of 1799 Sebaga ou Birk Marioul17 (Maryut Lake) reappeared in the form of a triangle with what appears to be a sea connection to the west of Alexandria. The connection with the Nile River appeared through the Canal de Bahiré (Bahirey Canal) flowing into the Maryut Lake, and Lac de Natron (Natron Lake) appeared at the place of the ex. Nitraria Lake southeast of Maryut Lake. The Bahhr-bé-lâ-mâ ou Fleuve sans eau (canal without water) was also shown approaching from the Birket-il Kerun (Karoun Lake) which is located to the south in the Feïum region, extending towards the southwest of Maryut Lake but without any intervening connections.

Fig. 10: 1764, detail from Carte des Embouchures du Nil, et partie de son cours, Jacques Nicolas Bellin. Archives J.-Y. Empereur. • geo-referencing the maps in the same coordinate system, • digitising each map in AutoCAD program, • overlapping each digitised map one on the other, • analysing the digitised features in ArcGIS program.

Towns are named around the Maryut Lake such as the town of Mariout (Marea) located on the northern shoreline of the lake and Tour des Arabes (town of Abusir) near to the Mediterranean Sea.

2.01: 1801, French campaign, scale: 1:100.000 (Figs. 11 & 12)18 This is the first accurate topographic map concerning the Mareotic region. It was edited by the engineers of the French campaign who came to Egypt between 1798 and 1801. The map was constructed by accurate topographic instruments including a level and a goniograph.

Topographic Maps In order to extract precise locations and establish an accurate surface area, four steps were undertaken when studying the second type of map:

In this map, a cut is shown in the dike separating the Bohéïréh Marîoût ou lac Mareotis (Maryut Lake) from the Bohéïréh Maadîéh, lac Madié ou d’Aboûqîr (Madieh or Abuqir Lake). The dike was breached by the English army in April 180119 during the war between the French

13. Carte de l’Egypte ancienne et moderne, map of Ancient and Modern Egypt. 14. Carte des Embouchures du Nil, et partie de son cours, map of the Nile river mouths, and a part of its stream. 15. The Salty Lake is at the place of the Nitraria Lake situated in Wadi Natrun. 16. Focused map Carte particulière et détaillée du Delta (Detailed map of the Delta) in the Carte physique et politique de l’Egypte (Physical and political map of Egypt). There is another similar map Carte d’Egypte (map of Egypt) edited by Jean-Baptiste Bourguignon D’Anville in 1765. Charles-Roux 1910. 17. Sebaga is near to Sabkha ‫ ةخبس‬in Arabic language which means marsh, and Birk comes from Birka ‫ ةكرب‬in Arabic, too, which means pond.

18. Carte d’Alexandrie, sheet no. 37 extracted from the Atlas of the French Expedition Carte topographique de l’Egypte et de plusieurs parties des pays limitrophes. 19. The dike was breached 19 April, 1801, by the English army under the orders of General Sir John Hely-Hutchinson, in order to cut off the freshwater supply brought by the Alexandria canal from the French garrison located in Alexandria, and to cut the road relating them to the main French army situated in Cairo.

15

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig. 11: 1801, Carte d’Alexandrie, French campaign. Archives J.-Y. Empereur. Fig. 12: 1801, Carte d’Alexandrie, French campaign, original scale 1:100,000. Digitised by I. Awad. Archives CEAlex.

and English. As a result, the water in Abuqir Lake (which was situated at a higher altitude than the Maryut Lake and was directly connected with the Mediterranean Sea), was redirected into Maryut Lake, causing flooding in its eastern (125 km²) and western (90 km²) extremities.

The depth of Maryut Lake was measured in units of feet using depth soundings all around the lake. We can observe from the depths shown on the map that the deepest zone in the Lake was the north-eastern zone that was an average depth of about 4.5 m. Maryut Lake becomes shallower

Two lake limits are represented on the map, the first referred to the limit of the flood on 9 August, 1801,20 and the second showed the limit of the lake after the flood had returned to its preceding limit on 10 January, 1802.21

20. The flood limits were obtained from M. Tawiland’s map (scale: 1:128.000), engineer in the English army. Jacotin 1809-1828: 83. 21. Due to the effect of the northerly wind and to the reduction of water in the Nile River branches. 22. Maybe it was the connection mentioned before in the ancient illustrations between the Mediterranean Sea and the Maryut Lake at the west of Alexandria.

Another dried up old cut22 is illustrated linking Maryut Lake with the Mediterranean Sea to the east of the Anse de Marabou (Gulf of Marabou), west of Alexandria. 16

I. AWAD: MARYUT LAKE THROUGH MAPS

Fig. 13: 1866, Carte des Environs d’Alexandrie, Mahmoud Bey El-Falaki. Archives CEAlex. Fig. 14: 1866, Carte des Environs d’Alexandrie, Mahmoud Bey El-Falaki, original scale 1:200,000. Digitised by I. Awad. Archives CEAlex.

time before the flood which was about 890 km² (211.300 feddans). At this point, the lake’s maximum width (from north to south) was about 45 km, and its maximum length (from east to west) reached about 56 km.

from the north-eastern zone towards the south (2 m in average) and the west (1.5 m in average) until the lake attends 0 m at its limits. In addition, two levelling profiles23 were undertaken across Maryut Lake, one at the Maryut Valley24 and the other along the dried old cut mentioned previously.

23. Le Père 1809-1828: Pl. 19. 24. Between the ruins of the town of Marea and the Mediterranean Sea passing through the saint tomb of Aboû el Keïr.

Thanks to digitising, we are able to ascertain with sufficient accuracy the precise surface area of the lake at that 17

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig. 15: 1911 (1914), Atlas of Egypt, Vol. I, Lower Egypt, Survey of Egypt (G. Daressy), original scale 1:50,000. Digitised by I. Awad. Archives CEAlex. Fig. 16: 1949, map series 1:100.000. Survey of Egypt, original scale 1:100,000. Digitised by I. Awad. Archives CEAlex.

2.02: 1866, Mahmoud Bey El-Falaki, scale: 1:200.000 (Figs. 13 & 14)25 This map is interesting because El-Falaki (1872) mentioned that Lac Maréotis (Maryut Lake’s) shape at that time (1866) was similar to the lake’s shape when the town of Alexandria was founded in antiquity, according to Strabo’s notes.26 The author also defined the ancient branches of the Nile as the Canopic Branch and Schedia Canal.

Lake was also reduced in area to 450 km² (which is about 50% of its area as indicated in 1801). This reduction in area occurred in just 65 years, and was especially apparent in the eastern part where its maximum length was about 24 km.

25. Carte des Environs d’Alexandrie, map of the surroundings of Alexandria 26. El-Falaki 1872: 99.

From this map, it can be observed that the Lac d’Aboukir (Abuqir Lake) has reduced in area. Proportionally Maryut 18

I. AWAD: MARYUT LAKE THROUGH MAPS

Fig. 17: 1970s, map series 1:25.000, Egyptian General Survey Authority, original scale 1:25,000. Digitised by I. Awad. Archives CEAlex. 2.03: 1911, Survey of Egypt, scale: 1:50.000 (Fig. 15) 27 The area of Maryut Lake continued to reduce particularly on its eastern margins. The Egyptian government began to have some interest in this zone by managing a system of irrigation canals and drains (for example: El-Omum Drain, whose waters flowed from the south-eastern side of the Maryut Lake) in order to cultivate the region around Maryut Lake. The area of the lake was reduced to 340 km² (38% of 1801s area). However, the dimensions of the lake from east to west appear to have extended compared to previously (about 75 km). The ancient cut between the Mediterranean Sea and Maryut Lake to the west of Alexandria is again mentioned.

north-eastern shores, particularly with the construction of Nozha30 and the seaplane base. By now most of the western part of the lake had dried up, perhaps partly due to the fact that this map was surveyed in the dry season.31

2.04: 1930, Survey of Egypt, scale: 1:100.000 28 The area of Maryut Lake continued to reduce in size particularly towards the south-east as the lake became increasingly exploited for agriculture purposes. An artificial salt marsh appeared for the first time in the middle of the lake in order to produce salt.

A great deal of development had also occurred in the lake:

2.06: 1970s, Egyptian General Survey Authority, scale: 1:25.000 (Fig. 17) 32 By the 1970s a big development had occurred at the eastern and southern parts of Maryut Lake with an expansion of irrigation canals and drains. The Nubareya Canal was completed in 1976 channeling waste water runoff from agriculture lands into the lake, and then into the Mediterranean Sea via the Mexx pump station.

In the middle of the lake, the salt marshes became much larger (it has expanded ten fold in just 20 years). Big changes also took place in the north-eastern part of the lake. The seaplane base was transformed into a big fish farm, and Nozha airport became a part of the land in order to separate the new fish farm from Maryut Lake.

2.05: 1949, Survey of Egypt, scale: 1:100.000 (Fig. 16) 29 The lake was further reduced on its eastern shores. This map also indicates early interest in expansion along its

Maryut Lake’s surface area has reduced further to just 210 km² (51.000 feddans) (which is about 1/4 of its area in 1801).

27. The maps of this series have been collected from: Survey of Egypt 1914. Eight maps have been used in digitising: sheet no. 009 Ramleh, sheet no. 010 Abuqir, sheet no. 020 Mallaha, sheet no. 021 Alexandrie, sheet no. 022 Kafr El Dawar, sheet no. 046 Hammam, sheet no. 047 Bahig and sheet no. 048 King Mariout. 28. German edition, four maps have been used in digitising: 88/42 Hammam, 88/48 Ghayata, 92/48 Alexandria and 92/54 Damanhur. 29. Arabic edition, four maps have been used in digitising: 0°31’/29°40’ Abu El-Matamir, 31°/29° Krair, 31°/29°40’ Alexandria and 31°/30°20’ Damanhur.

. 30. Known at that time as Fouad the First airport: ‫لوألا داؤف ءانيم‬ 31. The western zone’s water level is directly influenced by dry and wet seasons. 32. Fourteen maps have been used in digitising: 90/450 Hammam, 90/465 Mahatet El-Gharbaneyat, 91/450 Kom Makhoura, 91/465 Burg El-Arab, 92/465 Burg El-Arab, 91/480 Mahatet Bahig, 91/495 Aeroport of Amreya, 92/480 Kreir, 92/495 Amreya, 93/480 El-Deir, 93/495 Mexx, 93/510 Kom Echo, 94/510 Ramleh and 94/525 Munshaat Buline.

19

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig.18 : 1991, map series 1:50.000, Egyptian General Survey Authority. Digitised by I. Awad. Archives CEAlex. Fig. 19: 2001, Landsat satellite image, resolution 15m/pxl. Digitised by I. Awad. Archives CEAlex.

2.07: 1991, Egyptian General Survey Authority, scale: 1:50.000 (Fig. 18) 33 A huge development occurred in the southern and eastern parts of the Mareotic region in the early 1990s which enabled the entire lake to be part of the same irrigation system, and large areas of the middle zone of the lake were transformed into fish farms.

2.08: 2001, LANDSAT satellite image, resolution: 15m/pixel (Fig. 19) 34 Satellite images were used to study more recent changes in the lake morphology due to the unavailability of more

These were the last map series produced concerning the Mareotic region, and edited by the Egyptian General Survey Authority.

33. Six maps have been used in digitising: NH35-L5d Hammam, NH35L6c Burj al-’Arab, NH35-L6d King Mariout, NH35-P3a Sidi Kirayr, NH35-P3b Alexandria and NH36-M1a Kafr El-Dawar. 34. Downloaded from the University of Maryland site (www.umd.edu).

20

I. AWAD: MARYUT LAKE THROUGH MAPS Table 1: Comparison table between different periods of Maryut Lake.

Fig. 20: Evolution diagram of Maryut Lake’s surface areas (1801-2001).

contemporary maps since the last map series was produced in the 1990s.

as the construction of fish farms and artificial salt marshes, all of which require stable quantities of water.

A topographic map was extracted from a LANDSAT satellite image of a medium resolution of 15 m/pixel, which is enough to digitise the limits of the lake, marshes, fish farms, wide canals and drains, but not enough to define the other small features such as buildings, narrow canals, secondary drains or roads.

On the other hand, natural influences to the shape of Maryut Lake have a limited impact nowadays (such as rain and ground water). However, any loss in the lake’s surface area in the future depends on the rate and types of uses of the lake and the implemented management plan. Forecasting the Shape of Maryut Lake Digitising the topographic maps does provide an opportunity to forecast the future shape of the Maryut Lake region, if left to evolve naturally without any external interference. A forecasting study is undertaken by extracting and digitising different contour levels of the Maryut Lake region. The maps utilised in this digitising exercise were the map series of scale 1:25,000 edited in 1940s, and the same map series edited in 1970s; these maps series were chosen because they are the most accurately contoured maps that covered the whole Mareotic region. The accuracy of the results of such an exercise is after all dependent upon the precision of the maps used, and on the data extracted.

The map showed that the area of Maryut Lake’s is still about 210 km² (51,000 feddans). Its maximum dimensions attended 66 km in length and 11 km in width. From the table (Table 1) and the diagram (Fig. 20) showing the evolution in surface areas of Maryut Lake from 1801 until 2001, we can notice that the lake has lost 75% of its surface area in just 200 years. The greatest loss occurred between the period from 1801 to 1866, in which the lake lost half of its area due to drying and land fill. Maryut Lake became more stable with regard to its rate of surface area lose from the 1970s until now. During this 30 year period it lost 0.1% of its surface area. The main contributor to the reduced rate of water loss was irrigation and the increase in activities around the shores of the lake, such

The 0 m contour line was the first digitised, which refers to the limits of the Maryut Basin, which is the same as sea level (0 m). By comparing this contour map with Maryut 21

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig. 21: Comparison between Maryut Lake 1801, French campaign map series 1:100,000 (top) and 0 m level (bottom) in the Mareotic region. Digitised by I. Awad. Archives CEAlex. Results (Fig. 22) The surface area of Maryut Lake reduced by less than a quarter (24%) over a period of 200 years (1801-2001). This does not bode well for the future of Maryut Lake, particularly when you consider the intensive construction projects in and around the lake especially at its eastern end, its northern and at its north-western parts having been continuously divided up, dried out and filled to construct tourist villages and settlements.

Lake’s limits in 1801 (without flood), it can be seen (Fig. 21) that they are approximately identical. That means that the lake’s limits in 1801 had reached the 0 m level. Using this kind of study, the Maryut Lake’s limits can be compared to any topographic map with the different digitised contour levels such as the levels of 0 m, -1 m, -2 m and -3 m, in order to study the lake water’s level in any period. We chose these precise levels because the deepest altitude of the Mareotic region is about -4 m which is limited to a small area in the north-eastern part of the lake and the highest level that the lake’s water reaches did not exceed 1 m.

The use of Maryut Lake itself such as fish farms or salt marshes has increased from 1 km² (0.3% of Maryut Lake’s area) in 1930 into 39.6 km² (18.6% of Maryut’s Lake area) 22

I. AWAD: MARYUT LAKE THROUGH MAPS

Fig. 22: Comparison between Maryut Lake 1801-2001. Digitised by I. Awad. Archives CEAlex. Charles-Roux, F., 1910, Les origines de l’Expédition d’Egypte. Paris. Carte d’Alexandrie, 1801, French campaign. Carte topographique de l’Egypte et de plusieurs parties des pays limitrophes [MAP]. El-Falaki, Mahmoud Bey, 1872, Memoire sur l’antique Alexandrie, ses faubourgs, ses environs. Copenhagen. Jacotin, M., 1809-1828, Mémoire sur la construction de la carte de l’Egypte. Description de l’Egypte, Etat Moderne, Tome 2, 2ème Partie. Paris. Jondet, G., 1921, Atlas historique de la ville et des ports d’Alexandrie. Cairo. Le Père, G., 1809-1828, Planche 19, Vallée du Nil et Lac Maréotis. Description de l’Egypte, Antiquités, A. Vol. 5. Paris. Livio Sanuto, 1588, Africae Tabula VII. Biblioteca Nazionale Centrale di Roma [MAP]. Lucas, P., 1664-1737 (1724), Voyage du Sieur Paul Lucas, fait en MDCCXIV & par ordre de Louis XIV, dans la Turquie, l’Asie, Sourie, Palestine, Haute et Basse Egypte. Paris. Pococke, R., 1745, Observations on the Islands of the Archipelago, Asia Minor, Thrace, Greece and some other parts of Europe. A Description of the East and some other countries. Vol. II. Part II. London. Sauneron, S., & Martin, M., 1982, Oeuvres III - Parallèle géographique de l’ancienne Egypte et de l’Egypte Moderne. Cairo. Survey of Egypt, 1914, Atlas of Egypt, Vol. I, Lower Egypt, Cairo. de Vaugondy, R., 1753, Carte de l’Egypte ancienne et moderne [MAP].

in 2001. That means that the lake’s use has increased about 40 times in just 70 years for salt production. Conclusion The map can be a very useful tool in a study of temporal evolution of any geographic feature, such as the case of Maryut Lake. When assessing the results of such studies, the following conditions must be considered to increase the reliability of the product: • Data must be accurate to all the region of study (the more accurate the data the better the results). • Data (maps and satellite images) must be accessible. This is a particular problem when investigating the Maryut Lake region where some maps of some of the series were not found due to negligence of archiving or due to security reasons.35 • Data must be updated regularly to make the study of the region reliable and more accurate. In the case of the present study, the last series of maps produced concerning the whole region was in 1991. This means that for the last 17 years, no other map series concerning the Mareotic region have been produced, despite continued development and huge observable changes having occurred in the Mareotic region since this time. Bibliography D’Abbeville, N.S., 1655, Royaume et désert de Barca et l’Egypte [MAP]. Bellin, J.N., 1764, Carte des Embouchures du Nil, et partie de son cours [MAP].

35. Several military zones exist around the Mareotic region causing the exclusion of some maps which include these.

23

LAKE MAREOTIS: RECONSTRUCTING THE PAST

24

L. BLUE: LAKE MAREOTIS RESEARCH PROJECT

Lake Mareotis Research Project Lucy Blue

Alexandria was by any standards one of the great cities of the Mediterranean. Since its foundation in 331 BC and for almost a millennium to follow, it was the political, economic and cultural capital of Egypt, and one of the most significant emporia and complex ports in the Hellenistic and Roman worlds. As such, Greco-Roman Alexandria has been the subject of much historical and archaeological research (Fraser 1972; Haas 1997; Empereur 1998; Goddio, et al. 1998) that has revealed the wealth of the city, much of which was generated by trade through its important and monumental harbour. However, the complexity of the Alexandria harbours system, which included not only harbours on the sea, but also on Lake Mareotis, has never been thoroughly understood. The significance of Lake Mareotis in the history of the Greco-Roman port-city of Alexandria, is undeniable (Fig. 1). Settlements and industrial units located along the shores of Lake Mareotis, located to the south and west of the city, are known to have produced amongst other things, glass, textiles, pottery and wine (Empereur & Picon 1998), all of which were transported to the city. Yet the dynamics of this productive hinterland have not been fully appreciated in terms of the important contribution the region made to the economy of Alexandria and thus to Egypt as a whole.

The principal objective of the Lake Mareotis Research Project was to undertake a detailed systematic survey of the western extension of the lake in order to record, quantify and assess the archaeological resources of the area in a comprehensive manner. As a result, we hoped to determine a better understanding of the nature and extent of economic and maritime activities in the Mareotic region and to explore the relationship between the component parts of this complex system and how they developed and changed over time, in order to determine a more comprehensive understanding of the economic functions of ancient Alexandria.

In this context, the Lake Mareotis Research Project set out to investigate and determine the role of the lake, the function of the sites around its shores and the part they played in supporting the metropolis of Alexandria in antiquity.

In ancient times, Lake Mareotis was connected to and fed by the Canopic Branch of the River Nile (Strabo 17.1.7). By the 12th century the lake had silted up but prior to its silting it was much larger and deeper (Warne & Stanley 1993: 53-8; Said 2002). The main body of Lake Mareotis currently covers an area of about 90 km2 and is defined to the north by a chain of lithified carbonate ridges extending parallel to the northern shoreline of the lake and the Mediterranean coast (Fig. 2). It is nearly rectangular in shape and extends south of Alexandria for about 12 km, merging along its southern and eastern shores into the Western Deltaic region. It would once have extended considerably further south, quite possibly a navigable distance of about 50 km. It is separated from the lake’s main body by causeways and shallows and its relative size varies according to the season.

Environmental and Political Context of Lake Mareotis Lake Mareotis represents one of the most distinctive geomorphological features along the north-west coast of Egypt (Warne & Stanley 1993). It is unique compared to other lakes on the north coast of Egypt, being the only one with no direct connection to the Mediterranean Sea, but originally with direct access to the River Nile and hence to the whole of Egypt. Therefore, it is believed that the location and characteristics of Lake Mareotis gave the city of Alexandria one of its major advantages (Strabo 17.1.7).

Context of the Lake Mareotis Research Project The Lake Mareotis Research Project is a collaboration between the Centre for Maritime Archaeology, at the University of Southampton, and Department of Underwater Antiquities (DUA) of the Supreme Council of Antiquities (SCA), Alexandria. The project first started in 2004 when a British Academy Small Grant was awarded to support a pilot survey season in the western arm of the lake, to the west of Alexandria. The survey covered an area of approximately 40 km east-west by 3 km wide, along the shores of the western extension of Lake Mareotis, from Sidi Kerir to El-Hammam. During this first season over 60 sites where identified, many equating to new discoveries (Blue & Ramses 2005). The potential wealth of the sites in the area was thus realised and subsequently further funding was sought and awarded from both the British Academy and the Leverhulme Trust to conduct a more detailed survey on both shores of Lake Mareotis and on an island that is located in the eastern region of the survey area.

Thus at present, Lake Mareotis is a body of shallow brackish water that occupies about 13% of an extensive subsea-level depression that is sometimes referred to as the Maryut Depression. The remaining 87% of the depression has been drained and is now used primarily for agriculture. The main body of the lake is heavily polluted by untreated sewage and industrial waste. It has also been divided into many sub-basins which are used for industrial and agricul25

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig. 1: Location of Alexandria and Lake Mareotis (Lake Mareotis Research Project).

26

L. BLUE: LAKE MAREOTIS RESEARCH PROJECT tural purposes. Moreover, the shores of the lake are subject to continuous irrigation, drainage and reclamation, which means that it is constantly changing and unstable (Warne & Stanley 1993: 29-30; Frihy, et al. 1996: 282). The average depth of the remaining lake is less than 1 m, although it is assumed to have been greater in antiquity to allow boats to sail safely.

and Boussac & El Amouri (this volume) on the Hellenistic port of Taposiris Magna. In addition, rescue work has been undertaken by the Supreme Council of Antiquities (SCA) at a number of smaller sites around the shores of the lake largely in advance of development. The majority of known sites have not been systematically studied however, and the pilot survey conducted in 2004 revealed that only a fraction of the available evidence is currently recognised. In addition, as indicated previously, the area is massively under threat from development and pollution. Large sections have been quarried, particularly of the northern limestone ridge of the study area, while extensive sections of the remaining lake are being exploited for aquaculture and its shores cultivated for agriculture. Much of the land has been sold for development, and along the edges of the lake harmful polluting industries such as salt factories, geochemical and petrochemical processing plants, as well as numerous sewage plants, all present a real threat to this lacustrine environment – the landscape is constantly changing and with it the archaeological record is being destroyed (see Ramses & Omar this volume). For all these reasons it was determined that a comprehensive survey of the area was essential in order to record what remains before the information is lost forever.

Due to its location and characteristics, Lake Mareotis played a significant role in the internal and international transport system of Greco-Roman Egypt (Rodziewicz 1998), particularly after it had been connected to the sea by a navigation canal at Alexandria (Strabo 17.1.7: 31). It became an important conduit of communication between the River Nile and the Mediterranean Sea, and river boats would have transported trade goods from the Nile Valley to the harbours of Mareotis and onwards to Alexandria and beyond. As indicated, Mareotis also supported economic activities around its shores such as the cultivation of vines and grain, and the manufacture of ceramics, wine and glass (Empereur & Picon 1998). However, despite the important role of Lake Mareotis, previous research conducted in the region has been largely limited to specific areas and topics, such as the work of El-Fakharani (1983) amongst others, on the Byzantine port of Marea (Philoxenite) (see now Haggag, Babraj & Szymańska, Rodziewicz and Pichot this volume), Empereur & Picon (1998) on the amphorae and wine production in Mareotis (also see Dzierzbicka this volume),

Objectives of the Lake Mareotis Research Project The objectives of the collaborative Lake Mareotis Research project were to:

Fig. 2: An approximation of the ancient and present limits of Lake Mareotis (E. Khalil). 27

LAKE MAREOTIS: RECONSTRUCTING THE PAST • Systematically survey the western arm of Lake Mareotis (shores 40 x 3 km wide, and Mareotis Island,1 3.7 km in length and up to 680 m wide), bearing in mind the complex palaeogeographical history of the lake. • Record, quantify and assess the archaeological resource. • Determine the extent, chronology and, where possible, function, of each site. • Determine the nature and the extent of the economic and maritime activities of the region. • Determine the significance of the Mareotic region in relation to Alexandria, particularly during the Hellenistic, Roman and Byzantine periods when the lake is believed to have been most active. • Identify the degree of threat to the archaeological resource in relation to each site before it is lost to development and pollution.

• A Fluxgate Radiometer Magnetometer geophysical survey was conducted to establish the layout of structures beneath the surface at specific sites.2 • Ceramic survey was undertaken to determine the chronology, nature and function of each site. • Sedimentological survey to determine changing lake levels and identify geomorphological changes was conducted at selected sites. • Limited site-clearing to articulate walls and identify more detailed building phases, was undertaken. • A photographic archive was compiled of each site and features within the site. Analysis of the five survey seasons conducted in the western arm of Lake Mareotis between 2004 and 2008 is almost complete, and brief preliminary observations are presented in this paper. The final results will be published in a single monograph detailing the work undertaken and an interpretation of the final results (Blue & Khalil forthcoming). In total, over 70 sites were documented, ranging in date from the Hellenistic period to the 7th century AD (Fig. 3). A clear distinction was noted between the nature of the sites on the northern shores compared to those in the south and on Mareotis Island. The north shore sites were clearly defined, but quite heavily eroded, tell sites. The sites to the south appear to represent more dense urban concentrations with associated agricultural and industrial areas, particularly towards the east where the island ridge may have once been an extension of the southern shore and the ridge associated with the site of Marea. Many of the sites, particularly on the southern shore, were partially buried under aeolian and alluvial sediments deposited by the eroding hillside and wind-blown material from the north.

Methodology As a result of the pilot survey conducted in 2004 a total of over 60 sites were identified, as previously mentioned. Some of these sites were already documented in the archaeological record, some having been partially excavated by the Western Nile Delta section of the SCA and foreign missions, or noted but not recorded in detail in previous publications. However, at least half of the sites identified are believed to be new discoveries. The sites were prioritised on the basis of their maritime significance, particularly where maritime installations such as moles, jetties, quays, mooring rings etc., had been identified. GPS coordinates were taken to fix the position of each site located, each site was photographed and basic data pertaining to its nature were recorded. Since 2004 four further seasons of survey have been conducted between autumn 2006 and summer 2008, with the objective of recording in detail the maritime installations and to place the sites in their environmental and economic context. In order to achieve these objectives, the following methodology was implemented: • A desk-based assessment of the region was undertaken with the aid of high resolution Quickbird satellite images, to establish the nature and extent of the sites already identified. This information along with all the data acquired as part of the survey project, is archived in a GIS-based database. • Each site was surveyed using a GPS-based Real Time Kinematic (RTK) satellite navigation system to produce detailed topographic maps of each site. • Specific buildings and/or features were planned using a Leica TCR705 Total Station, and the data was downloaded in the field into AutoCAD via TheoLt software.

North Shore Sites (Fig. 3) A total of ten sites were identified on the northern shore. They date from the Hellenistic to Late Roman periods with a concentration of activity in the Roman period. They were primarily tell sites and their size ranged from c. 7000 to 180,000 m2. The ceramic assemblages at the sites included a mixture of amphorae, finewares, tablewares and cooking wares. Many of the finewares and amphorae were imports. Many of the sites on the northern shore contain wells and cisterns as well as evidence for mortar-lined basins, probably from bath complexes, which would suggest a domestic context. Limited evidence of irrigation exists in the form of water wheels or sakkias, and qadus pots associated with water wheels or sakkias were found at some Early and Late Roman sites. The evidence suggests that these sites were largely domestic in nature and each site represents a fairly contained and distinct unit with a clear association with

2. Preliminary investigations in the 2007 season were not as insightful as we had hoped, due to the fact that the fluxgate radiometer FM36 we had been using had limitations. Thus, the 2008 geophysical survey was conducted with a Bartington Instruments Grad 601-2 dual sensor fluxgate gradiometer. The dual sensors survey at twice the rate of a single sensored instrument and with more refined results.

1. The Lake Mareotis Research Project refers to this ridge, largely surrounded by water located in the south-eastern region of the survey area some distance from the southern shore of the lake, as Mareotis Island. However, its island status in antiquity has yet to be determined and it is also acknowledged that it is not, and was not in antiquity, the only island in the lake.

28

L. BLUE: LAKE MAREOTIS RESEARCH PROJECT

Fig. 3: Sites located along the northern shoreline, the southern shoreline and on Mareotis Island (Lake Mareotis Research Project). sites extend across the lake towards each other. Thorough investigation of the ceramic samples collected from these sites will shed more light on their possible nature and relationship.

the waterfront. A number of the settlements featured jetties extending into the lake on their southern shores. The survey also revealed possible spatial and functional relationships between a number of sites on the northern shore and specific sites on the island and along the southern shore. For example, it was noticed that at least one of the largest sites on the northern shore of the lake (AlGamal, Site 204) is located immediately opposite a settlement located at the western end of the island (Site 23; see Hopkinson this volume). Quays and jetties from these

South Shore Sites (Fig. 3) A total of 26 sites were identified on the southern shore. The sites ranged in date from the Hellenistic to the Late Roman periods, but generally exhibited limited activity in the Early Roman period. The earlier period displays some imports, largely amphorae from the Aegean region, but 29

LAKE MAREOTIS: RECONSTRUCTING THE PAST

is predominantly composed of locally produced material, while the Late Roman material reveals a greater amount of imported pottery from all over the eastern Mediterranean.

Fig. 4: Topographic, ceramic and auger survey plan of Site 44 on Mareotis Island (image Lake Mareotis Research Project).

The sites towards the western extremity of the southern shores formed distinct and in some cases (Site 44 and Site 109) quite substantial units, and are spaced well apart. The archaeology at these sites was largely concentrated at the foot of the limestone ridge (Fig. 4). The sites extended north towards the shore of the lake where many exhibited a series of linear features that extend parallel to the shoreline down to the water’s edge and were open to the E-NE. They appeared to support low-lying areas of marshy ground between the main occupation area and the shoreline ridge, or in many cases a boundary wall to the north. This type of linear feature, known as a ‘lake wall,’ essentially contained a body of water and could have been used for agricultural purposes, perhaps for feeding animals or harvesting reeds, as is still practiced in the region today. This was noticed at a number of sites and also towards the western end of the island (see below) (Fig. 5).

Fig. 5: An example of a ‘lake wall’ feature (Site 44) (photo Lake Mareotis Research Project).

To the east, the sites were located on a limestone ridge that is effectively an extension of the ridge upon which both Mareotis Island and the site of Marea are situated. The nature of these sites appears to reflect similar dates and activities to that of the eastern end of the island. A number of sites have been identified that contained wine production facilities and/or sakkias designed to facilitate the shifting of water (for example Site 100, and Sites 01 and 106). Generally, however, these sites are less substantial than 30

L. BLUE: LAKE MAREOTIS RESEARCH PROJECT those on the island or by comparison with Marea, and they tend to cluster opposite the western end of the island, or along the ridge that extends between Marea and the island. Further investigation is needed to identify the relationship between the eastern end of the island, the south-easterly sites and the western extent of Marea.

of a few ceramics previously identified by El-Fakharani (El-Fakharani 1983: 176-178). Therefore, these buildings might contribute towards resolving the mystery of the so far elusive Pharaonic harbour of Marea (Fraser 1972: 146; El-Fakharani 1991: 28, 1983: 176). Mareotis Island (Fig. 3) Mareotis Island is a limestone ridge joined to the southern shore at its western end. It is possibly an adjunct of the southern shore that has extended out into the lake from the Marea Ridge. It is aligned roughly northeast – southwest and is 3.7 km in length and 680 m at its widest point. Some 40 sites were initially identified primarily concentrated at the eastern and western ends of the island, with a few sites in its central section that were largely industrial-scale Early Roman kiln sites (see Hopkinson this volume).

Two sites stand out as exceptions to the others noted along the southern shores; the Kibotos (the square-shaped harbour) (Site 09), and the ‘complex building’ (Site 13) (Figs. 6 & 7). The Kibotos (Site 09) is essentially a harbour area constructed of large limestone blocks (0.8 x 1 m), defined by a series of moles that enclose an area some 60 m long (N to S) and 36 m (E to W) wide (see Khalil this volume). A second site, known as the ‘complex building’ (Site 13), is located a short distance to the southwest of the Kibotos and is composed of a continuous wall constructed of huge blocks (1.3 x 1.1 m), with a smooth outer surface that slopes up and inwards from the foot of the building. It has been proposed that both these buildings date to the Pharaonic period (Haggag 1984: 277-280). No earlier Pharaonic material has yet been identified at any of the sites recorded in the survey area, with the exception

The majority of the material is Hellenistic in date, particularly towards the eastern end of the island, with a flurry of activity again in the Late Roman period, particularly towards the western end of the island. The ceramic assemblage indicates some imported material especially for the earlier Hellenistic period, but in general it is locally

Fig. 6: The Kibotos site (Site 09) (photo Lake Mareotis Research Project).

Fig. 7: The ‘complex building’ (Site 13) (photo Lake Mareotis Research Project).

31

LAKE MAREOTIS: RECONSTRUCTING THE PAST produced, particularly the production of amphorae. The nature of the structures, with the exception of the centrally located industrial-scale amphorae production kiln sites, is essentially urban in character. The scale of the buildings, particularly those located on the ridge towards the east, is impressive, and many are associated with numerous waterfront structures located along the northern shore on the edge of the coastal plain. At the western end of the island, the settlement extends from the ridge to the north over a flat plain. Whereas the ridge supports further urban structures, the plain to the north appears to be more agriculturally orientated, with features including possible ‘lake walls’ and sakkias similar to those identified along the south-western shores of the lake (see above).

the lake’s Mareotic Arm (see Khalil this volume). The frequency and abundance of these installations along the shores of the Mareotic Arm reflect the extent of maritime activity that was taking place in this region. Besides the jetties and quays which are positioned perpendicular to the shore, another type of maritime installation has also been identified, positioned parallel to the shore, and described as a ‘lake wall.’ The so-called ‘lake walls’ define the shores and protect them from the effects of silting and sedimentation. As many as five lake walls were discovered along the southern shore of the lake and the northern shore of Mareotis Island, areas which are more subject to silting as a result of the prevailing northwest wind. The lake walls could have been utilised as docking facilities for merchant vessels, but were more probably associated with agricultural activities, perhaps the cultivation of reeds as is still witnessed in the region today.

There is also evidence for quarrying on the island, as well as rock-cut tombs. Towards the eastern end of the island, a possible ‘town wall’ has been identified together with a series of jetties and platforms associated with the waterfront, possible warehouses and a small inlet embayment. All these are discussed in more detail in Hopkinson (this volume).

The Lake Mareotis Research Project has also recorded numerous archaeological sites of industrial and agricultural nature such as amphorae production sites, wineries and the remains of several water wheels (sakkia) which were used for lifting water for irrigation and for use in baths and other domestic purposes. With one exception, all the amphorae and wine production sites are located on the southern shore of the lake or on Mareotis Island. In antiquity, the southern shores of the lake were characterised by a fertile plain which supported and enabled agricultural activities to flourish. In addition, the abundance of calcareous clay, particularly suitable for amphorae production, resulted in a thriving large-scale amphorae industry. The study of the ceramic assemblages collected from the surveyed sites indicate that the amphorae production centres were active from the Ptolemaic to the Late Roman period.

What is worthy of note is the fact that the majority of sites on the island are concentrated along the ridge and the northern shores, and that several were clearly built in a very specific relationship with sites on the lake’s north shore (see above). The northern concentration of sites has also raised questions about the island’s status. Does the fact that so few sites have been found along the island’s southern shores suggest that it may have been linked to the southern shore of the lake in antiquity? Recent geomorphological work suggests that in fact the area to the south of the island has always been waterlogged to some degree, but that the waters were shallow and access by boat was problematic, and as such activities in this area were restricted.

Since no lake-side archaeological sites were recorded on the southern shore of the lake opposite Mareotis Island, questions have been raised about the nature of the island and whether or not it was indeed an island in antiquity. However, sedimentological analysis carried out during the final survey season revealed that Mareotis Island was probably an extension of a ridge that extended some 10 km to the west from the site of Marea. In addition, it was ascertained that the area between the island and the southern shore of the lake was either waterlogged or marshland in antiquity (Flaux forthcoming).

Preliminary Conclusions The Lake Mareotis Research Project has identified and systematically surveyed over 70 sites dating from the Hellenistic period through to the 7th century, in what still remains of the western arm of the once much larger Lake Mareotis. The sites vary extensively in nature, size and function, but demonstrate the importance this area had in relation to Alexandria and the mechanisms by which Mareotic products were produced and arrived at the city. Local pottery production is indicated by the numerous kiln sites discovered in the region, particularly on the island. There is also considerable evidence for imported pottery from as early as the late 4th century BC and as late as the 7th century AD. The nature and size of the sites identified varied from rural small holdings to large urban settlements, with associated production sites, water storage facilities and agricultural and industrial complexes.

Previous research conducted in the Mareotic region has been largely limited to specific sites; however, the Lake Mareotis Research Project has systematically investigated all the archaeological remains along the shores of the lake’s western arm. The project also assessed the results of previous archaeological work undertaken in the Western Deltaic region. Based on this, it is reasonable to suggest that the Mareotic Arm was in fact the most active area in Lake Mareotis in antiquity, and that most of the Mareotic products arriving in Alexandria were in fact coming from the western Mareotic Arm, rather than from settlements associated with the lake’s main body (see Khalil this vol-

Different types of waterfront installations were recorded, including more than ten anchorage facilities, which form the majority of maritime installation along the shores of 32

L. BLUE: LAKE MAREOTIS RESEARCH PROJECT que d’Athènes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole française d’Athènes, 11-12 décembre 1988. Athens, Paris. Empereur. J.-Y. (ed.), 2002, Alexandrina 2. Cairo. Empereur, J.-Y., & Picon, M., 1998, Les ateliers d’amphores du Lac Mariout. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athènes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole française d’Athènes, 11-12 décembre 1988: 75-88. Athens, Paris. Flaux, C., forthcoming, Geomorphological survey of the south-west shores of Lake Mareotis. In L. Blue & E. Khalil (eds.), A Multidisciplinary Approach to Alexandria’s Economic Past: the Mareotis Case Study. Oxford. Fraser, P.M., 1972, Ptolemaic Alexandria, Vol. I. Oxford. Frihy, O.E., 1996, Some Proposals for Coastal Management of the Nile Delta Coast. Ocean & Coastal Management 30.1: 43-59. Goddio, F., Bernard, A., Bernard, E., Darwish, I., Kiss, Z., & Yoyotte, J., 1998, Alexandria: The Submerged Royal Quarters. Oxford. Haas, C., 1997, Alexandria in Late Antiquity: Topography and Social Conflict. Baltimore. Haggag, M., 1984, Ports in Ancient Egypt till the Arab Conquest. Unpub. MA diss., Alexandria University. Hassan, F.A., 1997, The Dynamics of a Riverine Civilization: A Geoarchaeological Perspective on the Nile Valley, Egypt. World Archaeology 29.1: 51-74. Rodziewicz, M.D., 1998, Fom Alexandria to the West by Land and by Waterways. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athènes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole française d’Athènes, 11-12 décembre 1988: 93-103. Athens, Paris. Said, R., 2002, Did Nile Flooding Sink Two Ancient Cities? Nature 415: 37-38 Strabo, Geography, H.L. Jones (transl.), 1949. London. Warne, A.G., & Stanley, D.J., 1993, Late Quaternary Evolution of the Northwest Nile Delta and Adjacent Coast in the Alexandria Region, Egypt. Journal of Coastal Research 9.1: 26-64.

ume). Finally, the Lake Mareotis Research Project has outlined the true scale of activities in the Mareotic Arm with respect to industrial, agricultural and urban sites, being much more extensive than previously suggested, thus supporting the significant importance of the area and highlighting the role it played in the economy of Hellenistic and Roman Alexandria. Acknowledgments The author would like to thank the Supreme Council of Antiquities for permission to work in the Mareotis region, particularly colleagues from the Underwater Department who were instrumental to the success of this project, in particular the support and help of the project co-director Dr Sameh Ramses and project manager Dr Emad Khalil. The project would not have been possible without the financial backing of both the British Academy and the Leverhulme Trust particularly for funding Dr Khalil’s two-year post doctoral fellowship. Finally, thanks is due to all students of Southampton and Alexandria Universities and Egyptian and UK colleagues who worked on the project without whom the survey would not have happened. Bibliography El-Fakharani, F.A., 1983, Recent Excavation at Marea in Egypt. In G. Grimm, H. Heinen & E. Winter (eds.), Aegyptiaca Trevernsia II, Das Römisch-Byzantinische Ägypten. Akten des internationalen Symposions 26.-30. September 1978 in Trier: 175-186. Mainz. El-Fakharani, F., 1991, The Kibotos of Alexandria. Studi Miscellanei 28: 21-28. Blue, L., & Ramses, S., 2005, Lake Mareotis Research Project – Preliminary Report from Pilot Survey, August 2004. Newsletter of the Archaeological Society of Alexandria 15 (February): 5-16. Blue, L., & Khalil, E. (eds.), forthcoming, A multidisciplinary approach to Alexandria’s economic past: The Mareotis case study. Oxford. De Cosson, A., 1935, Mareotis: Being a Short Account of the History and Ancient Monuments of the NorthWestern Desert of Egypt and Lake Mareotis. London. Empereur, J.-Y. (ed.), 1998, Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du collo-

33

LAKE MAREOTIS: RECONSTRUCTING THE PAST

34

D. HOPKINSON: MAREOTIS ISLAND SURVEY

The Results of a Preliminary Survey at Mareotis Island Dylan Hopkinson

Introduction

northern and southern limits of the western arm of Lake Mareotis and form a natural flooded depression known as the Mallahet Maryut Depression (Fig. 1). The island measures 3.7 km long by an average of 400 m (maximum 680 m) wide and is connected to the mainland by an area of low-lying marshy ground to the southwest at the foot of the Gebel Mariout Ridge (Fig. 2). The lands around Lake Mareotis formed an important agricultural production centre in antiquity known to have been richly populated with farming villages, market-towns and villa estates (Strabo Geography 17.1), and would have been dependent on seasonal rainfall and Nile floodwaters for their irrigation and fresh drinking water.

Location The site known as Mareotis Island lies 65 km southwest of Alexandria within the western arm of Lake Mareotis, closely associated with its southern shore. It lies 4 km to the west of the site of Marea/Philoxenité, and directly north of the modern town of Bahig (Fig. 1). The island was subject to systematic survey in 2007 and 2008 as part of the broader Lake Mareotis Research Project (see Blue this volume). A large number of archaeological features are clearly visible on the ground which date to the Greco-Roman periods largely indicated by the substantial quantities of ceramics present in surface scatters. This paper presents the overall character of the archaeological features observed, and some preliminary interpretation with specific reference to the littoral zone on the northern coastal plains where ‘maritime’ and lakeside activities were largely focused.

Environmental Background In antiquity Lake Mareotis provided a transport connection between Alexandria and its agricultural and industrial hinterland in the Mareotic region; and via the Nile waterway to the Egyptian interior along the Nile Valley. Strabo states that the lake “is filled by many canals from the Nile, both from above and on the sides” (Geography 17.1.7). It was these canals that provided the fresh water which came

The island is formed from the remnants of a small limestone ridge which lies between the two much larger Abusir and Gebel Mariout Ridges. These two ridges define the

Fig. 1: The location of Mareotis Island and the sites mentioned in text, modified from De Cosson 1935.

35

LAKE MAREOTIS: RECONSTRUCTING THE PAST Fig. 2: Schematic of the main features of Mareotis Island (Hopkinson).

from the Canopic Branch of the Nile apparently keeping the lake water sweet, and filling the cisterns of Alexandria. The direct link to the Nile during the seasonal floods is also likely to have had a significant impact on lake levels. The exact location and number of these connections is not firmly understood, but it is likely that the ancient courses would have mirrored those of the more recent canals (Warne & Stanley 1993: 15-16).

of sites located on the southern shore of the lake, where Roman foundations are observed to have been buried under 3 m of deposits (Khalil 2005: 39). Holocene sedimentation rates between 0.02 and 0.07 cm per year are indicated by Smithsonian boreholes either side of the island (Warne & Stanley 1993: 50). This sedimentation is likely to have obscured archaeological features over time, and to have raised the level of the lake bed while consequently reducing the depth of the lake waters (Flaux 2008: 9-10, III.2).

There is sufficient environmental evidence to show that the climate of the region has not changed significantly in the past two thousand years (Ball 1942: 20; Brooks 1949: 333). Evidence for this includes an analysis of a weather diary kept in Alexandria by Ptolemy in the 2nd century, which concludes that “the number of rainy days per annum was much the same as it is nowadays, though the rainfall was more evenly spread out over the year” (Murray 1951). So we can conclude that the climate was semi arid as it is today. Today rainfall in the Western Coastal Desert region is greatest along the margin by the sea and is typified by its great variability from location to location and from year to year, which suggests that in antiquity even this fertile and productive region would have had a precarious existence were it not for the fresh waters of the lake and the reliable annual Nile floodwaters.

Modifications to the Lake since the Greco-Roman Period The water regime in the area has been greatly modified since antiquity, most notably by the silting of the connections between the lake and the Nile; when the Canopic Branch of the Nile finally silted up around the 12th century the lake also began to dry up and shrink in size. Strabo also tells us that the extent of the lake must have been much greater in antiquity than it is today, and extended around 40 km to El Bordan to the southeast and to Kom Al Trouga 70 km to the southwest of Alexandria (De Cosson 1935: 26) (see Fig. 1). Warne & Stanley also support these dimensions which amount to around 700 km2 of land lying below the 0 m sea-level datum (Warne & Stanley 1993: 53).

Light north-westerly winds are generally observed in the region and these transport large quantities of clayey loess sediments from the degrading limestone ridges and silts and sand from the coastal plain. These silts are deposited in colluvial sheets at the base of the ridges and on north facing shores. This phenomenon has been noted previously (Warne & Stanley 1993: 53), and at a number

It is very likely that even without a direct connection to the lake the effects of the Nile flood would still have been noticeable in the water levels due to groundwater seepage (A.G.,Brown, Univ. of Southampton, pers. com.). When the Aswan Dam was constructed in 1964 there was therefore a significant change to the water level in the Nile 36

D. HOPKINSON: MAREOTIS ISLAND SURVEY Delta aquifers and the lake. As a result the hydrodynamic environment today is very different from antiquity and we have no direct way of understanding the range of water levels that would have been experienced during the floods. As Ramses & Omar (in this volume) indicate, modern infrastructure, fisheries and agricultural drainage ditches have further modified the lake, splitting it into several artificial basins and remodelling large stretches of coastline. These basins have water regimes that are artificially managed for modern needs, and large amounts of water are pumped into the sea to keep the waters artificially low for land reclamation projects.

this sakia to the cisterns of the town. To the south of this are the remains of two very interesting pottery kilns with a large heap of broken pottery thrown there as it was ‘scrapped’ from them. North of the sakia there is a long jetty running into the Lake from which ferry-boats once plied.” (De Cosson 1935: 130)

The reference here to waterwheels (‘sakia’ or ‘saqiya’), cisterns, ceramic mounds from pottery production and lakeside features characterise the island in general terms, but the description bears little direct relation to the position of the remains observed today. Consequently the archaeology of Mareotis Island represents a significant untapped resource that can enhance our understanding of the region.

Today the lake is fed only by water seepage from the sea, precipitation, and by irrigation ditches that channel agricultural waste-water from the Beheirah Governorate. As a result of all of these changes, it has been calculated that the current lake level is only 13% of the ancient lake surface area (Warne & Stanley 1993). Observations made during the 1970s show that the western arm of the lake was totally dry at least from Marea/Philoxinite and to the west. However, a recent geoarchaeological survey as a component of the Lake Mareotis Research Project indicates that the depositional environment close to the modern shoreline had always been lacustrine except during the recent centuries (Flaux forthcoming).

Fieldwork In August 2007 under the auspices of the Lake Mareotis Research Project, a six-week survey was undertaken to plot the visible archaeological remains on the island using RTK GPS and Total Station surveys. In order to provide an understanding of the dating of features a controlled surface ceramic collection strategy was adopted. An RTK GPS topographic model was also completed with 5 m transects across the island. In addition to this, a geoarchaeological auger survey was undertaken in 2008 to help contextualise the results and to begin to understand how site formation processes can inform our understanding of the site. Of particular interest here was the relationship between the ancient water levels and the archaeological features, and the impact of the changing lake levels on the development of the sites. A more detailed account of the Lake Mareotis Research Project and its methodologies is presented in this volume (see Blue this volume, see also Khalil and Ramses & Omar this volume)

Deltaic subsidence and tectonic shifts may also have changed the topography of the region as it is known to have done within the Nile Delta region further to the east; however it has not proved possible to use existing geoarchaeological datasets to understand the significance of these events. Consequently as a result of the various geographical factors described, we must accept that the current lake level cannot be used as a reference datum from which to reconstruct ancient lake levels and to assess the impact of the annual floods and the summer evaporation on the shoreline. However, understanding these relationships is a fundamental factor to our interpretation of the archaeological structures lying near the shore.

The majority of archaeological structures are located on high ground, occupying the central limestone ridge that forms a series of hills that extends as a discontinuous spine along the length of the island. Two main settlements were identified concentrated at the eastern and western extremes of the island, with further activities evident on the north facing coastal plains below the central ridge. The central section of the island is some 2 km long and is much less densely occupied than the western and eastern extremities. The activities here represent industrial amphorae production and possible agricultural activities.

Previous Research There has been a significant increase in our understanding of the archaeology of the Mareotic region in recent times brought about by the work of El-Fakharani (1974, 1983, 1984), Rodziewicz (1983, 1990, 1998a, 1998b, 2002), and Empereur (1986, 1998; Empereur & Picon 1986, 1998, 1992), however previous research on the island itself appears to have been negligible. There are numerous cartographic representations of the region, two of the most informative being the one on Sheet 37 of the Atlas Géographique in the Description de l’Égypte from 1809, and on Carte des environs d’Alexandrie produced in 1866 by Mahmoud Bey El-Falaki; however, the detail pertaining to the island in both cases is limited and difficult to reliably geo-reference. A further cryptic description of the archaeology on the island is given by De Cosson (1935):

Chronology The ceramics recovered from the island indicate that the settlements was well established some time during the Hellenistic period and that occupation potentially continued through to the 7th century AD. The western settlement appears to have been continually occupied as indicated by ceramic forms recovered, whereas the eastern settlement seems to have suffered a decline in activity in the Early Roman period which reached its nadir some time in the 4th to 5th centuries with a late recovery in the 6th century. The results of the ceramic survey are not commented on in detail here, but will be published shortly (Tomber & Thomas forthcoming).

“Many buildings can be traced, and at the eastern end the circular stone platform and oblong well of an ancient sakia will be found. Stone channels are traceable leading from

37

LAKE MAREOTIS: RECONSTRUCTING THE PAST Archaeology of the Eastern Settlement The eastern settlement (Fig. 3) is concentrated on the high ground of the limestone ridge which measures approximately 650 m by 350 m; the majority of this area is covered with building remains of various construction styles and materials. The layout and construction styles of the visible building remains indicate some degree of social organisation; such as possible administrative and commercial areas, some are grouped around what appear to be large, open areas at a number of locations on top of the ridge. A long boundary wall is located winding along the top of the ridge in the north-western limit of the settlement; this wall measures 240 x 1.20 m wide, and may have served a defensive function protecting the settlement from the west. To the west of the settlement there is a small low-lying depression that could well have formed a natural inlet prior to silting; around its northern and southern margins the remains of buildings that may have been possible wharf structures have been identified. This inlet appears to have served as an inland harbour which would have provided sheltered mooring. Auger cores taken during the 2008 season confirm that the sediments in this depression were indeed deposited in lacustrine conditions in antiquity (Flaux forthcoming).

are difficult to interpret in detail because there are several phases of activity. The relative dating of these features has not yet been possible. These structures appear to fall into several categories of linear features, walls and lakeside buildings. Linear Features (Group A) The first category of structures is a series of five linear features located on the low-lying northern coastal plain that extends between the present northern shoreline of the island and the foot of the limestone ridge (Fig. 3, Group A). These features range between 65 to 120 m in length, and appear as low, roughly built earthen ridges up to 4 m wide which are spaced roughly between 30 to 100 m apart. At the waters edge, additional features can be seen that extend along the current shoreline. As yet, these water front features have an undefined chronological relationship to the linear features (Group A) and represent primary and secondary activities over a period of time. The landscape between the base of the limestone ridge and the lakeside features appears to be largely devoid of additional structures and the linear features are relatively isolated with only a few abutting or adjacent features. It has already been established that little is known about the absolute water-levels of the lake in antiquity, and unfortunately the 2008 geoarchaeological survey did not target the area around the linear features therefore our interpretations

To the north of the eastern settlement, along the northern coastal plain of the island, between the lake and the foot of the ridge, are a series of structures and buildings which

Fig. 3: The littoral features mentioned in this paper and topographic survey of the eastern settlement of the island (Hopkinson).

38

D. HOPKINSON: MAREOTIS ISLAND SURVEY of the exact nature and function of these features remains speculative, however two possibilities present themselves. Firstly, they may represent crude earthen jetties projecting from the ridge into the lake during periods of higher water-levels. The other possibility is that the coastal plain was dry in antiquity and that the earthen ridges may have divided the coastal land into units, potentially for cultivation as is suggested by the interpretation of similar features identified in the western settlement. We know from ancient sources that the Nile had a dramatic effect on the Delta region; Strabo states “at the rising of the Nile the whole country is under water and becomes a lake, except the settlements; and these are situated on natural hills or on artificial mounds, and contain cities of considerable size and villages, which, when viewed from afar, resemble islands” (Geography 17.1.4). It is therefore possible that the annual variation in the level of the lake level was large enough for both scenarios to be valid.

‘quay’ wall to demarcate the edge of the lake either as protection from inundation or to form a working platform, perhaps for the mooring and loading of boats. In some instances the construction of these sections of wall seem to be related to the linear earthen ridges of Group A, and it would appear that in some cases they were built to connect or at least respect, the linear features lakeside termini. The elevations recorded from various sections of these walls indicate that the surviving top course of the walls are not all at the same level but in fact exhibit a difference in height of over 1.83 m. This may indicate that either the top course of some of the sections is missing, or that these walls were built at different times and at different mean lake-levels rather than as a single planned ‘quay’ building effort. Coastal Buildings Associated with the first group of linear features (Group A) are a large number of walls indicating the presence of a number of buildings that had been constructed along the shoreline in this area (see Fig. 3). Since the relative dating of structures has yet to be determined it is not possible to discuss the phasing of the structures. However, what is apparent is that if the linear features were jetties and if they were built at the same period as the coastal buildings, then these buildings would have been susceptible to serious flooding. It is vital that the relationships and phasing of these features is investigated as it will have important implications with respect to interpretation of the landscape and the features located within it.

Linear Walls (Group B) A second group of linear walls were observed to the west of Group A along the north-western shores of the lake in the eastern settlement area. This group is located to the west of the wall that winds along the top of the western side of the limestone ridge, and down onto the coastline (see Fig. 3). These linear features were observed as low rubble walls 0.50 m thick, their northern lakeside ends are close to the current waterline and appeared to extend inland for distances of between 25 to 60 m. None of these walls reached the foot of the limestone ridge and in most cases extended less than half way across the plain. Three of these walls were spaced approximately 35 m apart with a further group 30 m to the east that was composed of three roughly parallel walls in very close proximity. The coastal ends of the first three walls had associated low rubble linear structures or platforms which extended for a short distance along the coastline.

Amongst these coastal buildings are a group of structures that deserve particular attention. Three isolated groups of buildings of similar construction were identified at three different areas of the plain (see Fig. 3; Group D). Two of these are located on the north coast and a third on the eastern extreme of the island with a south-easterly coastal aspect. The buildings appear to be arranged in symmetrical blocks facing each other and around 13 m apart, each measuring approximately 10 m wide and between 30 and 55 m in length, and were aligned perpendicular to the coast. The building pairs consist of single rows of cellular rooms measuring roughly 10 m by 5 m. Their location and alignment relative to the coast suggests that their function was somehow related to the lake and that they could have been warehouses, shops or boathouses.

The alignment of the main walls of Group B is interesting since they were constructed on the same orientation as the linear features discussed above (Group A). However, the Group B linear features are not perpendicular to the shoreline since this section of the coast turns to the south and is west facing. As a result these features are on an oblique alignment relative to the current shoreline. The stone construction, relative close spacing, alignment to the coast, and their location to the west of the long winding ‘boundary’ wall, suggest a distinct function as compared to the Group A linear features. One possibility is that these features may have been concerned with land reclamation or stabilisation, serving perhaps as a series of ‘groins’.

Summary Overall there appears to be a great deal of established activity in the eastern settlement. Buildings on the ridge settlement hint at developed social organisation and those on the northern coastline show a strong association with the lake. Features interpreted as an inland harbour, lakeside walls, and possible jetties, all suggest an established interaction with the lake. There are a far greater number of buildings on the northern coastline as compared to the south coast in this part of the island, which may suggest that the southern coast was possibly dry or marshy and therefore inaccessible to boats in antiquity.

Ashlar Lake Walls (Group C) Large sections of dressed ashlar blocks are observed forming stretches of wall along the northern edge of the plain (see Fig 3). These walls closely respect the natural topography of the current shoreline and appear to have originally been laid as stretchers extending parallel to the shore. This might suggest that the walls were built as a sort of 39

LAKE MAREOTIS: RECONSTRUCTING THE PAST Archaeology of the Western Settlement The western settlement occupies a similar location to the eastern settlement, situated on the high ground of the limestone ridge which measures 820 by 200 m (Fig. 4). The area exhibits the remains of various buildings, with a mixed function suggested by the construction materials and surface finds. Overall the preservation of these structures is not good; a large area in the extreme west has been heavily damaged by earthmoving machines, and in other instances only patchy areas of construction and dislocated wall lines survive. As a result, it is not possible to identify any zoning of functionality or social organisation, although a few significant groups of structures were observed. As in the eastern settlement the northern coastal plain appears to be the focus of activities. Once again it is the lakeside features that are the focus of interest in this paper although some of the structures on the ridge will be discussed.

Ashlar Shoreline Wall Alignment (Feature E) The area along the waterline is densely overgrown with water reeds and has silted up making the water level very shallow and the ground marshy. Amongst the reeds can be seen the upper surface of sections of a long straight wall (see Fig 4). All the shoreline features in the western settlement appear to respect this northern boundary. The wall is built from limestone ashlars laid as stretchers with a flush north-facing aspect, suggesting that it was laid in a specific relation to the lake. The various sections of this feature appear to suggest a continuous wall with overall dimensions 0.40 m by as much as 245 m in length. No excavations of this feature were undertaken so it is not possible to indicate the number of courses or the height of the wall. The location and construction of this wall suggest that it has a significant relationship to the shoreline if water levels similar to those seen today are assumed. It appears to be a lakeside wall protecting the land behind or creating a working platform from which to conduct activities related to the lake. Some of the other features close to this wall appear to butt against this wall suggesting that it may also form part of a larger pattern of structures.

The features of the coastal plain are sparsely distributed and appear to fall into three clear categories: a section of ashlar wall aligned along the shoreline; long, thin linear walls forming low ridges perpendicular to the coastline; and localised masonry features typified by curvelinear walls and associated ashlar lined pits, there is an overall lack of secondary activities such as the coastal buildings on the eastern coastal plain, which makes interpretation somewhat easier.

Linear Walls (Group F) The second group of features located on the coastal plain of the western settlement, is a series of roughly paral-

Fig. 4: The littoral features mentioned in this paper and topographic survey of the western settlement of the island (Hopkinson).

40

D. HOPKINSON: MAREOTIS ISLAND SURVEY Schiøler identifies 14 different types of saqiya from Roman, Islamic, and modern contexts (Schiøler 1973: 13), including wheels with qudus pots attached directly to the rim, and wheels which powered a ‘pot-garland’ or bucket chain. In the case of the features associated with the western settlement it seems that the curved walls represent part of the animals’ circular track, while the ashlar lined pit is likely to be associated with either the wheel or the gear mechanism.

lel linear features aligned perpendicularly to the current coastline; they are constructed from faced ashlar blocks that in some cases have degraded to stone rubble and in some areas only survive as low earthen ridges. Seven potential features were observed in this group (see Fig. 4), all of which have some connection with the current coastline and appear to respect the alignment of the possible lakeside wall. These linear features extend back from the coast, south towards the main ridge; their lengths vary between 25 and 100 m but on average only reach halfway across the plain and appear to show no direct relationship with the ridge itself. Two of these walls are connected by a linking wall at their southern ends.

Many sherds of qudus pots were observed along the northern flank of the limestone ridge adjacent to the plain but were not recorded anywhere else on the island; the location directly adjacent to the water is also appropriate for this interpretation of water management, and places the northern coastal plain in an agricultural setting.

These walls have much in common with similar features at the eastern end of the island, they have the same orientation relative to the coastline and occupy the ground between the coast and the ridge, with their coastal terminus closely associated with sections of ashlar wall aligned along the coast. This suggests that the walls could have potentially functioned as jetties when lake waters were high or as boundary markers potentially prescribing plots of land along the coast that butt against the ashlar wall.

In the light of this interpretation it is possible to reinterpret the long linear walls between the ridge and the quay wall; these may potentially be seen as boundaries demarking areas of farmed land irrigated by saqiyas. Oleson concludes that rudimentary saqiyas from modern contexts could usefully irrigate similar portions of land to those possibly represented by the area of land contained within the linear walls (Oleson 1984: 369).

Curvelinear Walls and Ashlar Lined Pits (Features G) The land between the linear walls had a sparse distribution of ashlar constructed features that form the third group of structures. They are represented by sections of curved wall in two locations where additional associated wall lines were observed. These curved walls are incomplete but appear to represent rounded features with projected diameters of about 10 m (see Fig. 4).

Linear Mole Extending into the Lake towards a Small Island At the north-eastern extent of the western settlement in the marshy ground of the shoreline plain is a site that is unique on the island (see Fig. 2). This is significant because it is the only ancient feature on Mareotis Island that currently extends into the lake, and as such it represents a useful indicator of lake levels in antiquity. The feature is a 250 m long linear mole construction that extends out into the lake to join with a small island that measures roughly 30 by 15 m. This small island supports a rectangular building measuring 20 x 10 m that was built from substantial masonry blocks. Three jetties extended from the north of this building into the lake measuring between 7 and 15 m in length. The feature extends towards a promontory settlement on the opposite side of the lake where there is a similarly built reciprocal linear construction that extends for a short distance into the lake on the same alignment towards the small island. These features imply a functional relationship with the lake perhaps for controlling the movement of transport along the lake as well as between the two settlements.

The interior structure of one of these rounded features is preserved and takes the form of ashlar built components lining a rectangular slot which measured 4.80 by 0.90 m. This feature, although unexcavated and partially observed, bears a significant similarity to ancient water-lifting devices known as saqiya which have been observed archaeologically from broadly contemporary Greco-Roman sites in the region such as Abu Mina located some 15 km south of the island (Schiøler 1973: 129-136; Oleson 2000: 263; see Fig. 1), and Tuna Al-Gabal near Hermopolis in the Nile Valley (Schiøler 1973: 141-148). There are further saqiya features in the Mareotic region which have not been published in detail; examples in the immediate vicinity of the island are described as having a “circular stone platform and oblong well” (De Cosson 1935: 130).

Tower-like Feature (Feature H) A final feature in the western settlement is worthy of mention, this is a small artificial mound of collapsed building material some 4 m in height which lies just northeast of the main ridge on which the settlement is located. After cleaning the flat top of the mound, masonry elements were observed that formed a structure similar to the ashlar lined pits mentioned above. Rather than having a single pit slot this feature had a pit that was divided into two separate apertures with overall dimensions very close to the saqiya discussed above.

Waterwheels were in common use in Egypt from Hellenistic through to medieval times, and are simple devices using a gear mechanism to convert the horizontal movement of an animal walking round a circular track into vertical movement for lifting water. The water was held in pots called ‘qudus’ which are connected to the wheel. Saqiya are used predominantly in agricultural contexts (Oleson 1984: 126) but have also been recorded in the bath-house buildings, as at Abu Mina (see Fig. 1). Waterwheels are also known to occur in many configurations; 41

LAKE MAREOTIS: RECONSTRUCTING THE PAST Fig. 5: Longitudinal and transverse schematic sections through the ‘water tower’ on Mareotis Island (Hopkinson, after McCann, et al. 1987).

top of the tower and a further 1.5 m to the vat on top of ridge, an overall height in the order of some 5.5 m, the additional height of about 1.5 m being accounted for by the assumed superstructure which held the qudus pot-garland above the surviving tower.

It is possible that this feature is the top of a tower used to lift water from a well known to exist close to the bottom of the mound using a pot-garland passing through each aperture. An elevated ‘bridge’ constructed from earth and masonry blocks connects the tower to the main ridge at a point where a plaster lined vat is located as part of a larger building complex that could potentially represent a public bathhouse or cistern and water management feature (Fig. 5). An example of such a tower from the late 1st century AD is known from the port of Cosa in Italy (Oleson 2000: 258). The so called ‘spring house’ at Cosa initially lifted water to a height of 3 m but it burnt down and was replaced with a structure that lifted water over a height of 10 m to an elevated conduit. If this hypothesis is correct the structure on the mound would have lifted water to about 4 m to the

Summary The western settlement is generally poorly preserved, however there is an apparent lack of secondary activities on the northern plain which makes it easier to interpret the features that survive. The presence of water lifting devices and possible plots of agricultural land suggest that the majority of the coastal plain was above the mean waterlevel in antiquity. As such an alternative interpretation of the linear walls as possible jetties is not so well supported, 42

D. HOPKINSON: MAREOTIS ISLAND SURVEY although they could have served this function during the flood season. The absence of further buildings on the plain may suggest that this community did not have a strong association with the lake for their subsistence; however the presence of the promontory towards the small island may suggest that at some point the settlement had a significant role in administering activities on the lake.

plications for the interpretation of the other ‘coastal’ walls on the island. To the south of the central ridge of the island, there are two adjacent ‘plots’ of land which are enclosed by low walls of faced stone and rubble (see Fig. 2). These features resemble ‘karum’ or vineyard plots; features that are usually found inland in a 15 km wide band south of the Gebel Mariout Ridge, as indicated on Survey of Egypt maps from 1927. Karum are features that are poorly researched but the author believes they were used to raise the depth of topsoil on agricultural plots with the advantage of increasing the space for healthy root growth and for storing rainwater in the soil reservoir (Hopkinson 2007). There seems then to have been some small scale industry and subsistence agriculture in the middle of the island that is not directly related to either of the main settlements.

Archaeology of the Central Island The land between the eastern and western settlements on the island is roughly 2 km long and appears to have been sparsely utilised in antiquity for industrial and agricultural activities. There are four sites of large scale pottery manufacture (see Fig. 2; Kiln “A” & Kiln “B”), each with extensive associated storage rooms, and large mounds of ceramic sherds wasted during the firing process. It seems that these potteries, like many identified near the coast on the southern mainland, were exclusively producing amphorae forms known to have been used for making and packaging the large quantities of wine for which the region is renowned (Empereur & Piccon 1986). Two of the kilns originate in the Hellenistic period (see Fig. 2; Kiln “A”), one close to both the eastern and western settlements. These sites continued in use through to the mid-1st century AD, when amphorae production switched to two adjacent sites (see Fig. 2; Kiln “B”). Production of amphorae at the new sites continued from the late 1st to the 5th centuries AD.

Water Levels and Connection to the Mainland One important consideration for the interpretation of the features on the island is the relationship between the island and the mainland to the south, and the local hydrological environment throughout its occupation. The island is currently largely surrounded by water but it should not be assumed that this was the situation in antiquity because modern irrigation and dyke formation have modified the movement of water. A number of the features that have been identified on the island may suggest a greater connectivity to the mainland to the south. All of the potential maritime focused features such as the possible jetties, quays and lake walls are located on the north facing shores of the island, and the absence of these features on the southern shores suggests the absence of water in this area that could be usefully exploited. Moreover the features that are observed along the current southern shore appear to be associated with dry land activities. The ‘karum’ plots are agricultural features thought to be concerned with collecting and maximizing the use of precipitation and as such suggest that there is no alternative readily available water resource in close proximity. The second feature that appears to preclude the existence of a significant southerly body of water in antiquity is the large kiln structure observed in the west of the island (see Fig. 4; ‘kiln site’). These kilns are known to have significant subterranean fire chambers up to 4 m deep where the fire was lit, with a perforated firing plate above to allow the hot air to circulate (El-Fakharani 1983). However, the firing plate of this kiln is currently only a short height above the lake waters to the south allowing little room for the firing chamber and indicating that it could only have been used if the land to the south between the island and the mainland had been dry. The inference that there was connectivity between the island and the southern shores of the lake, is supported by what we know about the organisation of wineries of the region, consisting of three essential elements: the kiln, the winery, and the estate villa (Empereur & Picon 1992). At least three winery structures are known in the immediate vicinity of the island on the southern mainland, and no structures suitable for the production of wine have been identified on the island itself. It is possible to suggest therefore that

Further evidence of light industry on the island is indicated by two small quarrying sites on the limestone ridge (see Fig. 2), although their capacity is very small compared to the quantity of stone used in the settlements and it is more likely that the majority of the building stone was transported across the lake from the large quarry sites known to exist in antiquity on the Abusir Ridge to the north (Oliver & De Cosson 1938: 169-170). In the middle of the island, associated with the kiln sites, there are a group of structures that occupy the top of the ridge that could have been workshops or accommodation for the workmen involved in these industries. A number of features from these structures extend down towards the northern coastline, although on the plain below there is only one visible structure. This is a 75 m long section of ashlars laid as stretchers aligned with the current coastline (see Fig. 2). The wall lies 70 m inland from the current coastline, which is surprising since the flush face of the wall at first suggests a coastal function similar to the sections of lake wall or quay found associated with the eastern and western settlements. However, if this was its function then the land directly north of the wall equates to a large area of siltation and aeolian soil deposition. Small scale excavations either side of the wall in 2008, however, indicated that there was only one course of stones surviving and that the soils surrounding the feature appeared to display similar sediment regimes and were not deposited underwater. It therefore seems that this section of wall had no direct association with the lake, and its purpose is unclear. This finding is also significant since the construction of a flush face apparently respecting the coastline has im43

LAKE MAREOTIS: RECONSTRUCTING THE PAST the kilns on the island were producing amphorae for the wineries on the southern mainland. However, the recent geomorphological enquiry undertaken as part of the Lake Mareotis Research Project has confirmed that the body of water that currently extends from the southern shores of the island was in fact waterlogged in antiquity (Flaux forthcoming). The island ridge was an extension of the Marea Ridge and thus enclosed a body of shallow water to the south. How continually waterlogged this area was in antiquity has yet to be determined but the geomorphological survey has indicated that there was continuous lake sedimentation between the island and the southern shores since antiquity. Thus, it is assumed that access from this area into the main body of the lake was restricted by the ridge, thus deterring settlement on the southern shores of the island, as witnessed by the lack of archaeological sites in this region. It is clear however that the situation is very complicated and that our understanding of the archaeology of the island is incomplete.

El-Fakharani, F.A., 1983, Recent Excavation at Marea in Egypt. In G. Grimm, H. Heinen & E. Winter (eds.), Aegyptiaca Trevernsia II, Das Römisch-Byzantinische Ägypten. Akten des internationalen Symposions 26.-30. September 1978 in Trier, 175-186. Mainz. El-Fakharani, F., 1984, The Kibotos of Alexandria. Studi Miscellanei 28: 23-28. El-Falaki, Mahmoud Bey, 1872, Memoire sur l’antique Alexandrie, ses faubourgs, ses environs. Copenhagen. Empereur, J.-Y., 1986, Un atelier de dressel 2-4 en Ẻgypt au IIIe siécle’ de notre ére’. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athènes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole française d’Athènes, 11-12 décembre 1988: 599-608. Athens, Paris. Empereur, J.-Y., 1998, Alexandria Rediscovered. New York. Empereur, J.-Y., & Picon, M., 1986, A La Recherche des Fours d’Amphores. In J-Y. Empereur & Y. Garlan (eds.), Recherches Sur Les Amphores Grecques. Actes du colloque international organisé par le Centre national de la recherche scientifique, l’Université de Rennes II et l’École française d’Athènes (Athènes, 10-12 Septembre 1984): 103-126. Athens, Paris. Empereur, J.-Y., & Picon, M., 1998, Les atelier d’amphores du Lac Mariout. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athènes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole française d’Athènes, 11-12 décembre 1988: 75-88. Athens, Paris. Empereur, J.-Y., & Picon, M., 1992, La Reconnaissance des Productions des Ateliers Cérmiques: L’ Exemple de La Maréotide. Extrait des Cahiers de la Céramique Égyptienn 3: 145-152. Flaux, C., forthcoming, Geomorphological survey of the south-west shores of Lake Mareotis. In L. Blue & E. Khalil (eds.), A Multidisciplinary Approach to Alexandria’s Economic Past: the Mareotis Case Study. Oxford. Hopkinson, D.J., 2007, An Assessment of Greco-Roman Lakeside Agriculture and Trade in the Mareotic Region of Egypt. Unpub. MA diss., University of Southampton. Khalil, E.K.H., 2005, Egypt and the Roman maritime trade: a focus on Alexandria. Unpub. PhD thesis, University of Southampton. McCann, A.M., Bourgeois, J., Gazda, E.K., Oleson, J.P., & Will, E.L., 1987, The Roman Port and Fishery of Cosa: a Center of Ancient Trade. Princeton. Murray, G.W., 1951, The Egyptian Climate: An Historical Outline. The Geographical Journal 117.4: 422-434. Oleson, J.P., 1984, Greek and Roman Mechanical WaterLifting Devices: The History of a Technology. London. Oleson, J.P., 2000, Water-Lifting. In Ö. Wikander (ed.), Handbook of ancient water technology, 217-302. Leiden

Conclusion The archaeology of Mareotis Island consists of two settlements which appear to be concerned with agricultural, industrial and water management activities as well as lacustrine focused features which may indicate that fishing activities or lake transportation services were conducted from the island. The island represents a largely unexplored archaeological asset that has the potential to give insights into daily life in a non elite community; when looked at in conjunction with other known sites in the area it may provide an opportunity to begin to address settlement distributions and social organisation and make comparison between the lake region and those communities on the Delta, and perhaps archaeological findings from the Fayum or Nile Valley, where the most studied examples of agricultural settlements are known. Such comparisons with sites from the Egyptian interior may prove to be less than representative of ordinary Egyptian practices where the majority of the population was located on the Nile Delta (Bagnall 2007: 227). There is a great potential for future research on Mareotis Island and the broader Mareotis region and these provisional findings only begin to scratch the surface of what the island and Mareotic region in general has to offer. Bibliography Bagnall, R.S., 2001, Archaeological Work on Hellenistic and Roman Egypt, 1995-2000. American Journal of Archaeology 105. 2: 227-243. Ball, J., 1942, Egypt in the Classical Geographers. Cairo. Brooks, C.E.P., 1949, Climate through the ages. New York. De Cosson, A., 1935, Mareotis: Being a Short Account of the History and Ancient Monuments of the NorthWestern Desert of Egypt and of Lake Mareotis. London. El-Fakharani, F., 1974, The “Lighthouse” of Abusir in Egypt. Harvard Studies in Classical Philology 78: 257272. 44

D. HOPKINSON: MAREOTIS ISLAND SURVEY le Laboratoire de céramologie de Lyon et l’Ecole française d’Athènes, 11-12 décembre 1988: 27-36. Athens, Paris. Rodziewicz, M.D., 2002, Mareotic Harbours. In C. Déco bert (ed.), Alexandrie Médiévale 2. Etudes Alexandrines 8: 1-22. Schiøler, T., 1973, Roman and Islamic Water-lifting Wheels. Denmark. Strabo, Geography, H.L. Jones (transl.), 1949. London. Survey of Egypt (27/135) 1:100,000 Map; Sheet 88/48 “El Ghayata”. Survey of Egypt/Ministry of Finances, Cairo. Tomber, R., & Thomas, R.I., forthcoming, Pottery from the Lake Mareotis Research Project. In L. Blue & E. Khalil (eds.), A Multidisciplinary Approach to Alexandria’s Economic Past: the Mareotis Case Study. Oxford. Warne, A.G., & Stanley, D.J., 1993, Late Quaternary Evolution of the Northwest Nile Delta and Adjacent Coast in the Alexandria Region, Egypt. Journal of Coastal Research 9.1: 26-64.

Oliver, F.W., & De Cosson, A., 1938, Note on the Taenia Ridge – with especial reference to quarries, sites and an ancient road between Alexandria and Abu Sir. Bulletin de la Société Royale d’Archéologie d’Alexandrie 32: 163-175. Rodziewicz, M.D., 1983, Alexandria and District of Mareotis. Graeco-Arabica 2: 199-216. Rodziewicz, M.D., 1990, Taenia and Mareotis: Archaeological Research West of Alexandria. Annual of the Egyptian Society of Greek and Roman Studies 1: 62-78. Rodziewicz, M.D., 1998a, From Alexandria to the West by Land and by Waterways. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athènes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole française d’Athènes, 11-12 décembre 1988: 93-103. Athens, Paris. Rodziewicz, M.D., 1998b, Classification of Wineries from Mareotis. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athènes, organisé par le CNRS,

45

LAKE MAREOTIS: RECONSTRUCTING THE PAST

46

M. HAGGAG: EXCAVATIONS AT MAREA/PHILOXENITÉ

The City of Marea/Philoxenité. Reflections on the Alexandria University Excavations, 1977-1981 Mona Haggag

In 1977 Alexandria University patronized an expedition under the direction of the late Prof. Fawzi El-Fakharani to search for the ancient city of Marea, mentioned by different classical authors as the capital of the Mareotic Nome during the late Pharaonic period (Thucydides I.68; Herodotus II.30; III.12; 15; IV.161; Diodorus I.68.).

Sadek concluded that all the remains identified were not earlier than the 4th-5th centuries AD (Sadek 1978: 67-80). All trial trenches proved to have remains of stone buildings dating to the Byzantine period. Pot sherds were entirely of Byzantine manufacture (Shahin 1983). The same is true for the architectural material, construction methods and techniques. These results raised a major question about the location of the pre-Byzantine town mentioned in classical literature (Thucydides I.68; Herodotus II.30; III.12; 15; IV. 161; Diodorus I.68). . The Pre-Byzantine Remains (Map 1) By exploring the area some 5 km west of the Byzantine remains, earlier remains of Marea proved to exist beyond the main perimeter of the Byzantine site. A group of harbour installations were found on an island to the west.1

Mahmoud Bey El-Falaki located the site of Marea some 45 km west of Alexandria on the southern shores of the western extremity of Lake Mareotis near El-Hawaria (ElFalaki 1872: 85, 86, 91, 96, 97, 101; see other Haggag chapter in this volume). El-Falaki based his identification of the site according to its location on the map of Ptolemy the Geographer (IV,5. 16.17). This location is confirmed by Ball (1942: 25-30), Breccia (1922: 337), De Cosson (1935: 131-135), Rowe (1954: 128-145) and Fraser (1972: 143-146) due to four visible jetties which still project into the lake today. The jetties are associated with a broad ridge of hard limestone extending from east to west across the town parallel to the southern coast of the lake and bordered, from both north and south, by calcarenite land.

The ‘Pr’-Shaped Harbour (Fig. 1, Map 1) The main harbour of the island is a three sided structure projecting into the lake basin about 30.5 m wide and 57 m long. It is built of undressed blocks of limestone of Cyclopean size. A number of the upper blocks in this structure have drilled cylindrical holes which were either bollard holes or mooring rings (Fig. 2). According to the traditions of ancient Egyptian harbour architecture, ports of rectangular plan in the shape of the sign “pr” were designed for harbours facing the open seas. Earlier examples existed in the seaport of Alexandria (Jondet 1916: 13-20; El-Fakharani 1991: 25), the most conspicuous of which is the one inside the Western Harbour mentioned by Strabo as the Kibotos (Geography, Xvii, I, 10), another one is associated with the Island of Antirrhodes which lies inside the Eastern Harbour of Alexandria discovered by Goddio (Goddio, et al. 1998: 21-29, pl. I). This feature consists of two rows

The Survey As part of the 1977 campaign and in collaboration with the University of Guelph, Canada, Mohamed Sadek conducted a proto-magnetometer survey at the site, concentrating on the shoreline area from which the four jetties projected. Data was collected in lines parallel to the lakeshore from west to east in a grid of 800 m long and 100 m wide at intervals of 4 m. Soundings were taken in the area west of the first quay, and some trial trenches were investigated in three locations on the waterfront near the jetties. The trenches went deep into the bedrock which was found at intervals of 4.6-6.00 m below the present ground surface. Fig. 1: Kibotos harbour structure on the south shore of Lake Mareotis (Site 9 – Lake Mareotis Research Project) (photo M. Haggag).

1. It is worthy of notice here that Gauthier, referring to some hieroglyphic text, stated that the ancient city of Marea was located on an island in the lake (Gauthier 1925-1931: vol. III, 53-54).

47

LAKE MAREOTIS: RECONSTRUCTING THE PAST Map 1: Marea in Lake Mareotis (V. Atef).

Fig. 2: Kibotos harbour structure, mooring ring (photo M. Haggag).

Fig. 3: Rock-cut Kibotos harbour structure at Ikingi Mariut (photo M. Haggag).

48

M. HAGGAG: EXCAVATIONS AT MAREA/PHILOXENITÉ of stakes or piles approximately parallel, running southwest to north-east, along the axis of the main branch of the island, that forms a rectangular shaped-harbour against the island’s shoreline. A third one, hewn in the rock at Ikingi Mariut, about 20 km west of Alexandria (Fig. 3), was identified by El-Fakharani2 (2002: 203-208).

to a dromos. Another similar dromos on the opposite side leads to a funerary chamber. The ceiling is carved to imitate wooden beams in the manner common to Pharaonic cemeteries. False doors were also carved in the walls with a shaft in front of each one leading to the burial chamber below. All tombs have been looted. The only remains were a few local pointed bottom jars which are likely to be Late Pharaonic in date (El-Fakharani 1983: 176-178).

The Fort (Map 1) To the south-west of the pr-shaped port remains of a substantial structure built on a monolithic platform about 21 m wide and 24 m long was uncovered. The structure has an L-shaped plan. The walls seem to have consisted of rough core masonry with outer casing walls. The blocks of the outer casing are carefully cut and fitted together in oblique joints without using any mortar except for a foliage layer of non reddish mortar used in some of the courses to facilitate the sliding of such gigantic sized blocks. These external walls are slanting inward at an angle of about 40 degrees, and are based on larger foundation blocks with squared edges. The building is approached from the south by means of a ramp of dressed and carefully fitted stone blocks (Haggag 1984: 277-280). The use of rough masonry for the core while the huge blocks of the facing are very carefully polished, the oblique joints of the blocks, the use of such a thin layer of mortar, the inward slanting of the outer walls as well as the general plan of the building, are clear indications to a Late Pharaonic date. The building is interpreted as the fort of Psammaticus III (525 BC), from which his son Inaros sailed with his army, aided by some Athenian troops (Thucydides I.104), against the Persians in 465 BC, in an unsuccessful attempt to liberate Egypt.

The Byzantine Harbours (Fig. 4, Maps 1 & 2) The focal point of the town to the east of the Pharaonic/ Ptolemaic remains is the Byzantine harbour area which extends 2 km along the lake shore with four well preserved jetties projecting into the lake at an average height of nearly 2 m above the lakebed. The area comprises three harbours marked by the jetties and a natural promontory namely the west port, the middle and the east port. Along the shore there is a stone built seawall occasionally interrupted by either steps or stone ramps leading directly down to the water. The wall extends some 500 m with an average width of 1 m. At the outer extremity of the western quay of the middle port, there are a number of circular foundations about 4 m in diameter. These features are assumed to have been the base of a light beacon or lighthouse that would have aided navigation at night. Some mooring features can be detected on the upper course of this jetty in the shape of cylindrical bollard holes (Petruso & Gabel 1983: 63). The East Port (Fig. 5, Map 2) Although the east port was completely buried under the sands of the lake which by the time of excavation had dried out, digging revealed that it differs from the other ports of Marea in both shape and function. It is located between the promontory and a north-easterly island which has its own landing place formed by a short artificial dyke to the north-east and a natural promontory to the northwest. This northern island is connected to another smaller one to its south-west. A third quay connects this island to the promontory forming a wide entrance to the eastern port, at least in its first stage of construction. In the second stage, the shape, size and function of the port seem to have been modified. Almost 100 m of the quay that extended from the eastern small island to the promontory were removed and an extension to this quay constructed to the south. It was located parallel and adjacent to the eastern shore of the promontory and was about 200 m in length. Thirty metres before reaching a third island to the south, the quay ends forming an opening between its southern extremity and this small island. This opening provides an exit or entrance to the east port by way of the southern body of the lake, which extended considerably inland at this point. By this modification, the very long quays of the east port became totally isolated from both the mainland and the promontory. Accordingly, the link between the anchored ships with their cargoes and the town was severed. This fact is very important as far as the function of the port is concerned.

The Ptolemaic Quay (Map 1) To the east of the island, a stone built jetty about 104 m long extends into the lake. Near its southern end, the jetty inclines to the east at an angle of about 175 degrees to extend a further 20 m in length. It is constructed of local regularly sized rectangular blocks, arranged in alternate courses of headers and stretchers (Haggag 1984: 280-283). The fine workmanship, medium size nature of the blocks as well as the ashlar masonry construction style with its vertical joints, are indications of a Ptolemaic date for the quay. The Cemetery (Map 1) To the southwest of the island on the limestone ridge, the team came upon a group of burials that have nothing in common with Byzantine tomb types. The cemetery consists of some anthropoid pit tombs. A square shaft tomb is carved into the limestone ridge to a depth of 6 m. Loculi with gabled ceiling are cut either side of the shaft. A nearby chamber tomb is composed of a sloping passage hewn into the rock that ends with six steps leading to an open court. An opening cut in one side of the court leads

2. For detailed discussion of this pr-shaped ports of Pharaonic date as well as examples, see Haggag 1984: 263 ff.; El-Fakharani 1991: 25, fig. 1,2.

49

LAKE MAREOTIS: RECONSTRUCTING THE PAST Map 2: Marea’s public buildings on the lake shore (V. Atef).

Fig. 4: Byzantine harbour of Marea (photo M. Haggag).

El-Fakharani’s interpretation of the function of the east port (1983: 181) is that it worked as a one way traffic harbour for transit navigation: so that seafaring ships carrying goods from different parts of the Mediterranean could enter the east port via its northern opening, their cargoes were to be unloaded on the long quays in order to be reloaded on board river boats to transport goods to the Nile and vice versa.

depth of 0.25 m below the pavement, a square drainage canal, coated with a thick layer of plaster and covered by a row of huge rectangular blocks, extends across the decumanus. This drains into a similar canal that extends under an unpaved side street and from south to north crossing the decumanus at a right angle. Thus, the drainage waters pour directly into the lake (El-Fakharani 1983: 178-179). The Shops (Map 2) The arcade of the decumanus lines a group of 12 constructions (Fig. 7), with different ground plans. The buildings are located side by side with doors that opened out onto the sheltered arcade. They seem to have been built to function as shops with residential quarters for their owners. Remains of white plaster coating are still visible on some of the walls. An upper storey for each of the shops is confirmed by the existence of stairs at the rear of some of the shops and also in some of the back rooms (El-Fakharani 1983: 178-179; Soleiman 2004: 145-164).

The Public Buildings (Map 2) Many, if not most, of the public buildings of Byzantine Marea were mainly erected along the shoreline. Parallel to the lake shore in front of the harbour installations runs the main arcaded street of the city. This decumanus (Fig. 6) extends eastward until it reaches the natural promontory which forms the eastern part of the middle port. It is 10 m wide marked on its southern side by an arcade of which some of the bases of columns can still be seen in situ. The street is paved by big rectangular limestone blocks. At a 50

M. HAGGAG: EXCAVATIONS AT MAREA/PHILOXENITÉ The Basilica (Fig. 8, Map 2) To the west of the decumanus, at the intersection of the western side street, a huge complex of structures came to light during excavations (Sadek 1992: 549-554). The structures consist of two conjoined apsidal basilicas furnished with what seems to be public bathing facilities for sanitary purposes, similar to those discovered and interpreted by Grossman at St Menas (Grossmann 1986: 12-13; Haggag 1984: 284-289). The Bakery At the promontory which forms the eastern extremity of the middle port, a huge building with an area of about 700 m 2 was uncovered. A large granite mill inside one of the building’s multiple rooms indicates the relation of this building with grain grinding processes. Inside one of the rooms, a marble tile with a carved Greek cross was uncovered. In a compartment annexed to one of the rooms in this building, a clay jar containing a hoard of coins was also found. Courses of Byzantine burnt bricks are inserted in the stone masonry of the walls. The building is surrounded by an external wall supported with buttresses typical of Byzantine methods of construction (Moussa 2002: 479480). Fig. 5: (above) Plan of the eastern harbour of Marea, after El-Fakharani 1977 (thanks to El-Fakharani).

Fig. 6: Decumanus between the eastern and middle ports of Marea (photo M. Haggag).

Fig. 7: Plan of shops off the main decumanus (thanks to El-Fakharani). 51

LAKE MAREOTIS: RECONSTRUCTING THE PAST Fig. 8: Basilica, Marea (photo M. Haggag).

Fig. 9: Enigma Buliding, Marea (photo M. Haggag).

The Enigma Building (Fig. 9, Map 2) Next to the western edge of the promontory at its meeting point with the lakeshore along the middle port, an enigmatic building has been uncovered. The building stands completely in the water. It is composed of podia and walls leaving canals 0.75 m broad between them. Some of the canals run in parallel lines with the coast, while others run at right angles with it and the other canals. The entire building is sliding towards the water leaving the outer parts of the podia under water. The canals were studded with small artefacts. The finds, which were mostly intact, included nearly one thousand pieces of small bronze coins, about 200 jugs of red or yellowish clay, St Menas ampulae, other ampulae with crosses, small statuettes of horsemen, fishing hooks and broken glass bottles. This unusual plan made scholars differ in their interpretations of the building’s function. Some suggestions included its possible function as a dry-dock owing to its distinct slope into the lake water (Sadek 1978: 70), but the huge amount of finds inside the canals stands against this hypothesis. Another explanation is that it served as a fishery (Nikkos Lianos pers. comm.). While the discoverer tended to interpret it as a building for votive offerings in honour of both St Menas

and St Marcus, due to the large quantity of coins, ampulae and other artefacts that were found inside (El-Fakharani 1977: 16; Soleiman 2004: 121-138). Boston Field School In 1979, the Boston University’s Summer Term and Study Abroad Program supported an archaeological field school at the site of Marea. Their work lasted for three consecutive seasons, one month each year.3 In addition to the fruitful student training, the mission made new and important discoveries, the most important of which was the Diolkos of Marea. Along the shore between the middle and western jetties, the team of Boston University came upon remnants of a dockyard (Petruso & Gabel 1982: 12). It consists of two sloping walls, each is over 7 m wide and 20 m long,

3. The team was directed jointly by Profs. El-Fakharani & Gable in collaboration with Petruso and the architect, Boyd. The author of this article had the honour of participating in this field school first as an undergraduate trainée and then as an assistant archaeologist.

52

M. HAGGAG: EXCAVATIONS AT MAREA/PHILOXENITÉ Fig. 10: Winery excavated by El-Fakharani 1983 (photo M. Haggag).

pletion of the excavation of the building, it proved to be a huge double peristyle that occupied an area of more than 1,500 m2. Both El-Fakharani (1983: 184-186) and Rodziewicz (1988a: 175-178, 1988b: 267-276) asserted that this discovery represents the first and largest peristyle built for residential purposes to be discovered in Egypt hitherto.

emerging from massive stone foundations going down to the bedrock, to a depth of about 2 m. The runners slip into a central aisle that lies between the jetties. This aisle has some sort of drainage facilities in the form of holes cut into the floor in order to drain water and keep the floor dry. The entire building slides down into the lake basin at a gradient of approximately 1:16 which allows for the use of manpower and log rollers to hall ships out of the water (Petruso & Gabel 1983: 76).

In the middle of the town, some Byzantine dwellings were uncovered. Each is composed of two rooms. Byzantine local pottery ware were scattered inside. Muslim burials were dug inside the room. Green glazed pottery sherds of the Islamic period were found as well as one piece of a Fatimid coin.

The City’s Residential Area Far from the harbour area to the south of the limestone ridge and south of the modern highway leading to Borg ElArab, in the area now called “Hawariya”. During a second season conducted by the Alexandria University expedition team in 1977, various buildings were uncovered in this region which seem to relate to residential and daily life activities. The Peristyle Complex (Map 1) Adjacent to the modern highway (Map 1), some 2.5 km to the south of the harbour, a peristyle court which leads to a variety of rooms, was uncovered. One of the rooms has a staircase leading to an upper floor of the building. It seems that the building went through two stages in its construction; the first stage is indicated by a layer of white plaster coating on the façade. The second stage is marked by the addition of a massive thick walled façade with two windows which have tunnels underneath leading to watercourses that are connected to cisterns. A modern ditch for irrigation water that was dug through the rooms halted a complete investigation of the remains. In 1980, Rodziewicz4 replaced that ditch with a pipe in order to continue uncovering the rest of the building. On com-

Wineries (Map 1) About 200 m to the southwest of the peristyle building, in the middle of the town, a big elaborate winery (Fig. 10) was uncovered (El-Fakharani 1983: 182-184). Another identical in layout but smaller than the first, was also discovered 2 km to the southwest (Arafa 1985: 78-80). Both are of the type known in other places in the Mareotic region.5 They represent the archaeological testimony for the excellence of the Mareotic wine praised by the classical authors (see Dzierzbicka this volume). Each of the two buildings comprises two rooms separated by a low screen wall for squeezing grapes. One is smaller than the other. Both are entirely coated by four layers of red plaster to prevent any possible seepage of the juice. In the center of the smaller room is a raised round base covered with plaster possibly to support a movable squeezer or a hand press. By means of a lion headed spout, the juice from both rooms pours into a large square basin which is dug in the ground, with another smaller and deeper basin dug in its floor for collect-

4. Rodziewicz was then a consultant for the Egyptian Antiquity Service (EAO) and was in charge of the restoration of the area.

5. A third smaller one that lies to the south of the big winery has been reported by Rodziewicz, see Rodziewicz 1998a: 27-36.

53

LAKE MAREOTIS: RECONSTRUCTING THE PAST ing residues. The inner surface of the big basin’s walls is scratched to allow the plaster coating to hold firm. The inner sides of the basin are surrounded by a shelf approached from all sides by a flight of steps. Other steps lead to the basin’s floor. It is thought that this shelf is designed to support wooden beams that held a piece of cloth for filtering the juice poured from the lion shaped spout. In the middle of the north side above the spout, there is a small podium with two funnels pierced each by a hole opening at the basin. El-Fakharani’s hypothesis is that these funnels acted as measures for adding certain amounts of some aromatic flavours that produced the famous taste of Marea’s wine6 (1983: 182-184).

town (Ball 1942: 114, 117, fig. 17, pl. II, III). More than two and a half centuries after Ptolemy, we hear nothing about Marea. From the 5th century onwards, Byzantine Marea appears. During this time, the city has to play another vital role in Egypt’s history. The town in which the Martyr Abu Mina was buried lies about 15 km to the south of Marea. The sanctuary of St Menas was famous for its miraculous healing capacities. The importance of the area increased gradually until it reached its climax at about the end of the 5th and beginning of the 6th centuries (Kaufmann 1910: 103). Flux of pilgrims and patients from all over the Roman provinces headed to the place seeking cures for their ailments (Grossman 1986: 12-13). Flasks (ampulae) filled with the holy waters of Abu Mina were specially made for pilgrims to take with them on their way back home as a blessing.8

The Byzantine Tomb (Map 1) To the north of the railway line, and at the southern boundaries of the inhabited area of Byzantine Marea, lies a rock cut chamber tomb (El-Fakharani 1977: 19-22). The entrance to the tomb is an arched doorway with basalt jams and lintel. It opens onto a passage with a staircase leading to a square mourning chamber. The ceiling of the passage is barrel vaulted with a helicon vault covering the part which turns to the east towards the arched entrance of the mourning chamber. The latter is surrounded by a stone bench and covered with a cross vault with pendentives. Three burial chambers with cross vaulted ceilings are cut into the walls of the tomb, forming a trefoil plan for the Hypogeum.7

The most suitable and more convenient way for those who are coming from both the Mediterranean and the Nile is to reach the lake harbour of Alexandria and navigate the lake to the nearest harbour before taking the land route to St Menas. In this respect, Marea is the nearest point. At that time, Marea’s earlier harbours were isolated on an island and pilgrims had to be ferried to the southern coast of the lake in order to take the caravan route to the sanctuary. To solve this problem, the old town of Marea was shifted 2.5 km to the east of the earlier harbour area, where the line of the ridge inclines far from the lake shore leaving a larger area including a promontory, a bay and several islands in front of the shoreline, that combined offer a naturally sheltered harbour. This area seemed more suitable to quickly establish the new harbour facilities and provide accommodation for pilgrims (Rodziewicz 1998b: 101-103). Other factors may have contributed to the shifting of the site, the most important of which is the changing lake levels, a fact indicated by the different levels of rising water during the flood seasons over many centuries.

The Name of the City More than two centuries before the establishment of Alexandria, Marea, according to Herodotus (II.30, IV-161; III.12, 15), Thucydides (I.104) and Diodorus (I.68) functioned as a strategic staging post on Egypt’s northern frontiers. After the establishment of Alexandria, the former military role of Marea appears to have come to an end and the city takes on a more civic and commercial role (Rodziewicz 1998b: 101-103). Excavation works yielded neither weapons nor fortifications of neither Ptolemaic or Roman date. At the same time, classical authors, such as Athenaeus (I.33), Vergil (Georgics, II.91-92), Horace (Odes, I.37) and Columella (De Rustica, III.24), give us glimpses of the reputation of Mareotic wines. Marea became an agricultural centre as well as an intermediate station for goods moving between Alexandria and the Nile Valley (Rodziewicz 1983: 199-208). After the time of Strabo (XVII.I.14) who spoke of the city’s wealth in papyrus, bean and vine plantation, the city seems to have shrunk to a small village as stated by Athenaeus (I.33) and as could be inferred from Ptolemy the Geographer who mentioned it as “Palae-Marea” by which name he indicates the existence of the earlier Pharaonic

This leads us to the question that Rodziewicz (2003: 27-39) first raised: Was this new site of Byzantine Marea the city of Philoxenité?9 Philoxenité was built by the Praetorian Praefect Philoxenus upon orders of the Emperor Anastasius (AD 491-518) in order to serve the needs of pilgrims on their way to the shrine of St Menas. Rodziewicz counted on the Coptic Encomium in praise of St Menas, first published by Drescher (1946: 35-72; 126-149), to reach the conclusion of interpreting the later site of Byzantine Marea as the city of Philoxenité. According to the text of the Encomium, Drescher placed Philoxenité on the shores of the main basin of the lake, somewhere close to Kom Truga 35 km to the east of the Holy Shrine. Rodziewicz believes

6. For more detailed interpretations, see Rodziewicz 1998a: 27-36. 7. Various rooms on top of the Hypogeum and its vicinity were uncovered by a team of the Supreme Council of Antiquities (SCA) later in 1987/88. They represent a complex of funerary chapels for the tomb (Soleiman 2004: 283-286).

8. Just as Muslims nowadays do with the waters of the holy well of Zamzam at Mecca. 9. Before tackling this problem, it is worth noting that Rodziewicz in addition to his responsibility for the restoration works carried on at the site of Marea, made additionally impressive discoveries at the site.

54

M. HAGGAG: EXCAVATIONS AT MAREA/PHILOXENITÉ Bibliography

that the site of the Byzantine harbours at Marea accurately fit the location of the landing place of pilgrims seeking accommodation, food and beverage, before taking the short caravan route to the shrine.10 In this respect, it is important to point out that the Encomium, which is the sole source we know of to date about the city of Philoxenité, includes the following passage:

Ancient Sources Athenaeus, Deipnosophists, C. Burton (transl.), 1927. London. Augusti Meinekii, 1849, Stephani Byantii, Ehnicorum quae supersunt, I. Berlin. Columella, On Agriculture, H. Boyed (transl.), 1931. London. Diodorus Siculus, Library of History, C.B. Welles (transl.), 1963. London. Herodotus, Histories, H. Carry (transl.), 1912. London. Horace, The Odes, C.H. Bennet (transl.), 1925. London. Ptolemy, Geographica, E.L. Stevenson (transl.), 1932. New York. Strabo, Geography, H.L. Jones (transl.), 1949. London. Thucydides, History of the Peloponnesian War, C.F. Smith (transl.), 1969. London. Virgil, Georgics, J. Jackson (transl.), 1908. Oxford.

“...When it was finished he gave it the name Philoxenité… He ordered to build water stations, where he placed water jars, every ten miles along the route between the hospices and the shrine in order to serve the needs of these multitudes” (Drescher 1949: 15-16).

Yet, the distance between the Byzantine harbours of Marea to the sanctuary do not exceed 15 km, that is less than 10 miles, a distance that can not take more than a few hours if riding on camels and half a day if walking on foot. Then, why did the Praefect order the building of water stations every ten miles along this route?11 Rodziewicz’ hypothesis is mainly based upon believing that there is nothing evident at the site to be dated earlier than the Byzantine era. This assumption has been proved to be incorrect through what was discovered on the island and the ridge by the Alexandria University expedition. More recently our Polish and French colleagues have also uncovered some pre-Byzantine remains (Szymanska & Babraj 2001: 37-42, 2002: 47-58; see Babraj & Szymańska this volume; Pichot 2004). Moreover, if the newly established Byzantine centre had the name of Philoxenité, why did a Christian author, who lived and wrote after the time of Anastasius and his Prefect, Stephanus of Byzantium (Augusti Meinekii 1849: 432), retain the old name of Marea in his writings?12

Secondary Sources Arafa, D., 1985, Wine Production in Egypt from Alexander to the Arab Conquest: As Compared to the GraecoRoman World. Unpub. MA diss., Alexandria University. Ball, J., 1942, Egypt in Classical Geographers. Cairo. Botti, G., 1902, Studi sul III Nomo dell’egitto inferiore E piu specialmente sulla regione Mareotica. Bulletin de la Sociéte Archéologique d’Alexandrie 4: 41-84. Breccia, E., 1922, Alexandrea ad Aegyptum. Bergamo. De Cosson, A., 1935, Mareotis: Being a Short Account of the History and Ancient Monuments of the NorthWestern Desert of Egypt and of Lake Mareotis. London. Drescher, J., 1946, Apa Mena: A Selection of Coptic Texts Relating to Saint Menas. Cairo. Drescher, J., 1949, Topographical Notes for Alexandria and Distric. Bulletin de la Sociéte Archéologique d’Alexandrie 38: 13-20. El-Fakharani, F.A., 1977, A Preliminary Report on the First Season of Excavation at Marea. Excavation Report presented to the Antiquities’ Department, Egypt. Cairo. El-Fakharani, F.A., 1983, Recent Excavation at Marea in Egypt. In G. Grimm, H. Heinen & E. Winter (eds.), Aegyptiaca Trevernsia II, Das Römisch-Byzantinische Ägypten. Akten des internationalen Symposions 26.-30. September 1978 in Trier, 175-186. Mainz. El-Fakharani, F.A., 1991, The Kibotus of Alexandria. Studi Miscellanei 28: 21-28. El-Fakharani, F.A., 2002, The Pharaonic Port on the Mediterranean: Its Shape, Development and Importance. In Z.A. Hawass & L.P. Brock (eds.), Egyptology at the Dawn of the Twenty-first Century, Proceedings of the 8th International Congress of Egyptologists, Vol. 2: 203-208. Cairo. El-Falaki, Mahmoud Bey, 1872, Memoire sur l’antique Alexandrie, ses faubourgs, ses environs. Copenhagen. Empereur, J-Y., 1998, Alexandria Rediscovered. New York.

Finally, one would say that it is not unusual for a certain town in antiquity to have more than one name over the different periods of its history. Rhacotis/Alexandria, Thonis/ Heracleion, are very distinctive examples in this respect. It seems reasonable that the old town of Marea shifted eastward to build new facilities suitable for the new role the city had to play.13 Whether the new site took a new name or not is a question that cannot be answered yet with any certainty. In this respect, further readings of the original Encomium are required as well as any other material evidence that may appear through investigations currently being undertaken by different missions at the site.

10. Haas and Empereur share Rodziewicz’s opinion, see Haas 1997: 349; Empereur 1998: 229-239. 11. Whether this is due to miscalculations of the wording of the text, as Rodziewicz mentions, it is something that cannot be judged unless further readings of the original Encomium are undertaken, see Rodziewicz 2003: 27-39. 12. It is not evident that the various citations of Marea in the ancient literature refer to exactly the same spot as in the case of Procopius: Iacobus Haury 1913: vol. III 2, 171. We find some confusion in the use of “Marea” and “Mareotis” used to describe both the town and the lake or sometimes the whole lake district (Petruso & Gabel 1980: 1-27). 13. In 1902 Botti stated that there were two towns named Marea, the ancient and the new (Botti 1902: 73-75).

55

LAKE MAREOTIS: RECONSTRUCTING THE PAST Fraser, P.M., 1972, Ptolemaic Alexandria, Vol. I. Oxford. Gauthier, H., 1925-1931, Dictionnaire des nomes géographiques contenus dans les textes Hiéroglyphiques, Vol. III. Le Caire. Goddio, F., Bernard, A., Bernard, E., Darwish, I, Kiss, Z., & Yoyotte, J., 1998, Alexandria: The Submerged Royal Quarters. Oxford. Grossmann, P., 1986, Abu Mina: A Guide to the Ancient Pilgrimage Center. Cairo. Haas, C., 1997, Alexandria in Late Antiquity: Topography and Social Conflict. Baltimore. Haggag, M., 1984, Ports in Ancient Egypt till the Arab Conquest. Unpub. MA diss., Alexandria University. Haggag, M., 2002, Two Religious Buildings at Byzantine Marea. In Z.A. Hawass & L.P. Brock (eds.), Egyp tology at the Dawn of the Twenty-first Century, Proceedings of the 8th International Congress of Egyptologists, Vol. 2: 284-289. Cairo. Jondet, M.G., 1916, Les Ports Submergés de l’ancienne Ile de Pharos. Le Caire. Kaufmann, C.M., 1910, Die Menasstadt und der national heiligtum der altchristlichen Ägypter, I. Leipzig. Moussa, F.S., 2002, Quelques aspects de la vie quotidiènne représentés á Marea Byzantine. In Z.A. Hawass & L.P. Brock (eds.), Egyptology at the Dawn of the Twenty-first Century, Proceedings of the 8th International Congress of Egyptologists, Vol. 2: 478-486. Cairo. Petruso, K.M., & Gabel, C., 1980, Marea: An Environmental and Cultural Study at Lake Maryut, Lower Egypt: A Research Prospectus. Boston University African Studies Center Working Papers 25: 1-27. Petruso, K.M., & Gabel, C., 1982, Marea: A Byzantine Port in Northern Egypt. Boston University African Studies Centre, Working Papers 62: 1-23. Petruso, K.M., & Gabel, C., 1983, Marea: A Byzantine Port on Egypt’s Northern Frontiers. Archaeology (Sept./Oct.): 62-77. Pichot, V., 2004, La fouille de l’Ile de Marea: Le site, Pros pections et campagnes de fouille: http://www.cealex. org/sitecealex/activities/terrestre/marea/genef.htm, CEAlex (accessed March 2008). Rodziewicz, M.D., 1983, Alexandria and District of Mareotis. Graeco-Arabica 2: 199-208.

Rodziewicz, M.D., 1988a, Remarks to the Peristyle House in Alexandria and Mareotis. Praktika, International Congress of Classical Archaeology in Athens, September 4th -10th, 1983: 175-178. Athens. Rodziewicz, M.D., 1988b, Remarks on the Domestic and Monastic Architecture in Alexandria and Surroundings. In E.C.M. vanden Brink (ed.), The Archaeology of the Nile Delta: 267-276. Amsterdam. Rodziewicz, M.D., 1998a, Classification of Wineries from Mareotis. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athènes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole française d’Athènes,11-12 décembre 1988: 27-36. Athens, Paris. Rodziewicz, M.D., 1998b, From Alexandria to the West by Land and by Waterways. In J.-Y. Empereur (ed.), Commérce et artisant dans l’Alexandrie héllénistique et romaine. Actes du colloque d’Athènes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole française d’Athènes,11-12 décembre 1988: 95-102. Athens, Paris. Rodziewicz, M.D., 2003, Philoxenité - Pilgrimage Harbour of Abu Mina. Bulletin de la Sociéte Archéologique d’Alexandrie 47: 27-47 Rowe, A., 1954, A Contribution to the Archaeology of the Western Desert: II. Bulletin of the John Rylands Library 36: 128-145. Sadek, M., 1978, The Ancient Port of Marea. Cahiers des Études Anciennes VIII: 67-80. Sadek, M., 1992, The Baths at the Ancient Harbour of Marea. The Sesto Congresso Internazionale di Egittologia, Vol. I: 549-554. Turin. Shahin, B., 1983, Local Pottery in Byzantine Egypt: A Study of the finds at the City of Marea. Unpub. MA diss., Alexandria University. Soleiman, N.M.S., 2004, Marea: An Archaeological Study and the Manner of its Tourist Investment. Unpub. PhD diss., Alexandria University. Szymanska, H., & Babraj, K., 2001, Marea: First Interim Report, 2000. Polish Archaeology in the Mediterranean XII: 37-42. Szymanska, H., & Babraj, K., 2002, Marea: Second Interim Report, 2001. Polish Archaeology in the Mediterranean XIII: 47-58.

56

V. PICHOT: MAREA PENINSULA

Marea Peninsula: Occupation and Workshop Activities on the Shores of Lake Mariout in the Work of the Center d’Études Alexandrines (CEAlex, CNRS USR 3134) Valérie Pichot General Introduction to the Site of Marea The site, identified as Marea on the plate of Alexandria in the Atlas géographique d’Egypte1 (Description de l’Egypt 1809: Pl. 10) and then by Mahmoud Bey El-Falaki in 1866 (El-Falaki 1872: 96), is one of the rare examples of a harbour town on the shores of Lake Mariout that has essentially remained untouched since antiquity and accessible for archaeological studies. Situated some 40 km to the south-west of Alexandria, it stretches more than 25 ha along the southern shore of Lake Mariout at a point where

the width of the lake from north to south is less than one kilometre (Fig. 1). Since 1977 the site of Marea has been the subject of archaeological excavations and those discoveries made prior to 2003 all suggested a somewhat late occupation of the site. The excavated remains, dating from the 5th to 7th centuries AD, indicate a harbour town of considerable capacity with large storage facilities as well as public buildings of high quality (Fig. 2) (El-Fakharani 1983;

Fig. 1: Map of Egypt: location of Lake Mariout and the site of Marea (CEAlex Archives).

1. Hydrographic map of Lower Egypt drawn in 1801-1802, Description de l’Egypte, Etat moderne, vol. I, Paris, 1809, plate 10.

57

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig. 2: General plan of the site identified as Marea (CEAlex Archives).

Petruso & Gabel 1983; Sadek 1992; De Cosson 1935: 131; Rodziewicz 1983, 1998; Szymanska & Babraj 2008).

will be more precise once we have the results of the ceramological and numismatic studies presently under way.

Since 2003, the Centre d’Etudes Alexandrines (CEAlex) has been working on the peninsula situated some 100 m to the northeast of the ancient town of Marea: topographical surveys, prospecting on foot, geophysical examinations and archaeological excavations have all begun to reveal the general organisation of the peninsula and the existence of a large workshop quarter that extended over much of its surface.

An imposing building constructed of large blocks and hydraulic mortar stood on this causeway about 100 m before its junction with the peninsula. It is usually interpreted as a lighthouse or landmark. A jetty stretching 26 m into the lake is situated to the south-east of this (Figs. 3 & 7). The ancient causeway is far from straight. Oriented south/north for less than 300 m, it takes a turn to the east to follow, firstly, a south-west/north-east direction then west/east to the point of the lighthouse or landmark. It then returns to the south-west/north-east and reaches a workshop quarter. The rather particular shape of the causeway and the jetty to the south-east of the landmark create a mooring basin that is relatively well protected from the prevailing winds.

General Organisation of the Peninsula (Fig. 3) Originally an island, access to the peninsula was possible either from the lake onto a pier to the north that allowed for boats to moor, or by land across a causeway that was some 5 m wide that connected with the mainland (Figs. 4a & 4b).

On the north of the peninsula there was a building over 100 m long also built of large blocks and hydraulic mortar (Fig. 3). Its northern limit ends in a system of terraces and stairways cut into the rock and leading to a pier. Its general layout around a large rectangular courtyard might suggest a public building for commercial purposes concerned with trade and storage, though it might also be a rich villa. The presence of cistern(s) under this building is attested by a descending passage situated in the centre of the west wing. Structures in brick and hydraulic mortar visible on the surface of the east wing could be connected with the cisterns, or could be evidence of bath facilities. Surface surveys and drawings have not yet made it possible to understand the exact role and organisation of this building or the function of its associated hydraulic system.

In 2003 the first two excavation sectors were established upon this causeway. Sector 1 (Figs. 3 & 5) was opened on a very badly preserved part of the track. This excavation allowed us to reveal the presence of foundations of a structure probably associated with controlling movement on and off the causeway. Sector 2 (Figs. 3 & 6) was established at the point where the track disappears toward the southern part of the concession. The causeway effectively stops at this point and ends in a sort of platform made of numerous amphora fragments and mortar. The archaeological material found here was relatively homogenous and dates the construction of the causeway to between the 1st and 3rd centuries AD. The chronology of its construction 58

V. PICHOT: MAREA PENINSULA

Fig. 3: The peninsula of Marea (CEAlex Archives). 59

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig. 4: A) (above) pier to the north of Marea peninsula, view from the north-west. B) (right) ancient causeway connecting the peninsula to the mainland, view from the west. Photo V. Pichot (CEAlex Archives). Fig. 5: (left) Ancient causeway under excavation of Sector 1, view from the east. Photo V. Pichot (CEAlex Archives).

Fig. 6: Ancient causeway under excavation of Sector 2, view from the north. Photo V. Pichot (CEAlex Archives).

Fig. 7: Landmark and jetty to the south of the peninsula, view from the south-west. Photo V. Merle (CEAlex Archives).

60

V. PICHOT: MAREA PENINSULA Occupation and Workshop Activities CEAlex is currently focusing upon the central part of the peninsula where a quarter developed and evolved during the Hellenistic and Roman periods. This was principally a workshop quarter the presence of which is attested from the 3rd century BC.

a later period, occupation re-adopts the north-north-west/ south-south-east orientation present during the Hellenistic period on the elevated part of the peninsula. Survey of the peninsula led to the discovery of numerous archaeological vestiges connected to light industry. Many zones with concentrations of slag and furnace walls, point out the presence of what were workshops, the greater part of which appear to have been connected to metallurgic activities. The numerous plots showing clear anomalies detected by the magnetic survey certainly correspond to areas of intense metallurgic activity. Other anomalies may possibly represent the presence of potteries and more certainly chalk furnaces from the later era (5th-6th centuries AD), as has been shown by the excavations.

Geophysical magnetic surveys undertaken on the central part of the peninsula have revealed the existence of three major orientations, subsequently confirmed by excavation, that correspond more or less to the layout of the area at different periods (Fig. 8). In the Hellenistic period (3rd-1st centuries BC) there are two orientations: one south-west/ north-east for the low-lying part of the peninsula and the other north-north-west/south-south-east for the elevated part. At the end of the Hellenistic period and the beginning of the Roman era, the major orientation used within the low-lying areas follows a north-north-east/south-southwest axis, while the north-north-west/south-south-east orientation appears to continue in the elevated areas. At

Metallurgic activities The excavations undertaken in Sector 4 (see Fig. 3) have revealed extensive furnace activity: extensive not so much in terms of production, which is still difficult to estimate,

Fig. 8: Results of the geophysical survey undertaken on the Marea peninsula. (T. Herbich & CEAlex Archives). 61

LAKE MAREOTIS: RECONSTRUCTING THE PAST but rather in the number of hearths discovered. In fact, in this sector where explorations both geophysical and visual suggested an empty terrain, except for a ditch situated to the western edge, some 40 hearths have been unearthed most of which are lying a mere 0.15 m beneath the surface and all are concentrated in the west and north-western part of the sector. Of different sizes – the size depending to a large part on the extent of destruction – they are either circular in plan or occasionally rectangular. While certain examples still hold fragments of objects and/or slag, others are practically destroyed. They are not all contemporary, but they may have functioned in groups of two, three or four, several groups being in operation at the same time. The best conserved hearths were cleared when excavating rubbish ditch FS4024 situated on the western edge of the sector. Its fill composed fragments of amphora, ceramics, bones, some fragments of faience objects, charcoal, some fragments of iron and seashell. The majority of the material unearthed was characteristic of the 1st centuries BC and AD, however, certain elements can be dated to an earlier Hellenistic period, e.g. a worked lamp of the 3rd century BC and amphora and ceramic fragments of the 3rd and 2nd centuries BC.

Fig. 9: Sector 4 - Hearths in FS4024, under excavation. Photo V. Pichot (CEAlex Archives). tic period occupation itself potentially related to workshop activities, can be broadly dated to the first centuries of the Roman era. Sector 3, presently under excavation (see Fig. 3), was opened in 2003 on the western part of the island and was extended in 2007 into a zone of noticeable magnetic anomalies towards the east. Work on the western part was temporarily suspended to allow, firstly, for the completion of the study of Sector 4 and then the excavation of the eastern part of Sector 3. In 2003 two ensembles of buildings separated by an alley were unearthed in the western part of the sector (Fig. 10). The ensemble oriented north-north-east/ south-south-west was disturbed in its north-eastern part by large constructions orientated on a north-north-west/ south-south-east axis. Each building is composed of small workshop rooms opening onto the alley, while at the back there was a large room used for storage or habitation. One of the activities identified in this workshop quarter is polymetallurgical (Cu and Fe). This is attested by the remains of hearths, flooring and occupation layers characteristic to this type of activity, but also by the discovery of refuse material connected to metallurgical activity (slags, drips, hammerscales, unfinished objects, etc.).

This ditch (FS4024) was cut by several groups of hearths (Fig. 9). Among these structures, two features are quite well preserved (FR4034 and FR4035). Although made up of hearths with different plans – one is composed of rectangular hearths, the other of circular – they are constructed in a similar fashion: a hearth of 0.30 m diameter to which is attached a much smaller but deeper hearth, the exact function of which remains to be determined. These features were constructed of clay in which fragments of ceramics and amphorae were placed to act as support. The structures are well constructed and the interior walls are perfectly smooth. Their fill is of charcoal, baked clay, small fragments of iron and small furnace slag. The debris discovered suggest that metal working activity was taking place, associated with the fabrication and/or repairing of small iron objects with the use, in certain cases, of copper alloys and lead. Associated with the hearths, traces of post-holes reveal the presence of light structures that could have been more or less permanent. Stake holes near the hearths, however, indicate the probable use of fireguards, windbreaks and/or walls marking out certain specific zones, e.g. for storage.

Four workspaces are presently being excavated. The outline of walls marking out storage and circulation zones, have been recognised in three of these rooms. In Space 10 a series of re-laid beaten earth floors have been revealed beneath the last occupation level. According to the initial studies of the archaeological material, it seems that levels of workshop activity can be dated to the 1st century BC to the 1st century AD. A limited trench has been opened in the low-lying zone to the south of Sector 3. The water table was quickly reached at a little more than 1 m below the surface. Three successive floors and associated metallurgic activity layers (lots of charcoal, numerous small metal drips) dating to the 2nd to 1st centuries BC were revealed, as well as a wall whose orientation revealed the existence of an ensemble laid out on a north-west/south-west axis.

The remains of certain furnaces consisted of no more than the very bottom traces of stakes and posts, a few strips of beaten earth floors and a part of the first foundation course of a wall (MR4040 oriented west/east) set upon the bedrock. The vestiges of this sector are very worn down, hence the difficulty in envisaging the spatio-temporal organisation of these workshops. Without going into considerations of contemporanity of function and the length of occupation, and having yet to complete the study of the archaeological material, the important metal working activity in this zone, implanted upon the remains of Hellenis62

V. PICHOT: MAREA PENINSULA

Fig. 10: Overall plan of the western zone of Sector 3 (CEAlex Archives). 63

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig. 11: Overall plan of the eastern zone of Sector 3 (CEAlex Archives). Area Devoted to Cult Activity This occupation is contemporary with that of a construction (ST300) partially unearthed in 2007 in the eastern part of Sector 3 (Figs. 11 & 12). A paved way some 1.15 m wide situated to the north, leads to the entry of the building. The building2 is rectangular in shape approximately 10.35 m by 11.75 m, oriented north-north-west/southsouth-east along its longitudinal axis. The peripheral walls of an average thickness of 1.4 m and with concave external surfaces, enclose a space divided into several cells separated by walls that are 0.8 and 0.9 m thick.

The building consists of five different spaces. From the “outside” one enters into a large rectangular room of 3.5 m by 7.55 m, divided into two by a pavement of slabs that cuts through the middle. This pavement is approximately 0.85 m wide, thus narrower than the opening of the doorway (1 m approx.), and is paved with irregular limestone slabs. It connects the entrance with the only door within the building that is pierced through the interior transversal cross-wall. Its slight irregularity and the fact that is deliberately off the axis of the room, leads one to consider the nature of the flooring that once lay on either side. There are few clues left to satisfy this query. Three fragments of slabs attached to the pavement at three different points along its line might indicate that the missing flooring was made up of the same materiel as the pavement itself, at least on the edges. The walls of this room were covered with a plaster painted to represent alabaster. The remains of another painted coating made to represent red, white and black marble facing might have belonged to a later stage or more probably to the decoration of a room on the upper floor. In front of the interior door, a rectangular cavity 0.33 m by 0.18 m had been dug into one of the slabs of the pavement. It is centred on the median axis of the doorway and in the bottom one can see in the middle, a small depression, oval in plan and conical in section, which is filled with a plug of unbaked clay. This rectangular cavity

In the north-eastern part of the building, the double-faced masonry is composed of medium sized cut blocks of limestone with an inner filling. This construction method differs clearly from the opus incertum of the other walls of the construction. This could represent an act of restoration during a possible second phase of the building’s use. The upper part of the walls were built of mud-brick that were found fallen and disintegrated in the interior of the building.

2. The preliminary study of this structure was undertaken by I. Hairy, architect-archaeologist, CNRS, USR3134, Centre d’Etudes Alexandrines.

64

V. PICHOT: MAREA PENINSULA Fig. 12: (left) Eastern zone of Sector 3 under excavation. Photo V. Pichot (CEAlex Archives). Fig. 13: (below) Sector 3 - One of the two limestone sphinxes discovered in the tower building ST300. Photo V. Pichot (CEAlex Archives). Fig. 14: (bottom) Sector 3 Bronze candelabra after restoration in the CEAlex laboratory. Photo A. Hussein (CEAlex Archives). probably held a support, a base bearing either an element of decoration, e.g. statuary, or a functional object. The interior doorway was framed by two pilasters, of which only the bases remain, that bore two little, limestone sphinxes (Fig. 13) that were found dismantled in the nextdoor room. This interior doorway opens onto a room some 2.40/2.35 m by 4.45 m. At the end of this space, to the south, two small assemblies of blocks, the sides of which have been covered by cleanly cut stone facing, stood against the walls in the corners. These two assemblies form the base of two staircases that led to the upper floors. The two lateral spaces situated to each side of this distribution space are completely closed and inaccessible at this level. The eastern space was filled with clay and other rubble, while within the filling of the other space there is a block of masonry in opus incertum that most probably supported the weight of a now-disappeared superstructure. When considering the different aspects of this construction, its interior layout, as well as the archaeological material discovered here, a hypothesis may be that the function of this building could relate to the representations of the “tower-house” on Nilotic mosaics, or that it was religious. This building was part of a much bigger ensemble, the extension of which to the north-west was partially uncovered by excavation in 2008 and was still relatively intact beneath the construction levels of building ST301. Two spaces clearly similar in size and limited to the east by wall MR3114, were discovered. Their western limit has not yet been determined. The most southerly space is difficult to interpret. It was practically destroyed in its entirety by the implantation of wall MR3082 that lies in part upon wall MR3121 that separates the two spaces. The excavation of the northern space, delimited to the north by MR3116, has revealed the remains of flooring of compact clay in which were set several ceramic storage jars. Although we do not yet know for sure the duration of its use, the evidence of the archaeological material, including a Hellenistic lamp with seven wick-holes, a faience pendant and a bronze candelabra (Fig. 14), would suggest that this ensemble was 65

LAKE MAREOTIS: RECONSTRUCTING THE PAST settlement and organisation of a sizeable town on the lakeshore. Up until the present, our knowledge of the occupation of the site of Marea was limited to the Byzantine period (5th to 7th centuries AD). The work of the CEAlex on the peninsula has proven a settlement existed here well before the 5th century AD. The excavations undertaken in Sectors 3 and 4 have revealed a large occupation dating to the end of the Hellenistic and the beginning of the Roman era, as well as numerous signs of a Hellenistic occupation prior to the 2nd century BC. The opportunity at Marea to excavate and study a workshop quarter of this size, in an environment so well defined by the fact that it stands on a peninsula, is really quite exceptional. In addition to gathering information on the production itself and on the production lines, it is also possible to study the spatio-temporal organisation of the site in the very heart of the peninsula and to resituate it within its immediate environment (habitation, religious area, etc.). Bibliography De Cosson, A., 1935, Mareotis: Being a Short Account of the History and Ancient Monuments of the North-western Desert of Egypt and of Lake Mareotis. London. Description de l’Egypte, Etat moderne, Vol. 1, 1809. Paris. El-Fakharani, F.A., 1983, Recent Excavation at Marea in Egypt. In G. Grimm, H. Heinen & E. Winter (eds.), Aegyptiaca Trevernsia II, Das Römisch-Byzantinische Ägypten. Akten des internationalen Symposions 26.-30. September 1978 in Trier, 175-186. Mainz. El-Falaki, Mahmoud Bey, 1872, Memoire sur l’antique Alexandrie, ses faubourgs, ses environs. Copenhagen. Petruso, K., & Gabel, C., 1983, Marea: A Byzantine Port on Egypt’s Northwestern Frontier. Archaeology 36.5: 62-63, 76-77. Rodziewicz, M.D., 1983, Alexandria and district of Mareotis. Graeco-Arabica 2: 199-216. Rodziewicz, M.D., 1998, From Alexandria to the West by Land and by Waterways. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athènes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole française d’Athènes, 11-12 décembre 1988: 93-103. Athens, Paris. Sadek, M., 1992, The baths at the ancient harbour of Marea. Sesto Congresso internazionale di Egittologia, Vol. I: 549-554. Strabo, Geography, A. Meineke (transl.), 1877. Lipsiae. Szymańska, H., & Babraj, K. (eds.), 2008, Byzantine Marea. Excavations in 2000-2003 and 2006, Marea, Vol. 1. Kraków.

Fig. 15: Sector 3 - Hoe chalk burner. Photo V. Pichot (CEAlex Archives). in use from the end of the 3rd century BC. It was partially destroyed by the installation of one or several occupations to the north during the Roman period that were in part connected to workshop activity,3 and thereafter by chalk burners that were active in a later period (Fig. 15), most probably in the 5th and 6th centuries AD, and that largely destroyed all construction in this zone. Conclusion The Mareotis area, today semi-desert, was once a fertile region with widespread agricultural activity. In the Graeco-Roman era, Lake Mariout was a veritable inland sea with intense traffic. As a zone of movement and exchange, it was connected to the Nile by canals and with the sea at several points through the course of its history. A passage also allowed traffic access to the canal of Alexandria. The Ptolemaic capital had a lakeside port with quays and warehouses that Strabo (17.1.7) considered richer than the maritime harbours of the city. Marea is one of the best examples of Mariotic lakeside towns whose development was tied to exchange between Alexandria and its hinterland: exports of agricultural production and local industry (glass, metal, etc.), imports of raw materials and pilgrimages during late antiquity. Its state of preservation allows us to study the problems of

3. The mission of 2008 has brought to light numerous structures connected to workshop activities of the Roman era to the west and north of the tower building. They are presently under study and will be the object of a supplementary mission in 2009.

66

M. RODZIEWICZ: MAREA AND PHILOXENITE’S ARCHAEOLOGICAL EVIDENCE

On Interpretations of Archaeological Evidence Concerning Marea and Philoxenite Mieczyslaw D. Rodziewicz

The largest and best preserved ancient lake harbour southwest of Alexandria was located on the map of the Mariout region by Mahmoud El-Falaki in 1872 with a description and interpretation of the site as the centre of the ancient Mareotic district being identified as the site of Marea (ElFalaki 1872: 96). For a long time this interpretation was considered as correct and unshakable. It was expected that with continuous advances and development of field archaeology in the region, this interpretation would be substantiated by more archaeological evidence. However, in the first half of the past century the site was not investigated by the most active archaeologists in Alexandria such as Breccia and Adriani. Instead, two other sites in the Mareotic region captured the attention of these great archaeologists of the last century. The first such place was Abu Mina discovered in the desert south of Alexandria and excavated from 1905 by German archaeologist C.M. Kaufmann (1908). The second important site was the large city of Taposiris Magna, west of Alexandria, that spread around the Ptolemaic temple of Osiris, with the neighbouring site of Plinthine, where Breccia (1922: 353) and Adriani (1940) concentrated their field research in the first three decades of the last century. There was also a short German campaign near Amreyia, situated in the desert south of Alexandria, where very informative architectural remains of early Christian date have been uncovered (Eilmann, et al. 1930).

ten by Breccia (1922: 335) and Forster (1922), the port of Marea is hardly mentioned compared with such important places west and south of Alexandria as Taposiris Magna and Abu Mina. The best equipped excavations in the Southern Desert were made by expeditions organized by the German Archaeological Institute in Cairo in the 1930s. They were carried out near the modern market place and settlement of Amreyia (Eilmann, et al. 1930). Specific rural types of early Christian structures characteristic of the region, although rather modest and of an introductory nature, were recorded and published very quickly. It is a great loss that the excavators did not come back to the region after the second world war to continue the research. In the second half of the past century, excavations at Abu Mina, undertaken by the German Archaeological Institute at Cairo, were resumed under the direction of architects, Müller-Wiener and later Grossmann. They extended the area of research around Abu Mina, the largest pilgrimage centre of early Christian Egypt, and surveyed the neighbouring territory of the Lake Mariout harbour. Müller-Wiener (1967) and Röder (1967), devoted their time to investigating not only the ancient settlements in the region, but also the sources of local building material, including ancient quarries located along the southern shore of the lake, particularly one located near the modern village of Bahig (Müller-Wiener 1967: 104-117; Röder 1967: 118-131). Their interests also extended to an analysis of the building materials visible on the surface of the land along the southern shores of Lake Mariout and the surrounding area, with the clear intention of identifying a link between the chain of settlements in the desert and the pilgrimage centre at Abu Mina. The article of Müller-Wiener (1967), entitled “Siedlungsformen in der Mareotis”, was based on rational observation and the description of visible remnants of ancient structures on the surface (Figs. 1 & 2). In the harbour of so-called Marea, prior to any excavations, he identified on the surface, a double bath, structured frames of the quays, with accompanying buildings, and the most important construction, the monumental three apsidal church (Müller-Wiener 1967: 106).1 All his identifications were correct, and they have not altered as a result of later excavations. More im-

The most complete description of the whole area was offered by Anthony De Cosson, the former director of the railway west of Alexandria, in the 1930s. His book on Mareotis was based on then available literature on the subject and his extensive knowledge of the land from Alexandria westwards far beyond Marsa Matruh (ancient Parethonium) (De Cosson 1935: 131). In his description of Marea he follows the description and name of the extensive lake port remains situated south-west of Alexandria given by Mahmoud El-Falaki. In both, El-Falaki’s and De Cosson´s descriptions, the site was presented as 1.5 km long, but not very broad, and limited to the chain of visible ruins of fallen walls along the lake shore. Nevertheless, in their opinions the place was the most important centre of the historical Mareotic region, as mentioned in ancient literary sources. Other modern publications concerning ancient Mareotis have dealt very little with the largest lake harbour in the region. Yet, in the first half of the last century the site was recorded on several maps of the Survey of Egypt series, but was rarely presented in published photographs. The best known are pictures showing the western jetty of the harbour surrounded not by water, but instead by marshes. In the most popular books of that time writ-

1. Műller-Wiener 1967: 106, n. 16: “In Marea sind – obwohl das aufgehende Maurerwerk weitgehend fortgeschleppt zu sein scheint –noch zahlreiche Bauten erkrnnbar (grosses Doppelbad, Hafenbefestigungen mit Kaianlagen und drei langen Molen, Kirche mit drei-apsidialem Grundriss usw.usw.); eine nähere Untersuchung des Ortes wäre – zumal angesichts seiner Bedeutung in der Frühzeit –äusserst wünschenswert.”

67

LAKE MAREOTIS: RECONSTRUCTING THE PAST Fig. 1: Central part of the plan of Mareotis drawn by MüllerWiener in 1966, with the pilgrimage road from the lake harbour to Abu Mina and settlements, among them settlement “M” in Huwaryia village (after MüllerWiener 1967: Fig.1).

Fig. 2: Abu Mina with surrounding settlements/karms, and pilgrimage road towards the lake harbour (after Müller-Wiener 1967: Fig. 4).

68

M. RODZIEWICZ: MAREA AND PHILOXENITE’S ARCHAEOLOGICAL EVIDENCE portantly, he identified a separate settlement (“M” on the plan reproduced on his Fig. 1) (see Fig. 1), at a distance of 1.5 km south from the ruins of the lake harbour known as Marea (Müller-Wiener 1967: 110). He presented its dimensions (600 x 800 m) and described briefly the remains of a brick cistern located at the edge of the modern road, the same that later was identified as a water reservoir serving the double peristyle building, published by its first excavator El-Fakharani, as the Byzantine House (El-Fakharani 1983: 175). Thus, Müller-Wiener clearly separated all these inland situated structures from the harbour to the north, and very clearly stated that settlement “M” was not a part of a larger town plan comparable with ancient Alexandria, as was to be suggested a decade later, and supported by his successor heading the excavations in Abu Mina (Grossmann 2003: 13).

saved most of the monuments excavated by the Alexandria University Expedition, headed by Prof. El-Fakharani. The winery in Huwaryia village, excavated in 1977, received limited reconstruction (mainly for protection but also didactic reasons), and was covered with a roof. The excavation of the so called Byzantine House, which most probably served as a hospice for pilgrims heading to Abu Mina, was completed. It was partly reconstructed to prevent further damage and also for didactic requirements. The remains of buildings at the lake harbour were treated similarly (Rodziewicz 2002, 2003). During these works many previously completely unrecognised structures were identified, thus extending the possibilities of defining more features of the local topography, its culture, economy and technology. If they would have been known to Prof. ElFakharani at the time of his activities in the area, he would certainly have changed some of his first hypothetical identifications, and his general opinion of the site.

In 1978 the first archaeological excavations of the socalled site of Marea were financed by Alexandria University under the direction of Professor of Archaeology, Dr Fawzi El-Fakharani, a very energetic organizer and gifted speaker. In a very short time his discoveries at the site became very popular both nationally and internationally. Prof. El-Fakharani published in 1983 an extensive report on the survey and excavations of a vast area of Mareotis, about 6 km long and 4 km broad. Other publications of the site of “Marea” by Sadek (1978: 67) and Petruso and Gabel (1983: 62) followed, although analysis of available material and structures excavated there were not completed. At that time, there arose the notion of a very large and prosperous city of Marea around the existing remains of the lake harbour, which some supposed to be equal in size to ancient Alexandria. The Mareotic settlements 1.5 km south of the port, and the Ptolemaic rock-hewn tombs situated about 5 km west (at the site called Quassimyia), previously investigated by Müller-Wiener, were included by El-Fakharani in the city of “Marea” (El-Fakharani 1983: 176, 186). The lack of any continuation and evolution of chronologically parallel architectural structures between these distant places was not taken into consideration by the enthusiasts of this notion of the great city of “Marea”.

The most required change in the general knowledge of the site concerns the extension of the urban area of “Marea” to the west, i.e. to the tombs of the Ptolemaic period (at modern Qassimiyia site, ca 5 km from the city), which is still considered by some archaeologists as a part of the western necropolis of the capital of the whole Mareotis region (El-Fakharani 1983: 176; Bagnall & Rathbone 2004: 75).2 Of three hypogea, still well preserved in the 1970s, only traces of the largest, located on the southern slope of the rocky ridge, are still recognisable. The identification by El-Fakharani of Qassimiyia site as the western necropolis of “Marea” is not archaeologically substantiated, because similar hypogea existed on the rocky ridge near Huwaryia village (1.5 km from “Marea”), and all along the same ridge were located other types of tombs such as chambers, shafts or pit tombs, which were dispersed over the area, also considered by some archaeologists as belonging to the city of “Marea”. Furthermore, the extensive ancient ruins with a subterranean tomb, now located near the modern Huwaryia railway station, which supposedly delimited the southern urban area of ancient “Marea”, are according to the results of the survey of 1990s EAO Delta West Inspectorate (unpublished), the remnants of an isolated structure with a church inside, which may indicate one of the numerous Christian monasteries that existed here in the early Christian period.3 There were no detectable traces of urban features such as street connections between this structure and the group of low quality houses with an associated wine press, that were excavated and published by El-Fakharani (1983: 183-184). This site was earlier identified by Müller-Wiener as a separate settlement (Müller-

In 1972 Fraser, the author of a major work on Ptolemaic Alexandria, wrote, “Although Marea had been an important town in Pharaonic times, these remains are evidently late structures, they are not Ptolemaic and may be Byzantine or early Arab” (Fraser 1972: 146). Some years later an essential question arose about the location of the pilgrimage lake harbour, known from the Coptic Encomium on St. Menas as Philoxenite, a place that is also mentioned in other Christian sources (Drescher 1946: 147-148, 1949: 15-16; Rodziewicz 1983, 2003). The next phase of research connected with the lake harbour, went together with the protective work undertaken by the Egyptian Antiquities Organization in 1982, which required a factual analysis of all preserved archaeological material essential for planning and intensifying the protection of the cultural heritage in this area. I was involved in this process, particularly in the protective works, which

2. El-Fakaharani 1983: 176: “The discovery of burials in the west and south marked the limits of the town westwards and southwards since ancient cemeteries were customarily located outside of inhabited area”; Bagnall & Rathbone 2004: 75: “In the cemetery to the west of the town …”. 3. Compare these structures to those excavated by the Swiss at Kellia, Kasser 1983, 1986 with further references in Actes du Colloque de Geneve, 13 au 15 aout 1984, Geneva 1986.

69

LAKE MAREOTIS: RECONSTRUCTING THE PAST ing a now horizontal stratigraphy, unlike most of the Egyptian ancient settlements which grew upwards and form characteristic hills called “koms”. Contrary to the first and still valid identification of this archaeological site by Müller-Wiener in the 1960s (in particular the water cistern belonging to the big Byzantine House that he published), as a separate construction in a rural environment, the site is still directly or indirectly considered by some archaeologists as a centre of the urban area of an extended town called “Marea” (Bagnall & Rathbone 2004: 74-76).5 Thus, despite the growing archaeological evidence that strongly opposes the identification of the lake port as Pharaonic Marea, we still read about the remains of an ancient town (“Stadtanlage”) located as far south as the modern village of Huwaryia (Grossmann 2003: 15). As a consequence of such interpretations of the ancient village, general conclusions were formulated concerning the location of wineries in urban environments (Bagnall & Rathbone 2004: 74-76; Grossmann 2003: 15, n. 27). Furthermore, another conclusion drawn, was that the so-called (by Prof. El-Fakharani) Byzantine House at Huwaryia is the most convincing archaeological document attesting the existence of the urban centre of ancient Marea, where in reality at this location there existed a large building structured in a rural manner, surrounded by cultivated land. Judging from its huge dimensions and the church built inside, it may have served as a very comfortable and safe pilgrimage hospice from the early 6th to late 7th century AD, i.e. exactly at the time of prosperous pilgrimage activities to Abu Mina (Rodziewicz 2003: 27-47). No older or younger constructions in the area of the building have been registered. The proper understanding of the function of this large building is crucial for the interpretation of the character of the area, which was described by MüllerWiener (1967; who undertook a detailed survey of the area in the 1960s), as a rural settlement, not the centre of any larger town. He mentioned visible surface remains of a brick built water cistern that was located right on the edge of the modern road, that are still visible today (see Fig. 3). In late antiquity, this cistern was situated in a garden or in an open space surrounding the large double-peristyle building (the Byzantine House of El-Fakharani). MüllerWiener thought that the cistern may eventually be connected with the bath (Müller-Wiener 1967). However, the extensive neighbouring building (the Byzantine House), whose excavation was completed in the EAO in the early 1980s, appeared to be the largest residential edifice known hitherto in the whole Mareotis region. It was equipped with a church in the central, eastern wing, between the two extensive colonnaded peristyles. The church had multi-

Fig. 3: Byzantine House/Hospitium at the pilgrimage road from the lake harbour to Abu Mina (drawn by M. Rodziewicz). Wiener 1967: 110) (see Fig. 1). In the southern part of this settlement an underground water cistern, a group of low quality rooms and, further to the west, a medium-sized winery, were partly uncovered by EAO in 1992/3. They were associated with pottery of the late 1st century BC to the middle of the 2nd century AD (author’s identification). This winery appeared to be much older than the winery discovered by Prof. El-Fakharani (1983: Taf. 36). Other architectural remains from this settlement are younger. At the most northern part of the ruins typical Muslim burials were located, which, according to Prof. El-Fakharani, were associated with glazed pottery sherds and a probable Fatimid coin (El-Fakharani 1983: 176-177).4 Further to the north, an extensive building identified by El-Fakharani as the Byzantine House was separately built outside the settlement (El-Fakharani 1983: 184-186). It occupied a surface of over 1,500 m2 and was surrounded on all sides by cultivated land (Rodziewicz 1983, 1988a: 267-277, 1990, 2002: 1-22) (Fig. 3). The house was not covered by structures belonging to other buildings, which means that it was situated in a typically rural environment, at the far edge of the older, but still inhabited village, hav-

5. Bagnall & Rathbone 2004: 74-76; p. 76: “Archaeological evidence for wine production has also come to light. South of the limestone ridge are the remains of two wine-producing establishments. The larger and more interesting is aligned with the middle of the town”. However, such installations as wine factories and pottery kilns are not known in the very centres of ancient towns, where life was regulated by city law. Quite the contrary, these are very typical rural establishments.

4. El-Fakharani 1983: 176-177: “Few pieces of Islamic pottery sherds with green glaze and some white lines, and a piece of Fatimid coin were discovered just above the dead”.

70

M. RODZIEWICZ: MAREA AND PHILOXENITE’S ARCHAEOLOGICAL EVIDENCE

Fig. 4: Byzantine pilgrimage harbour/Philoxenite (Nos. 5-17) and western/coastal rural remains (Nos. 1-4). On the eastern side, structured causeway (No. 18) (drawn by M. Rodziewicz). KEY 1 – Jetty of pre-Byzantine date 2 – Rural estate with large water wheel 2a – Small kiln 3 – Unidentified structure of rural character 4 – Structures with mooring place

Structures identified in the pilgrimage harbour: 5 – Pier 6 – Water wheel (saqia) 7 – Public double bath 8 – Insula with shops 9 – Public building 10 – Public lavatory 11 – Kibotos?

coloured, geometrical opus sectile mosaics on the floor and a richly ornamented northern baptistery (Rodziewicz 1983, 1988a, 1988b). The baptistery floor was made of mortar and painted red in a geometrical pattern composed of triangles. After extensive cleaning of the area, it appeared that beside the large church in the centre and numerous rooms around an extensive courtyard (and probably on the upper floors), a collective (nine seat) lavatory existed there with two smaller ones aside. They were all built in the southern part of the edifice, by the staircase leading to the upper floors, and close to the bathing equipment located in the south-western corner of the house. The bathing facilities were not sufficient to serve the needs of pilgrims eventually stationed there, but a full programme of cold and hot baths was available just about 1 km to the west, in the large public bath that was surveyed, excavated and restored by EAO Delta West Inspectorate in the early 1990s.6 The bath, datable to the Byzantine period, was not attached to

12 – Pier 13 – Transept Basilica (after Grossmann 2002: Fig. 9) 14 – Eastern harbour 15 – Water wheel (saqia) 16 – Public bath 17 – Rock-hewn tomb/hypogeum 18 – Causeway leading to the island 19 – Pilgrimage track to Abu Mina

any architectural remains. It was built in an open area and surrounded by cultivated land. The area of the settlement, which extended from the uplands to the lake, bears traces of various human activities, such as limestone quarrying, various burials on the rocky ridge, clay extraction, pottery firing in the numerous kilns registered there, and cultivation down to the shores of the lake (Röder 1967: 118-131). The individual buildings, whose traces have been excavated and surveyed there, were built in a rural manner. Strictly urban construction methods of building are identifiable only around the lake harbour, erroneously identified as Pharaonic Marea (El-Falaki 1872: 96). Most of the buildings at the port were constructed in one phase, with similar construction principles (Fig. 4). The construction methods are comparable to buildings at Taposiris Magna, Alexandria and also the pilgrimage centre of Abu Mina (Grossmann, et al. 1982; Grossmann, et al. 1984: 123-151). They are not similar to the structures of neighbouring rural settlements and the “karms/karum” located further south. West of the Byzantine passenger harbour Philoxenite (“Marea”), on a rough natural rocky coastline, are the pre-

6. The results of which were presented by Ahmed Abd El-Fatah at the Balnéorient conference, held in Alexandria, December 2006 (in press).

71

LAKE MAREOTIS: RECONSTRUCTING THE PAST served traces of a small size mooring facility and a quite well preserved short but wide and relatively high, jetty that differs remarkably from the two long Byzantine quays of the pilgrimage port (Fig. 4.1). The jetty is constructed from very large sandstone blocks without hydraulic mortar. In front of it a cluster of ruins (Fig. 4. 2) is still visible, which surround a large courtyard, water wheel (saqia), cistern and channels. They probably mark structures of a coastal rural estate similar to the one at Borg el-Arab. Early Roman building material has also been recorded. In addition, there are traces of another water wheel, a small kiln (Fig. 4.2a), and widely dispersed traces of furnace activities. These remains depict the kind of rustic-coastal human activities with small size mooring places (Fig. 4.4), created prior to the construction of a large passenger harbour at the promontory. It is worth noting that during the very last era of the passenger harbour of Philoxenite, with large buildings already abandoned and decaying, old local activities returned, motivated by the rural economy, with an oil press and a pottery kiln installed on top of the ruined constructions.

tures at the lake port (Grossmann 2003: 13-20).7 Arguments against this proposition vary from self-contradictory to those based on out of date, unproven archaeologically theories, particularly that of the existence in that place of a large town, “Marea”, the Pharaonic capital of the district, with a generalised opinion that the old town experienced an unprecedented economic boom in a period of well-documented and steady economic decline of Egyptian cites.8 With a new round of excavations at the lake port starting in 2000, we expected a flow of new information and objective interpretations of freshly unearthed structures and artefacts.9 However, this was not to be and the lack of a rational argument in the presentation of newly excavated material from the harbour during the conference on Medieval Alexandria in 2002, motivated me to publish a paper entitled “Philoxenite-Pilgrimage Harbour of Abu Mina” (Rodziewicz 2003). Three years later, I found a statement in the pamphlet published by the Polish Mission (2006) saying that there are several reasons to assume that Philoxenite was built within the city of Marea, thus supporting the interpretation of the site as that of the location of Philoxenite, which I had first suggested in 1983 (Rodziewicz 1983: 202, 2002). However, the belief that Marea lies under the Byzantine buildings of Philoxenite, despite the lack of any older urban structures there that should date back at least to the time of Psametic I, is still maintained. In the

It is the author’s belief that the lake port in Mareotis should be interpreted as a disembarkation point for Christian pilgrims travelling to Abu Mina, and not as a town with a long tradition matching the history of the ancient capital of a Pharaonic nome (Rodziewicz 1983, 1988a, 1988b, 2002). This opinion was formed on the back of extensive personal study of the whole lake harbour area from the late 1970s to the mid-1990s. The text of the Coptic Encomium and other Christian sources serve to support this view (Drescher 1946: 147). Since, in the early 1980s, the water of Lake Mariout was at a very low level, all jetties and other structures at the harbour were dry, and research, as well as all necessary examination of the construction phases, was much easier than nowadays. For identification of the chronology of the site I used my knowledge of the pottery acquired on other excavations in Egypt and elsewhere, but primarily Alexandria (Rodziewicz 1976). The most informative and numerous sherds were those spotted in the mortar of the lowest parts of the jetties. They belonged to Late Roman Amphora 1, which clearly fixes the chronological horizon of the big harbour in the Byzantine period and parallels the chronology of Abu Mina. The types of the structures, the extension of the urbanised surface, three large water wells (Fig. 4.6, 4.15) and a very small local cemetery hewn in a rocky hill (Fig. 4.17), and a very large church (Fig. 4.13) and extensive harbour, lead me to conclude that the site did not develop gradually, but that is was constructed in a short time and according to a very specific programme. Furthermore, there were no older or younger pottery sherds collected at the site.

7. Grossmann 2003: 13-20, considers my opinion unacceptable, but his arguments are not backed by any factual documentation concerning the chronology of the artefacts in the lake harbour and pilgrimage centre in Abu Mina (which are mostly not yet published). Thus, his opinion on differences in chronology of the bath in the lake port and of the pilgrimage centre at Abu Mina can be taken only as his personal supposition, not supported by the necessary documentation and study of the pottery collected there. 8. According to Bagnall (1993: 108), despite continuous economic decay in late antiquity, not all cities in Egypt suffered regress, and Marea in the immediate hinterland of Alexandria enjoyed a period of prosperity. He says that “the flourishing development at Marea, on the lake which served Alexandria´s back door by water, points to vigorous commerce by this route in the late period”. So, the old Pharaonic city of Marea, of which we still do not have any archaeological evidence and whose location is still unknown, in Bagnall´s publication flourished in the 3rd century AD. 9. The Polish Mission, active there since 2000, is headed by Dr. Hanna Szymanska from the Archaeological Museum in Krakow. With great sadness, we observe subjective methods of interpretation of the archaeological evidence obtained there. At the conference devoted to Medieval Alexandria organized by CEAlex in 2002, Mrs. E. Wipszycka from the Polish Mission spoke about the identification of Philoxenite in the light of new discoveries. Her weak documentation mobilised me to publish the 2003 the article “Philoxenite-Pilgrimage Harbor of Abu Mina”. In 2006, a special exhibition was prepared by the Polish team about the excavations in Marea. It accompanied the conference “Balnéorient” at which Mrs. Szymanska presented a paper on a new bath from the harbour. This exhibition entitled “Marea - Polish Excavations in Egypt 2000-2004 conducted by archaeologists from the Archaeological Museum in Kracow under the auspices of the Polish Centre of Archaeology of Warsaw University”, was supported by a pamphlet with a short text signed K.B. in which we can find the passage: “Marea or Pholoxenite? Identifying the accurate name of the town is a challenge that the Polish Archaeological Mission has undertaken. We have several reasons to assume that it was within Marea that the city of Philoxenite was built.”

Thus, in 1983 at a conference in Athens (Graeco-Arabica) I interpreted the harbour as a specialised port for pilgrims travelling to Abu Mina, as described in the Coptic Encomium, which also mentions its name as Philoxenite (Drescher 1946: 15). The description of the site in the Coptic Encomium matches quite well the preserved Byzantine struc72

M. RODZIEWICZ: MAREA AND PHILOXENITE’S ARCHAEOLOGICAL EVIDENCE Bibliography Abd el-Aziz, M., 1998, Recent Excavations around Abou Mina. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athènes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole française d’Athènes,11-12 décembre 1988: 65-73. Athens, Paris. Abd el-Fatah, A., 1998, Recent Discoveries in Alexandria and the Chora. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athènes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole française d’Athènes,11-12 décembre 1988: 37-53. Athens, Paris. Adriani, A., 1940, Annuaire du Musée gréco-romain 1935-1939. Alexandria. Bagnall, R.S., 1993, Egypt in Late Antiquity. Princeton. Bagnall, R.S., & Rathbone, W., 2004, Egypt from Alexander to the Copts. London. Breccia, E., 1922, Alexandrea ad Aegyptum: Guide de la Ville Ancienne et Moderne et du Musée Gréco-Romain. Alexandria, Bergamo. Décobert, C., 2002, Maréotide médievale. Des Bédouins et des chrétiens. In C. Décobert (ed.), Alexandrie Mediévale 2, ÉtudAlex 2, 127-167. Le Caire. De Cosson, A., 1935, Mareotis: Being a Short Account of the History and Ancient Monuments of the NorthWestern Desert of Egypt and of Lake Mareotis. London. Drescher, J., 1946, Apa Menas. A Selection of Coptic Texts Relating to St Menas. Cairo. Drescher, J., 1949, Topographical Notes for Alexandria and District. Bulletin de la Société Archéologique d’Alexandrie 38: 13-20. Eilmann, R., Langsdorff, A., & Stier, H.E, 1930, Bericht Über die Voruntersuchungen auf den Kurûm El-Tiuwal bei Amrijr. Mitteilungen des Deutschen Archaologi schen Instituts Abteilung Kairo 1: 106-129. El-Ashmawi, F., 1998, Pottery Kiln and Wine-Factory at Burg el-Arab. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athènes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole française d’Athènes,11-12 décembre 1988: 55-64. Athens, Paris. El-Fakharani, F.A., 1983, Recent Excavation at Marea in Egypt. In G. Grimm, H. Heinen & E. Winter (eds.), Aegyptiaca Trevernsia II, Das Römisch-Byzantinische Ägypten. Akten des internationalen Symposions 26.-30. September 1978 in Trier, 175-186. Mainz. El-Falaki, Mahmoud Bey, 1872, Memoire sur l’antique Alexandrie, ses faubourgs, ses environs. Copenhagen. Empereur, J.-Y., & Picon, M., 1998, Les ateliers d´amphores du Lac Mariout. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athènes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole française d’Athènes, 11-12 décembre 1988: 75-91. Athens, Paris.

same pamphlet, the author presents important archaeological evidence, previously published by Grossmann (1993). The pamphlet informs us that: “The most interesting structure in the town is a basilica situated on an eminence overlooking the lake shore near the longest of the harbour piers…The greatest surprise however was concealed under the apse. About 1.80 m below the surviving tops of the apse walls, the floor of the firing chamber of a great kiln for the production of amphorae was discovered. The part of the church where liturgical ceremonies were held had been founded on these manufacturing remains. The kiln, which is one of only a few known from the Delta so far, turned out to be 8 m in diameter and the firing chamber floor was ca. 50 cm thick. Inside it, fired amphorae of the 2nd-3rd century were discovered next to clay stacking rings that had separated the vessels during firing.”

Because no further evidence was presented, I feel obliged to conclude that, firstly, this kiln alone proves that the hill is to a great extent artificial, because it was formed by industrial refuse produced by the ruins of a vast pottery workshop; but that secondly, these ruins should be understood as part of an extensive rural establishment of Early Roman type, comparable to several groups of similar remains hitherto uncovered around the lake (Rodziewicz 2002: 12, Fig. 3; Abd-el Fatah 1998; El-Ashmawi 1998; Abd El-Aziz Negem 1998). One site discovered in the 1980s by the Marsa Matruh Inspectorate of the Egyptian Antiquities Organization, was located on the southern shore of this same Lake Mariout, near modern Borg el-Arab village (El-Ashmawi 1998).10 Thus, the kiln discovered under the church in “Marea” (Fig. 4.13) belongs to a chain of Early Roman rural enterprises located all along the southern shore of the Mareotis Lake and has nothing to do with the great city of Marea. Therefore, we should retain the view that the Byzantine port of Philoxenite, created for pilgrims travelling to Abu Mina, which was built not upon the urban ruins of Marea – ancient capital of Mareotis as described by Herodotus (II.149) – but on the ruins of a nameless, extensive Early Roman rural estate, similar to many others in the region, a number of which have been located along the shores of Lake Mariout.11

10. El-Ashmawi 1998. I had the privilege to work there in 1987/88 as an adviser in the protective works of this large Early Roman rural estate, situated on Lake Mariout. 11. Preliminary reports on excavations in Marea are scarce (see Polish Archaeology in the Mediterranean XII-XVII, 2001-2007, Warsaw University Press). The exhibition illustrating results of the Polish excavations in Marea presented in the Bibliotheca Alexandrina, December 2006, was supplied with texts describing photos. A pamphlet was also presented. Both ignored all previous extensive research and restoration contributions to the site done by individuals and institutions. Reconstruction done at the site in the 1980s (shown there on big panels) that was carried out by the Egyptian Antiquities Organization was not properly described and the suggestion was that the works had been done recently by the Polish Mission. See Szymanska & Babraj 2007, with references to previous reports.

73

LAKE MAREOTIS: RECONSTRUCTING THE PAST Forster, E.M., 1922, Alexandria: A History and a Guide (5th edn.). New York. Fraser, P. M., 1972, Ptolemaic Alexandria, Vol. I. Oxford. Grossmann, P., Kosciuk, J., Severyn, G., & Severyn, H.G., 1984, Vorl. Bericht. Mitteilungen des Deutschen Archaologischen Instituts Abteilung Kairo 40: 123-151. Grossmann, P., Jaritz, H., & Romer, C., 1982, Vorl. Bericht. Mitteilungen des Deutschen Archaologischen Instituts Abteilung Kairo 38: 131-154 Grossmann, P., 1986, Abu Mina. A Guide to the Ancient Pilgrimage Center. Cairo. Grossmann, P., 1993, Die Querschiffbasilika von Hauwariya und die Bauten dieses Typus in Ägypten als Repräsentanten der verlorenen Frühchristlichen Architektur Alexandreias. Bulletin de la Société Archéologique d’Alexandrie 45: 107-121. Grossmann, P., Arnold, F., & Kosciuk, J., 1997, Report on the Excavations at Abu Mina in Spring 1996. Bulletin a la Societe d´Archeologie Copte 36: 83-98. Grossmann, P., 2002, Christliche Architektur in Ägypten. Leiden. Grossmann, P., 2003, Nochmals zu Marea und Philoxenite. Bulletin a la Societe d´Archeologie Copte 42: 13-20. Herodotus, The Histories. A. de Sélincourt (transl.), 1971. Middlesex. Kasser, R., 1983, Survey Archeologique des Kellia (BasseEgypte). Rapport de la Campagne 1981, Vol. I-II. Louvain. Kasser, R., 1986, Le site monastique de Kellia. Sources historiques et explorations archeoloqigues. Geneva. Kaufmann, C.M., 1906, Die Ausgrabung der MenasHeiligtümer in der Mareotiswüste. Cairo. Kaufmann, C.M., 1908, La Découverte des Sanctuares de Ménas dans le Désert de Maréotis. Cairo. Kaufmann, C.M., 1924, Die Heilige Stadt der Wüste. Unsere Entdeckungen, Grabungen und Funde in der altchristrlichen Menasstadt. Kempten. Műller-Wiener, W., 1967, Siedlungsformen in der Mareotis. Archäologischer Anzeiger 82.2: 103-117. Petruso, K., & Gabel, C., 1983, Marea. A Byzantine Port on Egypt´s Northwestern Frontier. Archaeology, Sept/ Oct: 62-63, 76-77. Röder, J., 1967, Antiken Steinbrüche der Mareotis. Archäologischer Anzeiger 82.2: 118-131.

Rodziewicz, M.D., 1976, La céramique romaine tardive d´Alexandria, Alexandrie I. Varsovie. Rodziewicz, M.D., 1983, Alexandria and District of Mareotis. Greco-Arabica 2: 199-216. Rodziewicz, M.D., 1988a, Remarks on the Domestic and Monastic Architecture in Alexandria and Surroundings. In E.C.M. van den Brink (ed.), The Archaeology of the Nile Delta, Proceedings of the Seminar held in Cairo, 19-22 October 1986: 267-276. Amsterdam. Rodziewicz, M.D., 1988b, Remarks on Peristyle House in Alexandria and Mareotis. Praktika 1983: 175-178. Rodziewicz, M.D., 1990, Taenia and Mareotis, Archaeological Research West of Alexandria. Acta of the First International Colloquium of the Egyptian Society of Greek and Roman Studies I: 62-81. Cairo. Rodziewicz, M.D.,1995, Eco-Archaeology of Ancient Alexandria and Mareotis. In A.A. Hussein, M. Miele & S. Riad (eds.), Proceedings of the Seminar on Geosciences and Archaeology in Mediterranean Countries, Cairo, November 28-30/1993: 127-139. Cairo. Rodziewicz, M.D., 1998a, Classification of Wineries from Mareotis. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athènes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole française d’Athènes,11-12 décembre 1988: 27-36. Athens, Paris. Rodziewicz, M.D., 1998b, From Alexandria to the West by Land and by Waterways. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athènes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole française d’Athènes, 11-12 décembre 1988: 93-103. Athens, Paris. Rodziewicz, M.D., 2002, Mareotic Harbours. In C. Décobert (ed.), Alexandrie Médiévale 2, ÉtudAlex 8: 1-22. Rodziewicz, M.D., 2003, Philoxenité – Pilgrimage Harbour of Abu Mina. Bulletin de la Société Archéologique d’Alexandrie 47: 27-47. Sadek, M., 1978, The ancient port of Marea. Cahiers des Études Anciennes VIII: 67. Szymańska, H., & Babraj, K., 2007, Marea. Sixth Season of Excavations. Polish Archaeology in the Mediterranean XVII. Reports 2005: 55-66.

74

K. BABRAJ & H. SZYMANSKA: MAREA OR PHILOXENITE?

Marea or Philoxenite? Polish Excavations in the Mareotic Region 2000-2007 Krzysztof Babraj & Hanna Szymańska

Harbour installations of all kinds can be identified all along the south-western shores of Lake Mareotis. The most imposing ruins of Byzantine date located just 45 km southwest of Alexandria, extend along the coast for 1.5 km and are currently under exploration by a Polish mission (Fig. 1).

gion remained the most important agricultural producer in northwestern Egypt (Haas 2001: 47). The lake was an important communication route in Ptolemaic and Roman times. Goods were transported down the channels from inland, reloaded in Marea and sent to Alexandria, from where they were shipped to other parts of the Mediterranean. The information contained in the Periplus of Pseudo-Scylax (Desanges 1978: 404-405), dated presumably to the early 4th century BC, indicates that Lake Mareotis may have been navigable in Pharaonic times and that the water was potable. Strabo (XVII, 1, 7) reports that the lake port in Marea handled more goods than ever came to Alexandria by the sea route (Yoyotte & Charvet 1997: 83). South of Alexandria, there was a harbour, no longer visible, called Portus Mareoticus; it handled the goods that were shipped via the lake. From there they were transported down a canal named Kibotos through Alexandria to the Mediterranean. One of the canals linking Mareotis with the so-called Canopic Branch of the Nile (Rodziewicz 1983, 1998: 101). The fertile Mareotis region was celebrated in antiquity for its vineyards, olive groves, fruit orchards and papyrus plantations. It produced food for the capital city and rich landowners had their estates here. In the mid-7th century it was still a place for Egyptians to come to appease hunger (History of the Patriarchs 1. 14, 501). Ancient writers spoke warmly of the quality of the local wine exported to Rome (Virgil, Georg. II, 91; Strabo, XVII, 1, 14; Horace, Odes I, 37, 14; Athenaeus, Deipn. I, 33). Even today one of the most renowned Egyptian grapevine growers and wine producers is based on the southern shores of the lake. Almost 30 furnace sites for firing amphorae from the early Hellenistic to late antiquity have been identified on the southern lake shore (Empereur & Picon 1998: 75-91; El-Ashmawi 1990: 55-64). In 2003, the Polish expedition unearthed beneath the basilica at Marea, a pottery kiln for firing amphorae of 2nd-3rd centuries date (see below). This evidence, when considered together with the numerous remains of glass workshops in the region, gives an idea of the crafts that were an important part of the economic life of the region (Kucharczyk 2008: 129-143).

In 2000, excavations at the site were undertaken by a Polish team working under the auspices of the Polish Centre of Mediterranean Archaeology of Warsaw University and the Archaeological Museum in Kraków. The license granted by the Supreme Council of Antiquities (SCA) of Egypt, covers an area of c. 19 ha. A hypsometric grid was put in place and a surface survey of architectural remains carried out (16 objects were found; Szymańska & Babraj 2008: Fig. 1). In the course of the following eight field campaigns, the team has fully cleared a Byzantine bath of the 5th to 7th century connected with a well that supplied it with water, and a funerary chapel. The present focus is on the exploration of a huge Christian basilica standing on the promontory above the port. History of the Site The site was identified as ancient Marea (Μαρεία) by Mahmud Bey El-Falaki, court astronomer to Egypt’s Viceroy Ismail Pasha (El-Falaki 1872: 96; Kees 1930: 16761678; Amélineau 1973: 241-243; Calderini 1980: 233; Gomaà 1980: 647), but today this identification is being reevaluated as indicated in the discussion below. The extensive town ruins situated on Lake Maryut (ancient Mareotis), 45 km from Alexandria, near the village of Hauwariya, have fascinated researchers for a long time. De Cosson prepared one of the first maps of the coast presenting the location of mainly architectural remains (De Cosson 1935: 131). The history of this imposing harbour town with four grand jetties for ships to dock and port basins to protect against wind and waves, can be traced in the ancient sources. Herodotus (II, 30, 2) reported that in the times of Psammetich I of the 26th Dynasty, Marea was a strategic position garrisoned by troops defending the border with Libya. General Amasis defeated the armies of Pharaoh Apries at Marea and took the throne in 570 BC (Diodorus Siculus I, 68). Under Persian rule, the town was allegedly the capital of an independent Libyan-Egyptian kingdom extending from the Canopic Branch of the Nile to Cyrenaica. Its ruler by the name of Inaros, challenged the Persians and lost his life after the fall of Memphis in 454 BC (Thucydides I, 104; Winnicki 2006: 135-142).

The channels that fed the lake silted up in the 8th and 9th centuries due to lack of maintenance, and the fact that fresh water from the Nile, mainly the Canopic Branch, was cut off. The lake began to dry up (Rodziewicz 2002: 9) leading to the decline of towns particularly along its southern shores. In 1801, the lake bed was filled with sea water after the English opened the sluices in order to cut off the Napoleonic army from fresh water. The present surface of the lake is c. 90 km2, and its depth averages at 1.50 m (Blue & Ramses 2005: 7).

Marea never lost its importance as a commercial harbour despite the founding of Alexandria, and the Mareotis re75

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig. 1. Remains of structures along the southern shore of Lake Mareotis. (Drawn by M. Niepokólczycki and A. Błaszczyk). KEY 1. Tholos baths 1a. Stone pier 41 m long 2. Stone pier 111 m long 3. Stone pier 125 m long 4. Harbour facilities 5. Unidentified structure, presumably dry dock uncovered by Petruso

6. Steps belonged to the unidentified structure 7. Byzantine baths (Polish excavations) 8. Well operated with saqiyah (Polish excavations) 9. Double baths uncovered by F. El-Fakharani 10. Decumanus along the shops cleared by F. El-Fakharani

11. Remains of saqiya 12. Granary uncovered by F. El-Fakharani 13. Basilica (Polish excavations) 14. Funerary chapel (Polish excavations) 15. Tombs carved in the rock 16. Remains of the ancient road leading to the island

time of Byzantine domination the region had lost its leading role. The town fell due to lack of water as discussed above, but the decline was also due to political and social factors. Mareotis was overrun by semi-nomadic Bedouin tribes who plundered whatever stood in their way (Décobert 2002). At the beginning of the 8th century, in the rule of the Abbasids which coincides with the period when all life seems to have disappeared from the site, a new wave of Arab nomads appeared in the region. These people were politically active and they increasingly took control of local church institutions (Décobert 2002: 137). Obviously, the atmosphere of the place was no longer favourable to

Mareotis was Christianised eventually but the relations with the Patriarchate in Alexandria passed through different phases (Timm 1988: 1593-1603). The first evidence of Christianity is linked to waves of repression, the effect of which resulted in adherents of the new religion being banished to the region. In 538, by the power of Justinian’s edict, the entire Mareotis region, previously part of the Roman province of Aegyptus Prima, was incorporated into Libya (Edict 13.1, 9, 17-22). In the period immediately prior to the conquest of Egypt by the Arabs in 641, Mareotis was a Byzantine province. The infrequent reference to Mareotis in Coptic writings is proof that by the 76

K. BABRAJ & H. SZYMANSKA: MAREA OR PHILOXENITE? Christians, particularly as a stopover for pilgrims. The wine from Mareotis, however, still retained its reputation through to the 7th century; at least this is what we are led to believe from a story about the abstinence of monks cited by John Moschus (1946).

difficult in view of continued industrialization and numerous tourist villa complexes being built along the shore. Remains of harbour facilities can be seen along both the southern shore and the northern shores of the lake. The biggest of these is Marea. The quay at the waterfront in the eastern end of the site preserves stone blocks that were used in its construction. Beside it there are four almost wholly preserved jetties where the ships docked. These jetties formed the eastern, central and western harbours. They were built of several courses of stone blocks of different size; three or four courses have still been preserved and can be studied thanks to the overall drop of the level of water in the lake. The dimensions of the jetties are (counting from the west to east): I – L. 41 m, W. 6.50 m, blocks averaging: 0.50 m x 1.35 m; II – L. 111 m, W. 5 m; III – The longest jetty is located by the promontory with the ruins of the basilica, L. 125 m, W. 7 m, blocks averaging 0.65 m x 1.70 m; IV – L. 0.35 m (presumably only the part on land has survived), W. 4 m; found next to the island that was connected in antiquity with the mainland by an artificial causeway.

However, identification of early harbour remains at Marea was questioned mainly on the strength of field survey results, which revealed nothing earlier than 6th to early 8th century pottery. The architectural remains also proved to be essentially Byzantine in nature (Fraser 1972: 146). On these grounds Rodziewicz (1983: 202-204) proposed to identify the ruins recorded on the surface with a town established as a transfer point on the southern shores of the lake for pilgrims on their way to the sanctuary of Abu Minas some 20 km away. The Coptic Encomium of St. Menas written by the Patriarch Ioannes IV (775-789), mentions numerous conveniences “like hospices by the lake and rest-houses for those wishing to rest” (Drescher 1946: 147148), a market place, porticoes and even a facility that we would refer today as a left-luggage office. This impressive rest stop was said to have been founded by Philoxenos, Prefect of the Emperor Anastasius (491-518), hence the town’s name of Philoxenite. Wipszycka (2002) has identified this official as the consul Philoxenos Soterichos, who held office in 525 and who is known also from consular diptychs (Martindale 1980: 879-880).

Marea or Philoxenite? The size of the harbour installations, which are currently being researched by a Greek-Egyptian team of underwater archaeologists (headed by Nikos Lianos from the Hellenic Society for the Study and Preservation of Marine Cultural Heritage and the Department of Underwater Archaeology in Egypt) when considered in connection with the considerable size of the basilica on the lakefront and the innumerous but conclusive evidence of Early Roman potsherds, indicates that a large urban agglomeration flourished on the spot already in pre-Byzantine times. Neither can the extensive sewage network be considered as belonging to a town of little consequence. Philoxenos may have founded his city adjacent to existing harbour installations that had formerly been part of Marea, before it was abandoned as an urban centre prior to the silting up of the fresh-water channels that supplied the lake, but there is little doubt that settlement existed here prior to its foundation. Already in the 2nd century AD Athenaeus (33d) wrote of Marea as the source of a wine of fine quality; he called Marea a village (κώμη). Athanasios (85 col. 400b-c) had the same impression of the town 200 years later. Grossmann (2003: 16) thinks that the Hellenistic architecture of Marea was of mud-brick which is easily disintegrated. It should be noted that at the margins of the settlement mud-brick is a building material used in the lower layers of the basilica, chronologically contemporaneous with the pottery kiln of the 2nd-3rd centuries AD.

Location of the Site The geographical coordinates of the site are: 30o 59’ 32.85’’ N and 29o 38’ 58.34’’ E to 30o59’48.84’’ N and 29o39’28.32’’ E. The excavation area measures 19.7 ha. To the west it is limited by a double-lane road linking the waterfront with Shakush settlement; the lake is a natural boundary on the north and east, and to the south there is the desert which extends as far as Shakush settlement. The overall height difference does not exceed 8 m anywhere on the site. The site grid consists of 42 geodetic points, mostly dictated by the natural topography. The coordinates were assumed locally, while the heights are listed by the Egyptian state geodetic service. The topographical plan of the site so far only covers the area excavated in 2000-2003 (see Fig. 1). From a geological point of view, Marea lies in a broad stretch of land between Mallahet Maryût, the long westward arm of Lake Mareotis, and a parallel ridge of oolithic limestone. The soils on either side of the ridge are mostly calcarenite, very favorable for cultivation with proper watering (Warne & Stanley 1993; El-Fakharani 1983: 175; Rodziewicz 1995; Mycielska-Dowgiałło & Woronko 2008: 17-18). A similar ridge lies on the opposite side of the lake, separating Lake Mareotis from the sea. This is the Taenia (ταινία) Ridge which extends along the coast from Alexandria all the way to Abu Mina and Libya. Today these ridges are mostly destroyed as a result of heavy exploitation of limestone quarries. In the 4th and 5th century, Christian monasteries appeared in the Taenia; they were called after the mile-stones that served to measure the distance from Alexandria. Their localisation today is very

There is however, opposition to the idea that this was the site of Philoxenite. The distance from Philoxenite to Abu Mina is one area to question. According to Drescher’s (1946) translation of the Encomium, the distance amounted to about 30-35 km. The text also mentions rest stops 77

LAKE MAREOTIS: RECONSTRUCTING THE PAST where water was available, which the Prefect had arranged every 10 miles or so (=14.80 km) between the two towns. However, the distance in reality is some 20 km, not enough for even two water stops. Yet the key argument against the identification with Philoxenite is Grossmann’s research on the frequency of pilgrimages to Abu Mina. Pilgrimages to the sanctuary did not peak until the second half of the 5th century, and afterwards surges in the number of pilgrimages accompanied the feast days of St Menas (Grossmann 2003: 18). The ruins currently being investigated on the site are considered earlier; the double baths explored by ElFakharani (1983; see both Haggag’s papers in this volume) originate from the end of the 4th century and the pottery kiln under the apse of the basilica operated in the 2nd-3rd centuries AD. The size of this latter facility is unmistakable proof of a flourishing centre existing on the spot (possibly Marea) prior to the Byzantine foundations, perhaps producing the wine that Mareotis was famous for and expediting it to other towns of the Greco-Roman world in amphorae produced at the site. One of the sherds from the fill of the pottery kiln bore the stamp of a potential owner of the vineyard Διονυσίου (Szymańska & Babraj 2005: 130, n. 41). Many artifacts discovered in the neighborhood of Marea confirm the presence of pottery kilns for firing amphorae in Roman times (Empereur & Picon 1998).

block-joining technique that went out of use in Italy in the 1st century AD, evidently point to an earlier origin (Adam 2008: 57). Moreover, the apsed funerary chapel, discovered by the Polish mission, with its three chambers intended for an affluent family, could not have been erected by and for pilgrims transiting through the town. On the other hand, there can be no doubt that pilgrimages to the temple of the holy martyr must have passed through this area, if anything because of the close proximity to the lake. More evidence is provided by a Late Roman villa rustica, uncovered in Hauwariya village, transformed into a dormitory in the middle of the 6th century and furnished with a small church inside it (Rodziewicz 1988: 271-273, Fig. 2). A large urban centre with developed harbour facilities must have been perfectly suited as a stopover for wayworn and surely often ailing pilgrims. Décobert (2002: 129) identifies ancient Marea with the locality of Îkinjî Maryût to the east of the site, where excavations are currently being conducted. Yet the only arguments in favour of this theory is the strategic position of this locality, affording control of the road from Cyrenaica and Egypt to Alexandria and being accessible also from the opposite side of the lake thanks to a considerable narrowing of its width at this point.

The great jetties are of key importance for the dating of the port. The structural variety of jetty construction (different size blocks) suggests several building stages, but the characteristic imprints left by ‘swallow-tail’ metal clamps, a

Fig. 2. Marea. Byzantine baths. (Drawn by D. Tarara).

78

K. BABRAJ & H. SZYMANSKA: MAREA OR PHILOXENITE? Baths (Szymańska & Babraj 2008a: 27-83) The freestanding building follows an orthogonal plan, encompassing the baths itself (Fig. 2) and the well, which is furnished with a sāqiyah installation drawing water for the baths. The actual building of the baths, covered an area of 642 m2, including two courtyards and was encircled by a wall of stone blocks. The baths itself was built entirely of red brick on a brick foundation (Fig. 3). Shops and a latrine lined the north wall on the outside. Colonnaded courtyards were located to the east and west.

The underground part of the building included the service area cellars, two heating furnaces and one for heating water in a boiler, finally, there were four hypocaust cellars: two in the men’s part and two in the women’s part. The interior decoration of the building consisted of marble pavements, columns with Corinthian capitals, and a few coatings of painted wall plaster featuring a colorful vegetal frieze. The floors and selected walls of pools were faced with marble slabs. This was evidently a luxurious complex and definitely not the only one in town to judge by the solid water channel system to which it was connected.

The building was a double bath, consisting of two uneven parts separated by a brick wall: the southern one, for men, bigger than the northern one for women. Each of these parts was heated by a separate furnace supplying warm air to four hypocaust systems. The entrance for men was in the western facade; it led from the courtyard, following widely accepted circulation principles, that is, directly from the apodyterium into the tepidarium, and then into the two caldaria. The women’s part was not as extensive, accessible from a rather spacious courtyard to the east side. It consisted of three rooms: apodyterium and two caldaria. Fourteen small pools were used for bathing, eight of which were located inside the two parts of the baths, by the outer wall of the caldaria. Two of them were accessible from the courtyard, leading to the men’s part. Four pools were located outside the baths, symmetrically in the two courtyards, abutting the walls. The pools were differentiated in shape, from semicircular through rectangular to round, depending on the rebuilding phase. Calculations made on the basis of six fragments of brick arches found in the debris permit a reconstruction of the width of baths halls, which were presumably covered with barrel vaults. The hypothetical height of the building, derived from the proportions of the surviving walls, was 3.50 m.

Nowhere in the building was a clear stratigraphy of the construction and occupation of the baths visible and the fact that the building had collapsed (vide brick arches from the vaults lay immediately on the floors) precludes precise dating. An analysis of the pottery finds has provided a timeframe for the operation of the complex between the first half of the 6th century and the first 30 years of the 8th century, that is, before the appearance of Islamic Early Lead Glazed wares, of which not unsurprisingly, not one sherd has been found at the site (Majcherek 2008). Arab coins already issued after the monetary reform, that is, after 686, and remaining in circulation for 750 years, confirm the continued operation of the baths after the damages incurred during the Persian occupation and the Arab invasion (Malarczyk 2008). The most intensive use of the baths occurred in the first half of the 7th century, a fact confirmed by both the pottery finds and the prevalence of coins issued by Chosroes II (Lichocka 2008). Sāqiyah (Szymańska & Babraj 2008: 85-99) A well operated by a sāqiyah stood about 5 m north of the baths. The well itself was 5 m deep, rectangular in plan and built of stone blocks (Figs. 3 & 4). The water in the

Fig. 3. Marea. Baths and western courtyard. View from the south-west (photo P. Suszek).

79

LAKE MAREOTIS: RECONSTRUCTING THE PAST Fig. 4. Marea. Well operated with sāqiyah. View from the south (photo by T. Kalarus).

well came from a still functioning underground spring in its northwestern corner. The mineral content of this water is much more abundant compared to the lake water. South of the well and level with its rim, there was the floor of a reservoir made of fired bricks in waterproof mortar. The south side of this basin, which was filled with water from the well, was reinforced with a solid platform built of stone blocks. A capacious cistern closed off the complex of the sāqiyah to the north. It was presumably intended as a facil-

ity for collecting water in case of drought. On top of this installation was a treading circle for the animals turning the sāqiyah wheel. A very simple hydraulic system based on connected vessels was employed. Water was collected in the reservoir of the sāqiyah and carried to the baths via a system of terracotta pipes Only one conduit reaching Basin X in the baths south of the sāqiyah (see Fig. 3) has been preserved by the west wall Fig. 5. Marea. Funerary chapel. View from the west (photo P. Suszek).

80

K. BABRAJ & H. SZYMANSKA: MAREA OR PHILOXENITE? Fig. 6. Marea. Plan of the basilica (drawn by D. Tarara).

of the baths (see plan in Szymańska & Babraj 2008: Fig. 1). The pipe, a fragment of which is visible under the last block of the sāqiyah reservoir with two blocks of the limestone bedding, ran to the north wall of the baths. The pipe turned into Room B1, which it crossed until it reached the south wall, where it was raised c. 1.50 m in order to empty into Basin X. There is every reason to believe that there were more sāqiyahs operating in the town complex of Marea. A ground survey of the site has identified the remains of at least one other facility of the kind in the area to the west of the baths (Szymańska & Babraj 2008: 88).

Basilica (Szymańska & Babraj 2006: 107-117) The most interesting building at the site, however, is a basilica situated on the hill near the longest harbour jetty. It was discovered by Müller-Wiener (1967: 106, n. 16), former director of excavations at the nearby sanctuary of Abu Mina. But it was Grossmann who first determined the plan and dimensions of the building in 1986 (Grossmann 1993: 107-121). Surprisingly, none of the ancient written sources mention this church, which was one of largest buildings of its type yet known in Egypt (see Grossmann 2002, passim).

Funerary chapel (Szymańska & Babraj 2008: 177-185). One hundred meters to the south there were the ruins of a building which were identified as a funerary chapel. It was furnished with an east-oriented apse and three masonry grave chambers containing 23 burials associated with very poor grave goods (Fig. 5). They contained both genders of different age, all clearly of family character. It functioned for a little over a hundred years and consequently must have belonged to the town’s inhabitants. A Gaza amphora under the floor of the apse verified a 6th century date for the construction of this complex.

It was a squat-shaped building, divided by columns into three aisles, furnished with a wide transept terminating in rounded arms and a relatively tiny apse (Szymańska & Babraj 2004: 53-56, 61-64, 2005, 63-67, 2005a, 43-54). The liturgical rooms were not on the outside, as was the rule at other basilicas, but incorporated into the body of the building on either side of the apse (Fig. 6) (see Grossmann 2002, passim). The baptistery with round baptismal font apparently remained from an earlier building, possibly a small chapel, which preceded the great basilica at the site. 81

LAKE MAREOTIS: RECONSTRUCTING THE PAST Fig. 7. Marea. Apse of the basilica. View from the west. ‘A’ and ‘b’ locate the two burials (photo by T. Kalarus).

A kind of money exchange building intended presumably for pilgrims visiting the basilica, was discovered beyond the southwestern corner of the building. Rare examples of bronze weights used to verify coin weight were found inside this room (Szymańska & Babraj 2005a: 54).

rated columns of various sizes, all of them in Proconesian marble and probably imported via Alexandria. The interior decoration of the basilica also included pavement mosaics of which small marble cubes found in the building are the only surviving evidence.

Two burial chambers with multiple burials were discovered under the floor of the apse (Fig. 7, a & b). Anthropological examination identified over 100 individuals: men, women, children and even unborn children. They appear to have been buried there during the invasion of Chosroes II in AD 619 when Persian troops torched Alexandria and ravaged the region.

Amphora Kiln (Szymańska & Babraj 2004a: 53-56, 6164, 2004b: 26-28, 2005: 63-67). A complete surprise awaited archaeologists digging under the church apse. At a depth of 1.80 m below the preserved tops of the walls, a grate belonging to a large amphora kiln was subsequently discovered. The part of the church intended for liturgical practices used this earlier kiln as a foundation (Fig. 8). More of the grate appeared once the floor of the burial chambers under the apse had been cleaned.

The rich interior decoration of the basilica included several fragmentary column shafts and Corinthian capitals deco-

Fig. 8. Marea. Basilica. Amphora kiln. View from the west (photo by J. Kucy).

82

K. BABRAJ & H. SZYMANSKA: MAREA OR PHILOXENITE? Bibliography

The kiln had a diameter of 8 m, and the thickness of the grate was 0.50 m. One of only a few discovered in the Nile Delta, this kiln still contained upon discovery the last batch of amphorae. These vessels were dated to the 2nd and 3rd centuries AD. The wall of the kiln still stood 0.93 m high and was built of mud-bricks. The structure of the kiln is not unlike others from the period with pillars under the grate pierced with holes to support easy circulation of hot air (El-Ashmawi 1998: 55-64).

Ancient Sources Athenaeus, The Deipnosophists, Ch.B. Gulick (transl.), 1927. Cambridge, MA. Athanasios, Apologia contra Arianos, H.G. Opitz (transl.), 1934. Berlin, Leipzig. Diodorus Siculus I, Library of History, C.H. Oldfather (transl.), 1989. Cambridge, MA. Herodotus, Book II, A.B. Lloyd (transl.), 1976. Leiden. History of the Patriarchs of the Coptic Church of Alexandria, Patrologia Orientalis 1.14, B.T.A. Evetts (ed.), 1904. Paris. Horace, Odes and Epodes, N. Rudd (transl.), 2004. Cambridge, MA. John Moschus, Pratum Spirituale = N.H. Baynes, 1947, The Pratum Spirituale. Orientalia Cristiana Periodoca 13: 404-414. Justinian, Edict, Corpus Iuris Civilis. Codex Justinianus, Vol. II, P. Krüger (ed.), 1877. Berlin. Strabo, Geography. In Strabon. Le voyage en Egypte. Un regard Romaine, J. Yoyotte & P. Charvet (transl.), 1997. Paris. Thucydides, History of the Peloponnesian War, Ch. Forster Smith (transl.), 1928. Cambridge, MA. Virgil, Eclogues. Georgics, H. Rushton Fairclough (transl.), 1999. Cambridge, MA.

Conclusions The site of Marea with its Late Antique agglomeration remains a puzzle, generating considerable controversy as to its chronology and even its very name. The identification of the town ruins lying on the southern shore of Lake Mareotis, just 45 km south-west of Alexandria, is amongst the priorities of a team from the Polish Centre of Mediterranean Archaeology of Warsaw University and the Archaeological Museum in Kraków, which have conducted excavations here since 2000, based on a concession granted by the Supreme Council of Antiquities of Egypt that covers an area of 19.7 ha. The Polish expedition has concentrated on the extant ruins, investigating the role and importance of this Byzantine town as a religious center built around a huge basilica. During nine field seasons (2000-2008), three independent architectural complexes have been investigated: the baths with sāqiyah-well, a funerary chapel, and the Christian basilica. All the structures were dated to the 6th (end of 5th) through early 8th centuries. A site survey included an inventory of all the ruins discernible on ground level within the limits of the SCA concession (see Fig. 1).

Secondary Sources Adam, J.P., 2008, La construction romaine. Paris. Amélineau, E., (1893, reprint 1973), Géographie de l’Égypte à l’époque copte. Osnabrück. El-Ashmawi, F., 1998, Pottery Kiln and Wine-Factory at Burg el Arab. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athènes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole française d’Athènes,11-12 décembre 1988: 55-64. Athens, Paris. Babraj, K., & Szymańska, H., 2004, Die Vierte Grabungssaison in Marea, Aegypten. Grabkapelle und Basilike. Kemet 13.3: 61-64. Babraj, K., & Szymańska, H., 2005, Fünfte Grabungssaison in Marea, Ägypten: Basilika. Kemet 14.3: 63-67. Babraj, K., & Szymańska, H., 2008, Sāqiyah. In H. Szymańska & K. Babraj (eds.), Byzantine Marea. Excavations in 2000-2003 and 2006, Marea, Vol. 1: 85-99. Kraków. Babraj, K., & Szymańska, H., 2008, Funerary chapel. In H. Szymańska & K. Babraj (eds.), Byzantine Marea. Excavations in 2000-2003 and 2006, Marea, Vol. 1: 177-185. Kraków. Blue, L., & Ramses, S., 2005, Lake Mareotis Research Project – Preliminary Report from Pilot Survey August 2004. Newsletter of the Archaeological Society of Alexandria 15 (February): 5-16. Calderini, A., 1980, Dizionario dei nomi geografici e topografici dell’Egitto Greco-Romano, Vol. III, fasc. 3. Milano.

This large port operated in Roman and later Byzantine times and perhaps also in the Ptolemaic era. However, the question of the actual identity of the site remains still uncertain. The current excavated ruins of Byzantine date extend along the coast for 1.5 km. The great jetties are of key importance for the dating of the port. The size of the harbour installations, the small but important quantity of Early Roman potsherds, as well as huge sewage network of substantial build, indicates that a large urban agglomeration flourished on the spot from pre-Byzantine times. Philoxenos may have founded his city prior to its abandonment as an urban center due to the silting up of the sweet-water Nile channels supplying the lake. However, evidence suggests that a site existed prior to the establishment of Philoxenite, being located next to existing harbour installations that may have been part of Marea. The recent discovery of the pottery kiln under the apse of the basilica that operated in the 2nd-3rd centuries AD, is unmistakable proof of a flourishing center existing on the spot (possibly Marea) prior to the Byzantine foundations. Perhaps this site was were the wine that Mareotis was famous for was being produced, and from here the product was being exported to other towns of the Greco-Roman world in amphorae produced at the site. 83

LAKE MAREOTIS: RECONSTRUCTING THE PAST De Cosson, A., 1935, Mareotis: Being an Account of the History and Ancient Monuments of the North-West Desert of Egypt and Lake Mareotis. London Décobert, Ch., 2002, Maréotide médiévale. Des Bédouins et des Chrétiens. In Ch. Décobert (ed.), Alexandrie Médiévale 2, Études Alexandrines 8: 127-167. Le Caire. Desanges, J., 1978, Périple du Pseudo-Scylax 90 F 107 M = Périples libiques, 7. In J. Desanges (ed.), Recherches sur l’activité des Méditerranéens aus confins de l’Afrique, Collection de l’École Française de Rome 38: 404-405. Rome. Drescher, J. (ed.), 1946, Apa Mena. A Selection of Coptic Texts relating to St. Menas. Cairo. Empereur, J.-Y., & Picon, M., 1998, Les ateliers d’amphores du lac Mariout. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athènes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole française d’Athènes, 11-12 décembre 1988: 75-91. Athens, Paris. El-Falaki, Mahmoud Bey, 1872, Mémoire sur l’antique Alexandrie, ses faubourgs et et environs découverts par les fouilles, sondages, nivellement et autres recherches. Copenhagen. El-Fakharani, F.A., 1983, Recent Excavation at Marea in Egypt. In G. Grimm, H. Heinen & E. Winter (eds.), Aegyptiaca Trevernsia II, Das Römisch-Byzantinische Ägypten. Akten des internationalen Symposions 26.-30. September 1978 in Trier, 175-186. Mainz. Fraser, P.M., 1972, Ptolemaic Alexandria, I. Oxford. Gomaà, F., 1980, s.v. Kom el-Idris (Marea), Lexicon der Äegyptologie III: 647. Grossmann, P., 1993, Die Querschiffbasilika von Hauwarîya-Marea und die übrigen Bauten dieses Typus’ in Ägypten als Repräsentanten der verlorenen frühchristlichen Architektur Alexandreias. In N. Swelim (ed.), In Memoriam Daoud Abdu Daoud (Bulletin de la Societé Archéologique d’Alexandrie 45): 107-121. Grossmann, P., 2002, Christliche Architektur in Aegypten. Leiden. Grossmann, P., 2003, Nochmals zu Marea und Philoxenite. Bulletin de la Societé d’Archéologie Copte 42: 13-29. Haas, Ch., 2001, Alexandria and the Mareotis Region. In S. Burns & J.W. Eadie (eds.), Urban Centers and Rural Context in Late Antiquity: 47-62. Ann Arbor, MI. Kees, H., 1930, s.v. Marea, Mareotis. Real Encyclopedie XIV.2: 1676-1678. Kucharczyk, R., 2008, Glass finds from the baths and sāqiya. In H. Szymańska & K. Babraj (eds.), Byzantine Marea. Excavations in 2000-2003 and 2006, Marea, Vol. 1: 129-143. Kraków. Lichocka, B., 2008, The early Byzantine coins. In H. Szymańska & K. Babraj (eds.), Byzantine Marea. Excavations in 2000-2003 and 2006, Marea, Vol. 1: 145-152. Kraków.

Majcherek, G., 2008, The pottery assemblage from the baths and sāqiyah. In H. Szymańska & K. Babraj (eds.), Byzantine Marea. Excavations in 2000-2003 and 2006, Marea, Vol. 1: 105-127. Kraków. Malarczyk, D., 2008, Ummayad coins. In H. Szymańska & K. Babraj (eds.), Byzantine Marea. Excavations in 2000-2003 and 2006, Marea, Vol. 1: 153-154. Kraków. Martindale, J.R. (ed.), 1980, The Prosopography of the Later Roman Empire, II. Cambridge. Müller-Wiener, W., 1967, Siedlungsformen in der Mareotis. Archäologischer Anzeiger 82.2: 103-117. Mycielska-Dowgiałło, E., & Woronko, B., 2008, Evolution of the natural environment in the region of Marea. In H. Szymańska & K. Babraj (eds.), Byzantine Marea.Excavations in 2000-2003 and 2006, Marea, Vol. 1: 17-26. Kraków. Rodziewicz, M.D., 1983, Alexandria and District of Mareotis. Graeco-Arabica II: 199-216. Rodziewicz, M.D., 1988, Remarks on the Domestic and Monastic Architecture in Alexandria and Surroundings. In E.C.M. van den Brink (ed.), The Archaeology of the Nile Delta, Proceedings of the Seminar held in Cairo, 19-22 October 1986: 267-276. Amsterdam. Rodziewicz, M.D., 1995, Eco-Archaeology of Ancient Alexandria and Mareotis. In A.A. Hussein, M. Miele & S. Riad (eds.), Proceedings of the Seminar on Geosciences and Archaeology in the Mediterranean Countries, Cairo, November 28-30/1993: 127-139. Cairo. Rodziewicz, M.D., 1998, From Alexandria to the West. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athènes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole française d’Athènes,11-12 décembre 1988: 93-103. Athens, Paris. Rodziewicz, M.D., 2002, Mareotis Harbours. In Ch. Décobert (ed.), Alexandrie Médievale 2, Études Alexandrines 8: 1-22. Le Caire. Rodziewicz, M.D., 2003, Philoxenité – Pilgrimage Harbor of Abu Mina. Bulletin de la Societé Archéologique d’Alexandrie 47: 27-47. Szymańska, H., & Babraj, K., 2004a, Marea. Fourth season of excavations, 2003. Polish Archaeology in Mediterranean XV: 53-63. Szymańska, H., & Babraj, K., 2004b, The ancient port of Marea, Egypt. Four seasons of excavations by the Polish Archaeological Mission. Minerva (April 2004): 26-28. Szymańska, H., & Babraj, K., 2005, Fouilles archéologiques de Marea en Égypte. Saisons 2002-2003. Archeologia LV: 119-130. Szymańska, H., & Babraj, K., 2005a, Marea. Fifth Season of Excavations, 2004. Polish Archaeology in Mediterranean XVI: 43-54.

84

K. BABRAJ & H. SZYMANSKA: MAREA OR PHILOXENITE? Warne, A.G., & Stanley, D.J., 1993, Late Quaternary Evolution of the Northwest Nile Delta and Adjacent Coast in the Alexandria Region, Egypt. Journal of Coastal Research 9.1: 26-64. Winnicki, J.K., 2006, Der libysche Stamm der Bakaler in pharaonischen, persischen und ptolemäischen Ägypten. Ancient Society 36: 135-142. Wipszycka, E., 2008, Remarques sur l’identification de Piloxenité à la lumière de fouilles récentes. Paper delivered at the Third Conference on Medieval Alexandria at the Centre culturel français in Alexandria, 8-10 November, 2002 (unpub.).

Szymańska, H., & Babraj, K., 2006, Polish Excavations in the Basilica at Marea (Egypt). Bulletin de la Societé Archéologique d’Alexandrie 45: 107-117. Szymańska, H., & Babraj, K., 2008, Baths. In H. Szymańska & K. Babraj (eds.), Byzantine Marea. Excavations in 2000-2003 and 2006, Marea, Vol. 1: 27-83. Kraków. Szymańska, H., & Babraj, K. (eds.), 2008, Byzantine Marea. Excavations in 2000-2003 & 2006, Marea, I. Kraków. Timm, S., 1984, Das christlich-koptische Ägypten in arabischer Zeit. Wiesbaden. Timm, S., 1988, Teil 4 (M - P), Reihe B (Geisteswissenschaften) No. 41/4, Wiesbaden.

85

LAKE MAREOTIS: RECONSTRUCTING THE PAST

86

M.-F. BOUSSAC & M. EL AMOURI: LAKE STRUCTURES AT TAPOSIRIS

The Lake Structures at Taposiris Marie-Françoise Boussac & Mourad El Amouri

Introduction Among the harbours of Mareotis, which in recent years have been the subject of renewed attention, Taposiris on the north shore is of particular interest. Amongst many features of interest, the city offers the only example, besides that of Alexandria, of a closed basin which allows controlled maritime traffic management (Fig. 1). These structures were briefly mentioned at the beginning of the 19th century by Pacho, who in 1824 devoted a few lines to “the dam running from east to west (...) built in the south of the city”, a structure intended, he said, to prevent floods (Pacho 1827) and sketched by Coste1 in 1820 (Fig. 2). Thereafter, Breccia (1914), De Cosson (1935: 110-111) (Fig. 3), Ochsenschlager (1979, 1999) and Rodziewicz (2002), amongst others, have pointed out the peculiarities of the closed basin, speculating on its chronology and its connection with the wall of the Barbarians which closes, to the west of the city, the very narrow spit of land that separates the Mediterranean from Lake Mareotis at this point, and its relation to the causeway which blocked the lake to the south. At some point in their use, both systems acted as locks and seem related, but we do not know if they are part of the same feature, nor do we know during which period they were used.

When the French Archaeological Mission of Taposiris3 was launched in 1998, it was determined that an understanding of the harbour system would be a research priority. What was the chronology of the currently visible structures? To which phases of the site did they correspond? Because of the dimensions of the harbour basin (the eastwest artificial levee or causeway is about 1,700 m long), areas that were deemed likely to provide some answers with limited resources were selected for excavation, supplemented with an environmental study and geophysical survey. Since most of the data acquired has already been the subject of articles (Boussac 2007, 2009), this paper will briefly present the harbour structures and the conclusions reached concerning the southern causeway. The east side of the system will be the main focus as this gives the latest possible date for all the studied lake constructions. Overview of the Remains The ancient city of Taposiris, located on the south side of the taenia ridge, is organised into three sectors linked by a network of north-south routes: the upper town with the Brescia terrace and the temple, the middle town, and the lower town on the shores of Lake Mareotis. To the west, the Wall of the Barbarians forms the western limit of the city (see Fig. 1).

To answer these questions, the American mission from Brooklyn College carried out several soundings in 1975, during a one month campaign. Anxious to locate the harbour area,2 Ochsenschlager worked on two sectors, north and north-west of the dug-out channel and the causeway. Neither gave the anticipated results: in the north (Zone A), an elevated area, interpreted as a lake front, proved to be an accumulation of rubble (waste from amphorae workshops, as shown thereafter by Empereur & Picon 1998). In the north-west (Zone C), American archaeologists concentrated their efforts on what initially seemed to be a warehouse along the shore. They described it as a platform which had been redesigned several times (the so-called platform building), they did not understand its function, but they dated it to the 3rd century AD. They also uncovered a sophisticated system of water tanks carrying water towards the harbour from a terrace further north. No research was published to specify the chronology, and no interpretation was suggested (unpublished).

The topographical survey of the lower town area carried out in 1999 and 2000 (Fig. 4), and supplemented in 2009, extends from the Wall of the Barbarians in the west to Plinthine in the east, and highlights the irregular outline of the lake shore. A partially anthropogenic spit of land (promontory) divides the lake shore into two marshy plains, one of which, to the east, forms a bay which is interpreted as a harbour basin.4 This basin covers an area of approximately 8 ha and is closed off to the south by a causeway (approximately 1,700 m long), which extends east-west in the shape of a ridge, but it is not straight as the sketches of Coste (1820; see Fig. 2) or De Cosson (1935: 110; see Fig. 3) imply. It is also interrupted by at least six openings that may be deliberate or may be related to the disappearance of the retaining walls which in places supported the north slope of the causeway.

3. Since 1998 La Mission Française in Taposiris Magna has been supported by the Commission des fouilles of the French Ministry of Foreign and European Affairs. Authorisation has been kindly given by SCA and wee appreciate their assistance. 4. Analysis by C. Flaux, PhD Univ. Aix-Marseille, in 2008, in a PhD thesis on the geomorphology of Lake Mareotis.

1. Thanks to the Bibliothèque Municipale à Vocation Régionale in Marseille, especially Noëlle Colombié, who allowed us to reproduce two sketches by Pascal Coste from their archives. 2. Thanks to the American mission, especially M. Venit, who shared their unpublished documentation (esp. Preliminary Report of the 1975 season).

87

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig. 1. General map of the site and location of areas under study at Taposiris Magna and Plinthine. © MAFT, French Archaeological Mission in Taposiris.

88

M.-F. BOUSSAC & M. EL AMOURI: LAKE STRUCTURES AT TAPOSIRIS

Fig. 2. Drawing by Pascal Coste, made in 1820. 2a: General plan of Alexandria to Abousir. 2b: Detail of plan showing the Temple of Taposiris (1), the Tower of Arabs (2), bridge (3), eastern jetty (4) and causeway across the lake from the levee to the south shore (5). 2c: Plan and isometric view (south) of the bridge of Taposiris. © BMVR Fig. 3. Map of the remains of the ancient city of Taposiris by A. De Cosson, 1935.

89

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig. 4. Location of the different sectors of the site of Taposiris that were studied along the lake shore. © MAFT

The Elements of the Lake System The Causeway and the Channel Several operations (1998, 2000-2005) were conducted at the west end of the causeway, on both sides of the channel and near the bridge, where the basin forms a bend: it

expands to the nortwest to form a loop against which is aligned by a series of buildings, whose different orientations and especially superposition indicate different phases. This visible state has significantly altered the previous interpretation of the phases: the channel was dug in the first half of the 2nd century AD (Boussac 2009)7 in a location where a densely occupied area was previously developed whose architectural design suggests prosperity. This area was used for storage and trade, judging from the quantity of imported amphorae material found there (Boussac 2009).8 On the north shore (Sector 1) there is a row of shops, abandoned at the end of the Hellenistic period (un-recovered items left in place) perhaps as a result of flooding,9 before the digging of the channel. On the other side of the channel (Sector 3), earth that had been dredged from the lake covered buildings last occupied at the end of the 1st century BC and which had been in use since at least the 2nd century BC. The excavated building has the same

5. Breccia 1914. He only mentions the 1 km-long dyke, running parallel to the hills, and a “fine bridge whose construction certainly dates to Roman times. It seems clear that the lake extended to Taposiris, and that the dyke blocked the water in a kind of harbour”. 6. Prospecting by C. Benech, CNRS. See the map in Boussac 2007: 452.

7. See the analysis by S. Marquié in Boussac 2009. 8. Boussac 2009 (see the appendix by S. Marquié and the study by K. Senol). 9. Study suggests during the 1st century BC: see the analysis by S. Marquié in Boussac 2009: 137.

Topographic mapping and geophysical survey revealed the density of structures in the lower town, briefly mentioned as “walls and ruins” by Coste (1820) and ignored by Breccia (1914).5 Although the causeway that blocked the lake has virtually disappeared (only a stretch of about 10 m could be uncovered), its outlet was located near a warehouse built on the causeway. Most importantly, the geophysical survey revealed in the West Bay and south of the bridge many ruins which are currently silted over.6 These results underscore the extent of changes in the landscape, the mobility of the shoreline and the extent of the harbour area.

90

M.-F. BOUSSAC & M. EL AMOURI: LAKE STRUCTURES AT TAPOSIRIS architectural features as the north shops (foundations and architectonic elements in stone, mud-brick walls covered with a coating). It was supplied with drinking water by rainwater collected and discharged into a cruciform tank located to the north. The angle of the buildings on both sides of an east-west axis indicates the prior circulation of water in this area, which was later replaced by the channel. The level of the tank indicates that the site chosen to dig the channel was originally an area not liable to flooding.

harbour traffic to pass, while minimizing the size of the humpback bridge. This would partly explain the asymmetry of the two arches, with a humpback passage which best fits the top of the vaults, which are all of different widths and heights (approximately 4.5 m and 4 m). Furthermore, the construction of the bridge can only be explained by the existence of the causeway and channel system. A limited investigation (6.2 m by 3.2 m) conducted in the foundation pebble bed (radier) of the north pier of the bridge, to the flood level of the lake, confirms that the construction of the bridge is contemporary with the digging of the channel, however it does not allow precise dating. Indeed, the stratigraphic study of this area shows that the bridge was built over a level that dates back to the Imperial Roman period; it confirms that the channel, the bridge and the closed harbour system can be dated to the Imperial Roman period.

The whole area was disrupted by major works after a period of stagnation and neglect. The objective was to create a navigation canal by digging a channel, and ensure access by strengthening and shoring the banks formed. The channel was dug and the waste was deposited to the north and south creating two sedimentary causeways, of unequal length but of a similar configuration – the causeway to the north extends for about 150 m, to the south for 1,700 m.

The East Part of the Lake System The lake basin is closed to the east by a built jetty, which was the subject of two campaigns in 2005 and 2006, following a preliminary phase of topographical mapping in 1999 and 2000. The cancellation of the following two campaigns (2008 and 2009), due to an exceptional rise in the water level, meant that certain points remain to be determined, in particular the chronology: while the date of abandonment is well attested by the coins and ceramics (first half of 7th century AD), uncertainties remain concerning the previous phases.

An anthropic embankment was formed by successive heaps of backfill piled up, stabilised by a gypsum screed. The most extensive works were carried out to the west, where the edge of the promontory needed to be cut back. This explains why there are only two artificial embankment slopes in this sector, to the north and the south, and why the causeway ends near the bridge at a height of some 5-6 m and then drops in irregular steps towards the east. The two sides are asymmetrical: the north slope, which follows the dug-out channel is steeper and supported by a series of retaining walls to prevent the collapse of the earth and therefore seal the passage, the south side has a shallower slope.

The jetty is located approximately 1,700 m to the east of the bridge (west entrance of the harbour basin) below the Temple of Taposiris. Between the end of the causeway and the tip of the jetty, there is an opening which is currently about 100 m wide. This probably allowed boats to enter the harbour basin from the east. We have not yet observed a structure that could act as a checkpoint like the bridge that closes access to the west or like the causeway running north-south which blocked the lake between the harbour complex and the southernmost bank (Boussac 2009).

The Bridge A dressed stone10 bridge11 guards the west entrance of the channel. Its present appearance has hardly changed since Coste’s sketch in 1820 (see Fig. 2-c): 47 m long including the ramps, and 9 m wide. Its maximum height is 2.55 m. It consists of two massive abutments of equal length (6.02 m for the south abutment and 6.55 m for the north abutment). The space of approximately 8.35 m between the two abutments is divided by a narrower intermediate pier (1.20 m). This pier is off-centre compared to the axis of the abutments, thus forming two passages of different widths (4.10 m to the south and 3.05 m to the north) which only allowed access to the channel to small boats (Bougia 1996). Study of the structure leads to several conclusions. First of all, the bridge was probably supported by masonry vaults (Figs. 5-7).12 The flatness of the natural terrain required the creation of a passage sufficiently high for the

Excavations have identified several structures: the jetty, which is interrupted by three openings – two flushes and a water supply channel – which cross its entire width; two basins, one square, the other in the shape of a hoop, located inside the dock and attached to the jetty, between the north flush and the water supply channel. The East Jetty The east jetty extends north-south and has a visible length of 230 m and a width varying of 5-6 m.13 It ends in the lake and it is possible to follow its extent for a further

10. Rodziewicz (1998: 102) only mentions a “structure”. In fact, the bridge was necessary to cross the channel for those coming from the south of the lake using the causeway and following the road north. 11. As is the case for most stone buildings in the region, it is calcarenite, white dune limestone, extracted from the quarries on the taenia. 12. This architectural analysis, and the model (Fig. 7) are the work of students at M2Pro Univ. Lyon II, especially R. Brunier and F. Jimenez, under the supervision of T. Fournet (CNRS, IRAA/HISOMA).

13. It is longer than the jetties built on the south shore at Marea (60 and 40 m long) at the site referred to as “Kibotos” (Blue & Ramses 2005: 10) and is similar to structures noticed during prospecting on an island of Mariout, at a site which played a major role in the management of lake traffic (ibid., Sites 23, 12).

91

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig. 5. View of the Taposiris bridge from north. © MAFT

Fig. 6. (above) Plan and elevation of the north pile of the bridge, November 2004. Lay out, drawing and DAO by M. El Amouri. © MAFT

Fig. 7. (left) Model and reconstruction of the bridge in Taposiris Magna (students of the Master Pro-Lyon II). © MAFT 92

M.-F. BOUSSAC & M. EL AMOURI: LAKE STRUCTURES AT TAPOSIRIS

Fig. 8. Overall plan and detailed outline of the eastern jetty. Lay out, drawing and DAO by M. El Amouri. © MAFT

20 m or so under water (Fig. 8). To the north it gradually disappears first towards the west and then the east. This asymmetry could indicate that the point of attachment of the causeway on dry land is different than at the southern lake end from that at the lake end. Further north, the silting of the area and recent constructions prevent the correct reading of its route. Situated on the axis, about 50 m to the north, is a masonry structure (ST900114) which may or may not be related to the jetty.

Fig. 9. The southern end of the eastern jetty. View from north. © MAFT

The jetty (Fig. 9) is built in blocks of regular size (0.50 x 0.20 x 0.25 m on average) arranged in headers, and the design is comparable to that of the harbour structures present at the neighbouring sites at Gamal and Quseir (Figs. 10 & 11). Excavated sections of the eastern jetty revealed four well preserved courses, and two additional courses which have now disappeared. Over its entire length, both sides of the jetty are punctuated with buttresses (Fig. 12). These 27 buttresses (16 on the east side and 11 on the west side) are maintained on several courses (three or four). Every second course has two headers while the alternate blocks are stretchers on their edge. They are on average 0.50-0.60 m wide and protrude by about 0.30 m. They stabilised the entire jetty, while encouraging micro areas of

14. An investigation was carried out in 2006 on this structure. The report on surface ruins shows an angle similar to that of the jetty. If it belongs to the jetty this would alter our interpretation of the extent of the closed harbour basin.

93

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig. 10. Jetty at Gamal. View from north-west. © MAFT

Fig. 11. Jetty at Quseir. View from south-east. © MAFT

Fig. 12. Plan and elevation of the inner south face of the north flush and buttress. Lay out, drawing and DAO by M. El Amouri. © MAFT

94

M.-F. BOUSSAC & M. EL AMOURI: LAKE STRUCTURES AT TAPOSIRIS Fig. 13. View of the north flush during excavation, from east. © MAFT

In addition, the east part of the flush is designed in order to circulate as much (surface) fresh water as possible (McCann 2003: 32)16 towards the interior of the harbour basin: a base of grey mortar17 at the base of the projection formed by the paving of the western part, has a slope to the east which hinders the entry of sediment into the harbour basin, while allowing the passage of a stream of fresh water. The western part of the bottom of the flush, paved in worked stone, is thus located more than 0.40 m above the mortar.

sediment accumulation, thus reducing the risk of sapping erosion due to currents and saline seepage.15 It is telling that particular attention was given to the construction of the eastern face of the jetty where the buttresses are more numerous. Indeed one notices a difference in treatment between the east side facing the open lake (the outside of the basin) that is subject to currents, and the west wall facing the interior of the basin. Investigation of the north flush revealed four courses of blocks to the east whereas there are only two to the west. A row of worked stones and a stone blockage line the east facing, but do not exist on the west side. In addition, the level of the upper beds of the blocks on the east side is slightly higher than that of the west side. These differences in the construction show that the jetty is a rampart against the open waters of the lake located to the east. To the west, the water in the basin is calmer and the construction of the jetty does not require these precautions.

Carved into the faces of the flush conduit on the paved western half, there are several vertical notches with corresponding horizontal grooves in the paving, which allow the recreation of the system of closing and filtering of the flush. It would appear that three of these locking systems could operate simultaneously. They probably held filter sluice gates or simple martelières type gates, which allowed the control of the flow of water according to currents and seasons while limiting the entry of sediments to the basin.18 Two twin notches have no corresponding grooves which implies the redesign of the paving and locking systems. A lightweight and retractable system to cross the flush probably existed for passage across the jetty, but no archaeological evidence shows its design.

The North Flush There are two flushes, designed to bring as much running water as possible into the basin and thus prevent silting; one near the south tip of the jetty, the other to the north, more than 70 m from the present shore line, only the second flush has been the subject of investigation (Figs. 12 & 13). Its east opening is shaped like a funnel with two facing walls built in a quadrant. The west opening of this passage is straight and 1.2 m wide. The axis of the flush is not perpendicular to the jetty but shifted 15 degrees to the north, which probably gives an indication of the direction of the wind and/or dominant currents in the region during antiquity.

16. Mc Cann 2003: 32; the author states that the constant flow of water helps control the temperature, oxygenation and salinity of the water, adding that the proximity of a source of fresh water is an advantage. 17. Grey mortar seems to have been chosen for the bed of the east entrances of the north flush and the water supply channel. However, pink hydraulic mortar is used around the upper areas of the north flush and water supply channel. 18. Does this system indicate seasons with high levels of sedimentation? Are these seasons connected to Nile floods? Indeed, the Nile is joined to Lake Mareotis via canals but underground water must also have raised the level of the lake during floods, bringing a great deal of sediment.

15. There is also a risk of erosion of the sides of the causeway facing the channel; this was prevented by the stone constructions, the ramp and the retaining wall.

95

LAKE MAREOTIS: RECONSTRUCTING THE PAST Fish Farming Systems One of the most interesting findings is the discovery of an original system for breeding/keeping fish. At the current stage of excavations, this consists of a water supply channel which crosses the jetty, a hoop-shaped basin and a fish-tank attached to the jetty inside the harbour basin (see Fig. 8). This complex, which has possibly not been entirely uncovered, and which shows signs of repairs, indicates intensive fish farming (Lafon 2001: 161, 165), i.e. breeding in artificial pools. From a typological point of view, these installations built on the lake are one of the types of fish farming installations analyzed in various recent studies, mostly devoted to the Roman world (Kron 2008): basins built by digging in the ‘sand’ and ‘masonry work’ (Type III) (see below Lafon 2001: 171). Plato tells us that fish farming was practiced on a large scale on the banks of the Nile and papyri from the Hellenistic and Roman eras reveal much information about ichthyotropheia, piscinai and other apodocheia (Chouliara-Raïos 2003; Brewer & Friedman 1989) of which our system is an example. The region of Canope is home to several piscinae dugout of the rocks, some of which have long been known (Breccia 1926; Abd El-Maguid forthcoming) but the installation uncovered in Taposiris is unique to this day around the shores of Lake Mareotis (Lafon 1998: 573).19 However, the association of harbour installations and fisheries is not surprising and the facilities found at Taposiris are similar for example to those found at Kenchreai dating to around AD 80 (Scranton, et al. 1978):20 basins are created near the jetty (Rothaus, et al. 2008; Lafon 2001: 162)21 and connected to one another and with the sea through channels.

tion basin. As with the flush, the grooves holding the filter sluice gates or locking martelières were modified during the rehabilitation of the water supply channel. The walls of the water supply channel are built in worked stones bound with pink mortar. The blocks which form the entrance to the channel are chamfered to facilitate the entry of water. It is a simple funnel system. At the entrance, the bottom of the channel, like the bottom of the flush, is made of grey mortar sloping towards the east. The water supply channel, like the flush, might have been covered by a crossing system, but we did not find any trace of it, although we noticed the last course on its north wall was slightly battered. The western part of the south interior face of the channel (Fig. 14) appears to have undergone repair with small unworked stones which contrast with the usual blocks of worked stone of the walls of the channel. Other repairs are visible in the south wall of the hoop-shaped basin. The meticulous cleaning of the water supply channel led to the discovery of seven coins wedged between the paving stones of the channel, and thus protected from the cleaning work (Field Unit [FU] 9240). Five of them belong to the Later Roman Empire (AE4) and two are Byzantine dodekanoummia of the 6th-7th century AD, including one dated to Heraclius (AD 610-641).23 Thus, the water supply channel must have functioned at least from the 4th century to the 7th century AD.

The Water Supply Channel The water supply channel crosses the jetty perpendicularly over its entire width (Figs. 14 & 15). This channel is 0.70 m wide, 6 m long and 0.85 m deep at the point where the third course is still preserved. It consists of a small decantation basin, a ledge and three closing systems with notches carved opposite one another, 0.50-1 m apart: they were used to create a kind of lock, like in Kenchreai (Scranton, et al. 1978).22

Its period of abandonment is suggested by traces of blocking up. Chamfered blocks the same size as those still in place, and a block with a groove identical to that found on the blocks still in place, were discovered during excavations. These blocks thus come from the courses of the water supply channel. The chamfered blocks were placed at the entrance of the water supply channel to block it: they fit perfectly in the entrance of the water supply channel while keeping one layer horizontal, indicating that they were probably deliberately placed there. The backfilling of the channel thus seems a deliberate action indicating the stopping of the water supply or even of fishing, but ensuring the continued functioning of the jetty.

One of these systems of notches still has its corresponding grooves cut in the bottom of the paving blocks (Fig. 16). This is the first paving block (to the east) and acts as a dam to stop the sediment accumulated in the small decanta-

The Hoop-shaped Basin The west end or exit of the water supply channel flows into a hoop-shaped structure (about 6 x 3 m) the base of which is paved (Figs. 8, 17 & 18). It is built in blocks of worked stone, arranged in a single row and preserved to two or three courses. The blocks are arranged in stretchers and their bases are covered with a hydraulic coating that forms a slight slope on the inside edge of the basin (see Fig. 18). This poorly preserved coating is only applied at the junction between the base pavement and the elevation.

19. Lafon 1998: 573, n. 3, notes that coastal speculation has caused the disappearance of many installations in the western Mediterranean over the past fifty years. Lake Mareotis is also threatened by intensive property speculation: during our first works on the east jetty area, it was divided up into plots of land, shown by boundary makers. 20. Scranton, et al. 1978. Their construction dates back to around AD 80. 21. Also see Lafon 2001: 162, n. 89, for Cosa: the author notes that in the port of Cosa, “canals joining the lagoon which has various dams, flow directly into the port, and form part of the same project”. 22. Scranton, et al. 1978. Note the presence of several sluices per channel, necessary to form intermediary basins and fish traps: the dimensions of the basins (1 x 0.7 m) are similar to those found at Taposiris.

23. Study by T. Faucher, Appendix 2, infra.

96

M.-F. BOUSSAC & M. EL AMOURI: LAKE STRUCTURES AT TAPOSIRIS

Fig. 14. Lay out, elevations and sections of north and south walls of the water supply channel through the eastern jetty. Lay out, drawing and DAO by M. El Amouri. © MAFT signed so as to create a slope down towards the fish-tank basin.

The hoop-shaped basin, fed with water through the supply channel, is closed by a double sluice gate system, to the north-east and south: to the north-east, vertical and horizontal grooves were cut into the blocks forming the junction between the channel and the basin. To the south, the funnel-shaped end is fitted with the same system, indicating the method of control of the flow of water into the fish-tank, a large square structure enclosed by walls with double facing (Fig. 19). The junction was also de-

The north wall of the basin is an extension of the north wall of the water supply channel and seems to be a renovation or addition. Indeed, this wall is not linked to the channel and is built with smaller, roughly hewn stones, some of which are reused. Perhaps the link with the channel did not initially exist, at least in this form. 97

LAKE MAREOTIS: RECONSTRUCTING THE PAST The Fish-tank The fish-tank is a square basin whose interior dimensions are about 8 m x 8 m. It features walls angled north-south and east-west, all of different form. The west face of the east jetty serves as its east wall. The northeast corner of the basin corresponds to the end of the west face of the jetty. Investigations carried out on the extension of this face showed that no repairs had been made at this location and that the jetty had its original form. It appears that the south and west walls of the basin were entirely built under water. The south wall of an average width of 1.5 m, consists of two parallel faces built in stone and filled with green clay sediment (FU 9221 to the east and FU 9225 to the west). This wall has an opening, built on at least two courses and equipped with a sluice gate system to control the flow of water. The proximity of the flush installed in the jetty provides a fresh water supply to this southern sluice gate (Leatham & Hood 1958-59: 265).24 The position of this pool is therefore also chosen because of the flow of fresh water into he harbour basin. Sluice Gates and Water Flow A system of fresh water flow was set up between the various structures composing this fishery. The flow or the blocking of water, or at least of fish in the case of filter sluice gates, is ensured by a system of sluice gates or martelières type gates. These are all designed in the same way: the vertical grooves cut in the facing blocks allow their positioning across passages, while horizontal grooves cut in the paving slabs ensure a good grip. Some of these gates must have let water filter through to ensure a good regeneration of water in the fish tank. However, no archaeological evidence indicates their construction material.25 The size – between 0.10-0.20 m wide for the different grooves in the facing or paving blocks – suggests that at least the external framework of the sluice gates was made of wood. They may have been solid and entirely made of wood, or with openings, made with a wooden frame and metal grid.

Fig. 15. Water supply channel through the eastern jetty opening into the hoop-shaped basin. View from east. © MAFT Fig. 16. Grooves and notches in the water supply chanel. View from north. © MAFT

The water from outside enters from two openings located to the south and north of the whole system: • In the south, fresh water arrives through the flush located in the jetty. It enters the harbour basin and indirectly feeds the fish tank by the passage located in its south wall. It is closed by a sluice gate installed upstream of the paving inclined towards the fish-tank. • In the north, the water supply channel feeds fresh water into the hoop-shaped basin. The arrival of water is regulated by four successive sluice gates, which leave water to flow towards the hoop-shaped basin while preventing fish from escaping. The frequency of these sluice gates also facilitates maintenance work (cleaning, repairs, etc.). A rearrangement may explain the position of the sluice gate located to the west end of the water supply channel, which opens directly onto the northeast corner of the hoop-shaped basin. The space partitioned off by this sluice gate and the one before it, is the largest in the channel. The last sluice

24. Leatham & Hood 1958-59: 265: n. 8, concerning the need for a flow of fresh water into the fish tanks; the authors cite Q. Hortensius criticising M. Luculus whose fish tank does not have enough. 25. Sciallano (1997: 17) notes bronze sluices with holes, like those recommended by Columella (viii.17,6) because they are fine nets which allow water to pass through while trapping the fish (Leatham & Hood 1958-59: 265, n. 9). But other systems exist in wood or composite (wooden frame and metal grid for example), or in stone like at Mochlos (Leatham & Hood 1958-59: 275 and plate 63(a)).

98

M.-F. BOUSSAC & M. EL AMOURI: LAKE STRUCTURES AT TAPOSIRIS Fig. 17. View of the hoopshaped basin, the water supply channel and the fish-tank. View from north. © MAFT

Fig. 18. The pavement in the hoop-shaped basin. View from north. © MAFT

Fig. 19. The opening between the hoop-shaped basin and the fish-tank. View from southwest. © MAFT

99

LAKE MAREOTIS: RECONSTRUCTING THE PAST gate is fitted across and at an angle to the axis of the channel and is therefore directed towards the centre of the loop, which may also be due to a repair. This all suggests that this sluice gate dates from a later date.

9216/9220) filling the hoop-shaped basin, principally uncovered storage containers (Amphorae LRA 1, 4 LRA and LRA 5/6), indicate a final phase of use of the fish tank during the first half of the 7th century. This hypothesis is supported by the absence of Egyptian amphorae Egloff 167, well documented on Alexandrine and Mariout sites from the mid-7th century. Similarly, among the coins wedged between the paving stones on the channel bed, there are two Byzantine dodekanoummia from the 6th-7th century AD, including one dated to Heraclius (AD 610-641; see Appendix 2).

The passage between the hoop-shaped basin and the fishtank, which is also blocked by a sluice gate, is equipped with the same system. Thus, all these adjustments help to prevent the basins from silting up while encouraging the constant flow of water, which is a key element for the survival of fish in captivity (Sciallano 1997: 17). Furthermore, the shallow and muddy bed is the perfect environment for fish farming (Lafon 2001: 159).26

This last phase of use of the system also relates to the last phase of the warehouse excavated in 2000 on the causeway, strategically located near the causeway which crosses the lake (Boussac 2009: 129). The finds included three Heraclius dodekanoummia and late amphorae (LRA 1, 4, 5/6 and 7).

The Fish The fish species identified in the Field Unit related to the use of the fish tank (FU 9215 and 9220 in particular) at this stage of the study are mainly marine fish or fish which live in brackish water (Lafon 1998: 575).27 A preliminary diagnosis carried out using photographs to visually identify the species28 in advance of a comprehensive study, indicates the presence of sturgeons and garfish (Belone belone). At the present stage of work, all the excavated structures seem too small to have housed real breeding farms (Guest-Papamanoli 1986: 301). We suppose it was for the temporary storage of live fish, like at Kenchreai.

The intensity of lake activities during late antiquity corresponds to what the texts and archaeology tell us of the role of Taposiris during this period: the identification in 2009 of late antique thermal baths in the middle town, comparable to those at Marea, further strengthens these findings. However, nothing is yet known about the Hellenistic installations, although we are collecting more data on the city from the 2nd - 1st century BC, and even from the end of the 3rd century: the temple, according to Dr. Z. Hawass, was founded by Ptolemy IV, suggesting that the city’s development is linked to (or accelerated by) this royal intervention (Hawass 2008: 29). However, it implies nothing about lake constructions. Nevertheless, in the west, the area covered by the backfill from the digging of the channel in the first half of the 2nd century AD (Boussac 2009: 137-141), revealed many imported amphorae (Rhodians and Cnidians) which show the intensity of trade from the outside via Alexandria, at least between the mid-2nd century BC and the early 1st century BC. Although their fragmentation prevents us from being precise, the site where these containers were found, close to the shore, suggests that we are not dealing with land transport, but transport on the lake (Bernard 2009). This storage area suggests specific installations of which nothing remains.

Chronology In the east sector of the harbour system, the various phases of installation of the fish farming facilities and the chronological link between the visible system and the construction of the jetty are currently difficult to specify. Whereas the west harbour system (bridge and channel) can be dated to the 2nd century, the state of visible installations in the east remains unclear: the traces of repairs in the system between the hoop-shaped basin and the channel do not make it possible to date its construction, nor that of the visible jetty. All that can be said about the jetty is that it cannot be later than the 4th century. The presence in the filling of the south wall of the fish tank, of amphorae remains LRA 1 and fragments of containers LRA 4, some of which are dated between the last third of the 5th century and the middle of the 6th century,29 provides an approximate dating for its construction (terminus post quem).

Nothing is known about the structures from the beginning of the Imperial era: in describing Taposiris as the gate (eisbolè) of Egypt in his Letter to the Alexandrians in AD 41, Claudius refers to the fiscal and military role of the city, which was probably associated with specific facilities (Boussac 2001). However, the oldest phase of the harbour that we were able to identify is that which includes the digging of the channel, which is not earlier than the first half of the 2nd century AD (Boussac 2009).30 If one considers that the warehouses and the fish farming facilities operated until the 7th century, then one understands to what extent the intermediate stages of evolution still escape us.

However, ceramic and monetary data provide consistent dates for the last phase of use of the system: ceramic material discovered in the last archaeological layer (FU

26. Lafon 2001: 159: “breeding requires very sheltered water with a shallow sandy or silty bed: lagoon shores provide the most favourable environment”. 27. Lafon 1998: 575. The author insists that brackish water is favourable for breeding as it helps the development of young fish and attracts fish. 28. S. Cravhino, pers. comm. (PhD Univ. Aix-Marseille I). 29. Study by D. Dixneuf, IFAO. See Appendix 1, infra.

30. See S. Marquié’s study in Boussac 2009.

100

M.-F. BOUSSAC & M. EL AMOURI: LAKE STRUCTURES AT TAPOSIRIS

Appendix 1 The Ceramics in the Harbour Area: the Eastern Jetty (Sector 9)

Fig. 20. Main types of ceramics found in Taposiris Magna (Sector 9) during the Late Roman era.

Archaeological excavations carried out in the area of the harbour basin, and more precisely on the eastern jetty (Sector 9) uncovered a few ceramics often very fragmentary and eroded. They amount, except for fragments of bodies, to more than 243 sherds, or 119 individuals.31 However, most of them, i.e. more than 66% of the ceramic assemblage (NMI = 79), consist of storage and transport containers, both locally produced and imported, characteristic of the Late Roman period, dating which is confirmed by the study of coins by Thomas Faucher (see Appendix 2 this paper).

pereur & Picon 1989: 236-243),33 and were used for the storage and transport of wine, and possibly olive oil. Fragments found in Area 9 (Fig. 20, a) are related to the B1 type according to the classification of Late Eastern Amphorae established by Dominique Piéri (2005). This form, generally attributed to the 6th and 7th centuries, is characterized by “thick rolled rim (an edge to mid-neck recalls the moulding of the previous type)” (Piéri 2005: 75). The handles are now almost unribbed. The clay is rather dense and a general trend emerges: marl clay, fine texture and a red to light brown section. Inclusions are of large size, and consists mainly of grains of quartz, some white and sometimes red.

Delphine Dixneuf, IFAO, Cairo

1. The Imported Amphorae and Ceramics • The LRA 1 Amphorae32 Among the imported containers at the site, LRA1 amphorae are the second most common type (NMI = 10). This evidence finds parallel in all Late Roman Egyptian sites, from the second half of the 4th century until the mid-7th century, or slightly beyond. These jars were produced on the southern coast of Turkey, Cyprus and Rhodes (Em-

• The LRA 4 Amphorae LRA 4 productions from southern Palestine represent 32.7% of NMI, i.e. 39 individuals. The general form of these amphorae (Fig. 20, b) is morphologically close to Pieri’s type B and dated between the middle of the 6th and the 7th century (Piéri 2005: 106-107). Many of these amphorae have a sandy and gritty marl clay, with a medium dense fabric and contain several grains of quartz, some grey and white particles of various sizes. The section is usually homogeneous, from buff to orange. Shaping and surface treatment are quite coarse, and clay accretions are frequently observed on the neck. The macroscopic charac-

31. All the sherds are analysed using the criteria of the clay and shape to determine the MNI – Minimum Number of Individuals – by context first, then for the whole sector. Rims, handles, bottoms and bodies fragments were counted. NMI or the global minimum, estimated for the sector, is the highest number among the different morphological features, except the bodies. Regarding the amphorae, the number of handles is divided by two. 32. The term Late Roman Amphorae (LRA 1 to 7) refers to the typology of Late Eastern amphorae established by Riley 1981: 85-122 for the material found in Carthage.

33. However, only two workshops have been excavated in Cyprus: Paphos and Ziggy. See Demesticha & Michaelides 2001: 289-296; Demesticha 2003: 469-476.

101

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig. 21. Ceramics from the filling of the hoop-shaped basin (FU 9216/9220).

teristics help us to locate the region of production in southern Palestine, especially in the area of Gaza, Ashkelon and Ashdod.34 However, it is interesting to note the similarity between the clay of Palestine amphorae and ceramics manufactured from Mareotic clays.

Sector 9 are morphologically close to Piéri’s “bag-shaped amphora, type 4” (Piéri 2005: 117, fig. 76). It is a globular amphora of small size; the neck is short and cylindrical or convex, ending with a small round and/or flared rim, with a sinuous profile (Fig. 20, c). On the Kellia site, Françoise Bonnet-Borel sets the appearance of these containers during the second quarter of the 7th century (Bonnet 1983: 442), however it is not impossible that they started as early as the late 6th century. They are still present during the Fatimid era, with no major change in the shape (Vogt 1997: 256, 258). As regards the identification of site productions, only one group of clay could so far be identified: marl clay, medium density fabric, sandy; the section is usually brown to buff. The inclusions consist of several grains of quartz, some grey and white particles of various sizes. The macroscopic examination of the amphorae, which can logically be assumed to be local or at least regional productions, questions the distinction between Palestinian and Mareotic productions. Thus, two hypotheses can be put forward: they are either Egyptian or Palestinian containers exported with the LRA 4 Amphorae. Only macroscopic and petrographic analysis would answer this question.

In addition to transport containers of types LRA 1 and LRA 4, a few fragments of amphorae and ceramics could not be identified with the exception of a fragment of the body of an African sigillata and two elements of Cypriot sigillata. It is a ring-base low bottom and a rim which seems to correspond to Hayes’ 9B shape and date from late 6th to late 7th centuries (Hayes 1972: 378-382). 2. Egyptian Production • Egyptian Production in Marl Clay With the exception of 12 residual fragments (NMI = 6) of AE 3 amphorae, marl clay productions are mainly illustrated by globular or ovoid containers, more commonly referred to as “bag-shaped”, which were produced in many eastern Mediterranean sites, more specifically in the Levant, Palestine and Egypt (NMI = 15). Containers found in

Quite logically, common ware is illustrated primarily by marl clay products, some with alluvial content, manufactured without any doubt in Mareotic workshops, on the site

34. For a state of question on LRA 4 amphorae and bibliography, see Dixneuf 2005: 54-62.

102

M.-F. BOUSSAC & M. EL AMOURI: LAKE STRUCTURES AT TAPOSIRIS

Appendix 2 The Coins from Sector 9

or at least nearby (Majcherek 2001: 60-64). The shape repertory includes dishes with rounded rim (Fig. 20, d), some of them with a high-placed carination in their external wall, cups with flattened carination (Fig. 20, e), fragments of storage jars and vases for liquids and some cooking pots (Fig. 20, f).

Thomas Faucher, Sorbonne, Paris

A number of coins whose study is currently in progress were found during the excavations in Sector 9. After some preliminary restoration, it is possible to make a few remarks here.

• Egyptian Production in Alluvial Clay These productions from the Delta and/or the Nile Valley are only a small part of the ceramics: 17 individuals, i.e. 14.2% of all ceramics identified. Apart from some fragments of three AE amphorae, a bottom and a carinated shoulder of a wine LRA 7 container (Fig. 20, g) have been identified. Common wares are illustrated by bowls (Fig. 20, h),35 dishes with rounded rims, and two cooking pots with a C shaped rim dated at the Kellia to the 7th century (Egloff 1977: 103, Pl. 52 (no. 4 & 6, type 138) and to the first half of the 7th century in Baouit.36

Most of the coins belong to the Late Roman and Byzantine periods. No coin dating to earlier than the 4th century AD was found, whether in or out of stratigraphical context. All the Late Roman coins (12 items) are small denominations (ca 10 mm) commonly called AE4 (AE for bronze and 4 for the size). The identification of these coins is not possible until the restoration is completed but it is unlikely that more information about the type and date of these coins will be gained, for two reasons. First, the proximity of the lake water has facilitated the corrosion of these coins which are often in poor condition. Secondly, these coins, commonly dated to the 4th-5th centuries, were often molded coins, unlike struck coins, produced more or less legally to overcome the lack of currency. This technique allowed for the production of a large number of coins in a short time, but makes coins almost illegible, even when they are just produced. Therefore, it is logical that their circulation and subsequent deposition in the ground has erased any information.

3. The Assemblages • The Fishery Complex The southern wall of the fish-tank consists of two faces with a filling of green clay (FU 9221/9225). Few ceramics were found in the filling but they provide a chronology for the construction of the fish-tank and probably the whole complex. They consist of bodies of some LRA 1 amphorae and fragments of LRA 4 containers including a bottom and a rim close to Piéri’s B1 form; this form is dated from the last third of the 5th to the middle of the 6th century (Piéri 2005: 105-106).

In addition to these Roman coins, there are also three dodekanoummia (a coin of 12 noummion) easily recognisable by the letters IB on the reverse, produced using the same molding process. Among the latter, were identified one coin from the reign of Justin I (518-527) and a second from the reign of Heraclius (610-641). It is difficult to give a precise date for molded coins since the coin might have been molded later than the date of production of the model.

The ceramic material discovered in the last archaeological layer (FU 9216/9220) of the hoop-net shaped basin consists mainly of storage containers, i.e. LRA 1, LRA 4 and LRA 5/6 Amphorae, and fragments of common wares which might be linked to food consumption by the fishermen (Fig. 21). Thus, this material attests to a final phase of use of the basin during the first half of the 7th century. This hypothesis is confirmed by the absence of Egyptian Egloff 167 amphorae, well documented on Alexandrine and Mariout sites from the mid-7th century onwards.

The Alexandrian excavations emphasised the importance of this phenomenon whose magnitude was often underestimated in the past. This manufacturing technique is attested as early as the Ptolemaic era and continues in Roman and Byzantine periods despite its apparent restricted use during the Early Roman Empire; in any case it is difficult to say if it was outlawed by the state and successfully repressed.

To conclude, the study of Late Roman pottery discovered at Sector 9 offers some interesting insights into the commercial activities and trade in the harbour area, which was the western customs of Alexandria on the Mareotis Lake since Roman times if not before. In addition to local marl clay productions, the repertory includes several imports, mainly of wine containers from the eastern Mediterranean and some ceramics from the Delta and/or the Nile Valley, reflecting the economic vitality of the region and fishing activities.

When fully restored, the coins from Sector 9 will hopefully provide a more precise datation and help discern more accurately the different periods of use of the eastern jetty area.

35. One should tentatively compare this rim fragments with a series of dishes found at Kellia and dated to the first half of the 7th century; however these are of alluvial clay and with a painted decoration on white slip. See Bonnet 1994: 378-379, fig. 231 (No. 138). 36. Unpublished material.

103

LAKE MAREOTIS: RECONSTRUCTING THE PAST Bibliography Bernard, S.G., 2009, Alexandrian Tainiai and Land Traffic Patterns: A Note on the Amnesty Decree (P. Teb. I 5) in Light of the Topography. Zeitschrift für Papyrologie und Epigraphik 168: 265-270. Blue, L., & Ramses, S., 2005, Lake Mareotis Research Project – Preliminary Report from Pilot Survey August 2004. Newsletter of the Archaeological Society of Alexandria 15 (February): 5-16. Bonnet, F., 1983, Poteries, verres, monnaies, décors et inscriptions. In R. Kasser & J.-M. Alès (eds.), Survey archéologique des Kellia (Basse-Égypte). Rapport de la campagne 1981: 423-480. Louvain. Bonnet, F., 1994, Le matériel archéologique récolté en 1977, 1982 et 1983 aux Qouçoûr er-Roubâ’iyât. In R. Kasser & F. Bonnet (eds.), Explorations aux Qouçoût er-Roubâ’îyât. Rapport des campagnes 1982 et 1983: 349-406. Louvain. Bougia, P., 1996, Ancient bridges in Greece and coastal Asia Minor. PhD thesis, University of Pennsylvania. Boussac, M.-F., 2001, Deux villes en Maréotide: Taposiris Magna et Plinthine. Bulletin de la Société française d’égyptologie 150: 42-72. Boussac, M.-F., 2007, Recherches récentes à Taposiris Magna et Plinthine, Égypte (1998-2006). Comptes rendus des séances de l’Académie des inscriptions et belles-lettres (janvier-mars): 445-479. Boussac, M.-F., 2009, Taposiris Magna: la création du port artificiel. In F. Dumasy & F. Queyrel (eds.), Archéologie et environnement dans la Méditerranée antique: 123-142. Paris. Breccia, E., 1914, Alexandria ad Aegyptum. Bergamo. Breccia, E., 1926, Monuments de l’Égypte gréco-romaine, I: Le rovine e i monumenti di Canopo. Bergamo. Brewer, D.J., & Friedman, R.F., 1989, Fish and fishing in ancient Egypt. Warminster. Chouliara-Raïos, H., 2003, La pêche en Égypte à la lumière des papyrus grecs. Ioannina. De Cosson, A., 1935, Mareotis: Being a Short Account of the History and Ancient Monuments of the NorthWestern Desert of Egypt and Lake Mareotis. London. Demesticha, S., 2003, Amphora production on Cyprus during the Late Roman Period. In C. Bakirtzis (ed.), De Rome à Byzance; de Fostat à Cordoue. Évolution des faciès céramiques en Méditerranée (Ve – IXe siècles), actes du VIIe congrès international sur la céramique médiévale (Thessalonique, 11-16 octobre 1999): 469476. Athènes. Demesticha, S., & Michalides, D., 2001, The Excavation of a Late Roman 1 Amphora kiln in Paphos. In E. Villeneuve & P.M. Watson (eds.), La céramique byzantine et proto-islamique en Syrie-Jordanie (IVe - VIIIe centuries AD. J.-C.). Actes du colloque d’Amman (3-5 décembre 1994): 289-296. Beirut. Dixneuf, D., 2005, Production et circulation des biens à Gaza durant l’Antiquité tardive: le témoignage des amphores. In C. Saliou (ed.), Gaza dans l’Antiquité Tardive. Archéologie, rhétorique et histoire: 54-62. Salerne.

Egloff, M., 1977, Kellia. La poterie copte. Quatre siècles d’artisanat et d’échanges en Basse Égypte. Genève. Empereur, J.-Y., & Picon, M., 1989, Les régions de production d’amphores impériales en Méditerranée orientale. Amphores romaines et histoire économique: dix ans de recherché. Actes du colloque de Sienne (2224 mai 1986): 236-243. Rome. Empereur, J.-Y., & Picon, M., 1998, Les ateliers d’amphores du Lac Mariout. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athènes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole française d’Athènes, 11-12 décembre 1988: 75-88. Athens, Paris. Guest-Papamanoli, A., 1986, Archéologie, ethnographie ou ethnoarchéologie des ressources marine de sites cotiers. Le cas de la pêche aux muges dans les lagunes de la Grèce Occidentale. In F. Gerard (ed.), VIèmes rencontres internationales d’archéologie et d’histoire d’Antibes. L’exploitation de la mer de l’Antiquité à nos jours. II: La mer comme lieu d’échanges et de communication : 281-303. Valbonne. Hawass, Z., 2008, In search of Cleopatra and Mark Antony. Horus (August/September): 26-29. Hayes. J.W., 1972. Late Roman Pottery. London. Kron, G., 2008, Reconstructing the Techniques and Potential Productivity of Roman Aquaculture in the Light of Recent Research and Practice. In E. Hermon (ed.), Vers une gestion intégrée de l’eau dans l’empire romain: 175-185. Rome. Lafon, X., 1998, Piscinae et pisciculture dans le bassin occidental de la Méditerranée. Journal of Roman Archaeology 11: 573-582. Lafon, X., 2001, Villa Maritima, recherches sur les villas littorales de l’Italie romaine (3ème s. av. J.-C. / 3ème s. ap. J.-C.). Rome. Leatham, J., & Hood, S., 1958-59, Submarine Exploration in Crete, 1955. The Annual of the British School at Athens 53-54: 263-280. El-Maguid, A., forthcoming, Maamourah Surveys: Preliminary report. In H. Tzalas (ed.), Tropis VIII: Eighth conference on ship construction in antiquity, Hydra, Greece, August 2002. Athens. Majcherek, G., 2001, Marea 2001: Note on the Pottery. Polish Archaeology in the Mediterranean 13: 60-64. McCann, A.-M., 2003, The Roman Port and fishery of Cosa, a short guide. Rome. Ochsenschlager, E.-L., 1979, Taposiris Magna: 1975 season. First International Congress of Egyptology: 503-506. Berlin. Ochsenschlager, E.-L., 1999, Taposiris Magna. In K.A. Bard (ed.), Encyclopedia of the Archaeology of Ancient Egypt: 759-761. Oxford. Pacho, J.-R., 1827, Relation d’un voyage dans la Marmarique et la Cyrénaïque. Paris. Piéri, D., 2005, Le commerce du vin oriental à l’époque byzantine (Ve – VIIe siècles). Le témoignage des amphores en Gaule. Beyrouth. 104

M.-F. BOUSSAC & M. EL AMOURI: LAKE STRUCTURES AT TAPOSIRIS Riley, J.A., 1981, Pottery from the Cisterns 1977.1, 1977.2 and 1977.3. Excavations at Carthage 1977 conducted by the University of Michigan 6: 85-122. Rodziewicz, M.D., 1998, From Alexandria to the West by land and waterways. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athènes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole française d’Athènes, 11-12 décembre 1988: 93-103. Athens, Paris. Rodziewicz, M.D., 2002, Mareotic harbours. In C. Décobert (ed.), Alexandrie médiévale 2: 1-22. Le Caire. Rothaus, R.M., Reinhardt, E.G., & Noller, J.S., 2008, Earthquakes and Subsidence at Kenchreai: Using Recent Earthquakes to Reconsider the Archaeological

and Literary Evidence. In W.R. Caraher, L. Jones Hall & R.S. Moore (eds.), Archaeology and History in Roman, Medieval and Post-Medieval Greece: Studies on Method and Meaning in Honor of Timothy E. Gregory: 53-66. Aldershot. Sciallano, M., 1997, Poissons de l’antiquité. Catalogue d’exposition, 1er avril-31 août 1997. Musée d’Istres. Scranton, R., Shaw, J.W., & Ibrahim, L., 1978, Kenchreai Eastern Port of Corinth, Vol. I. Leyde. Vogt, C., 1997, Les céramiques ommeyyades et abassides d’Istabl’Antar-Fostat: traditions méditerranéennes et influences orientales. In G. Démians d’Archimbaud (ed.), La céramique médiévale en Méditerranée. Actes du VIe congrès de l’AIECM2 (Aix-en-Provence, 13-18 novembre 1995): 243-260. Aix-en-Provence.

105

LAKE MAREOTIS: RECONSTRUCTING THE PAST

106

M. BERGMANN, ET AL.: SCHEDIA ON THE CANOPIC NILE

Schedia, Alexandria’s Harbour on the Canopic Nile. Interim Report on the German Mission at Kom el Giza/Beheira (2003-2008) Marianne Bergmann, Michael Heinzelmann & Archer Martin Historical Sources and Research History Shortly after the foundation of Alexandria, the new metropolis was connected to the Canopic Nile by an artificial, 30 km long channel. At the point where the canal branched off the Canopic Nile a new town, Schedia, was founded (see in general: RE 1921: 401-403; RE 1899: 2030; Fig. 1).1

to display them. At the same time they are an indication of the formation of a prosperous citizenry with an understanding of itself as inhabitants of Schedia. In Roman times Christianity seems to have established itself early under Alexandrian influence. Schedia received its own episcopal seat. In spite of its importance the site as well as the Nile Branch, is still virtually unknown. The river and the canal have now vanished from the landscape. Schedia itself was still a prominent stretch of hills at the beginning of the 20th century with a length of about 2.5 km and a width of 1.5 km. Intensive sebakh extraction and the increasing encroachment on the ancient settlement site by the villages of Kom el Giza, Kom el Hamam and Kom el Nashwa have, however, made the topography so unclear that the place is hardly to be noticed in the landscape. Only in a few places can the flat remains of the koms and single structures be recognized. The greatest height in the territory of Kom el Hamam is occupied by the present cemetery.

In its important river harbour all goods coming from Upper Egypt had to be transferred to smaller vessels. It was also a customs station wherefore a pontoon bridge was installed (σχεδία), which gave its name to the place. Schedia seems to have been a flourishing Greek-dominated polis throughout Ptolemaic and Roman times and was inhabited until the early Byzantine period. Ancient sources suggest the existence of ample harbour and storage installations with the personnel and administrative infrastructure that belonged to them (Strabo, Geogr. 17,1,16; Procopius, Aed. 6,1). The inscriptions found in the area of Schedia provide further evidence of the character of the place (Bernand 1966: 45 ff., 1970a: 329 ff.; Botti 1902). Thus, cults of Zeus Soter and Athena Polias are attested, along with a ship and military station already in the late 4th century BC. About 115 BC the garrison station at Schedia dedicated a Kleopatreion. These cults, as well as early Ptolemaic grave inscriptions with Greek names, appear to indicate a high percentage of Hellenic inhabitants or at least of strongly Hellenized people in the city founded ex novo. There was, however, a Jewish colony as early as the time of Ptolemaios III Euergetes (246-221 BC) with one of the oldest known synagogues attested by an inscription – evidence of the early multicultural character of this harbour city. In the Roman Imperial period Schedia calls itself a polis in inscriptions and apparently has its own magistrate (archon). Various honorary statues and bases for private citizens and emperors found in the area of Kom el Giza suggest that there must have been appropriate public spaces in which

After the first vague attempts to localize Schedia in the 18th century, it was the Napoleonic expedition of 1798-1802 that first concerned itself with Schedia and the Alexandrian canal (Jomard & Jacotin 1818/25; Bernand 1970b: 422). In this connection the place is shown on the first accurate map as one of largest tells in the Nile Delta. A participant in the expedition, Jean Michel Le Père, was the first to propose the localization of the beginning of the canal at El-Karyun and the identification of the group of mounds of Nashwa and Kom el Giza as Schedia. Subsequently, Ptolemaic-Roman finds begin to be reported in the area later identified as Schedia, mostly chance finds by farmers or people engaged in quarrying sebakh. These objects, which include a series of honorary and funerary inscriptions and a Nilometer, finished in part in private collections but mostly in the museums of Cairo and Alexandria.

1. This project is being carried out by the Institute for Classical Archaeology of the University of Goettingen in cooperation with the Archaeological Prospection Service of the University of Southampton. Further support came from the German Archaeological Institute, the Department of Geodesy at the University of Stuttgart, the Geology Department of Mansoura University and the Centre d’Études Alexandrines. The project has been funded first by the Fritz Thyssen Foundation and since 2004 by the German Research Foundation. We thank the Supreme Council of Antiquities and its Director General, Dr. Zahi Hawass, for the permission to work at Schedia. For help and support we thank the head of the Antiquities Service for the Department of Beheira, Fawzi el Choulani, and the director of the SCA’s excavations at Schedia, Ahmed Abd el Fattah.

It was only in the 1980s, though, that the first regular archaeological work was done at the site, when the Egyptian Antiquities Service carried out some rescue excavations because of the growing threat to the ancient site through modern construction (Abd el Fattah 1988, 1998, 20002001). Several areas of varying sizes were investigated on the eastern edge of Kom el Giza and at Kom el Hamam, that is in the eastern part of the ancient settlement (see Fig. 4). Numerous building structures of the late Hellenistic, Roman and Byzantine periods were uncovered. A high number of Hellenistic objects were found, among them numerous coins. The excavated areas were restored and protected by the Egyptian Antiquities Service. 107

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig. 1: Alexandria and its hinterland in 1925, Survey of Egypt, sheet 4 (detail). The New Evidence Since 2003 thanks to the Supreme Council of Antiquities (SCA), the Institute of Classical Archaeology of the University of Goettingen has conducted an interdisciplinary project consisting of a topographical survey, large scale geophysical surveying, geological investigations, selective stratigraphic excavations and a comprehensive study of the finds (Bergmann & Heinzelmann 2007).

boundaries can also disclose characteristics of the ancient topography, because some of them seem to document ancient boundaries, like the former edge of a river branch. Geophysical and Geo-archaeological Investigations Three seasons of geophysical surveys have been conducted at the site since 2003 by the University of Southampton, to assess the archaeological remains of the area surrounding the Schedia excavations (Figs. 2 & 3). The surveys were carried out using magnetometry. Furthermore, in 2005 a series of deep drillings and sedimentary analyses was carried out in collaboration with the Geology Department of the University of Mansoura, which will give further evidence of the course of the Canopic Nile and the canal.

Topographical Survey In order to integrate old maps and the results of the excavations, geomagnetic and geological investigations into a unified GIS system, it was necessary to establish a raster of fixed points for the whole site investigated. After the introduction of this geodetic net, relevant points of the newly established fixed point system were measured using a real time GPS to determine the parameters of transformation for the implementation into the Egypt Blue-Belt-system. Subsequently an up-to-date cadastral map was created on the basis of recent Satellite images (Fig. 2). This mapping of the modern surroundings of ancient Schedia was indispensable for several reasons. With regard to the planning of the geomagnetic surveying and the stratigraphic sondages, detailed knowledge of the modern topography (especially of the canals, streets and field borders) is of great importance. Furthermore, it is possible to draw some conclusions about the ancient landscape from its current shape. This holds true for the relief of the terrain, which reveals areas of settlement as well as the course of the probable harbour basin, but to a certain degree modern field

At present these results seem to suggest that the modern Kanubiye Canal follows the south-western embankment of the ancient Canopic Nile which may have passed between Kom el Giza and Kom el Hamam. In that case, its breadth amounted to 150 m. As there is also clear evidence that there was once a lagoon or a second branch of the river to the north of Kom el Giza, it can be assumed that the settlement of Kom el Giza actually was isolated on a kind of long and narrow island. Maybe it was this topographical situation which gave the stimulus to construct a pontoon bridge here. The investigations have also revealed a possible localisation of the Alexandrian Canal with a strict east-west orientation between Kom el Nashwa and Kom el Sherif and a 108

M. BERGMANN, ET AL.: SCHEDIA ON THE CANOPIC NILE

Fig. 2: Schedia. Topographical map based on quickbird satellite image (2005) indicating excavated areas (A) and corings (H). German Mission at Kom el Giza/Beheira. Fig. 3: Schedia. Map with reconstruction of the ancient topography. German Mission at Kom el Giza/ Beheira.

109

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig. 4. Schedia. Plan of the southeastern settlement (Kom el Giza, Kom el Hamam) with excavation areas 1980-92 and 2003-06. German Mission at Kom el Giza/Beheira. possible harbour basin immediately north of it. A second canal could have existed between Kom el Hamam and Kom el Sherif, branching at a right angle from the Canopic Nile and then turning in a westward direction. We thus appear to have a complex system of canals.

the architectural remains documented as a first step. Secondly, some stratigraphic trenches were excavated in order to learn more about the chronology and function of the structures. Furthermore, a series of new stratigraphic sondages have been carried out in previously untouched areas.

As for the extension of the ancient city, it is quite clear that the ancient settlement once occupied the whole area around the possible harbour basin (today’s Nashwa), the whole southern bank of the former Canopic Nile for a length of ca. 2.5 km (today’s Kom el Sherif, Kom el Hamam, el Karyun), as well as parts of the northern bank (today’s Kom el Giza). However it seems that the built up areas did not reach a significant depth but stuck rather close to the edges of the waterways, because of the necessities of a river harbour with its ongoing transhipment of goods. Thus, the picture that emerges is that of a rather amorphous and unhomogeneous settlement, consisting of a series of elongated narrow strips along the waterways, which also separated the different parts of the settlement from each other. Its nucleus has to be assumed in the vicinity of the harbour, from where it probably grew step-by-step without over-all planning along the river embankment. Unfortunately, the former kiman in the area of the harbour have suffered most from sebakh extraction, while we have to assume that the better preserved areas of the former settlement at Kom el Giza and Kom el Hamam were actually on the outskirts.

Summing up the results the following picture appears. In the western part of the older excavations (Area 3), there is a double-tholos-bath with bath-tubs to sit in (Fig. 5). It belongs to a Hellenistic type, which is well attested in the Nile Delta, the Fayum and other parts of Egypt, where it was used as a Hellenistic heritage much longer than anywhere else in the ancient world (see Boussac, et. al. forthcoming) A special feature of the Schedia bath is the visible division between the baths for men and women, otherwise attested in the contemporary papyri. Coin finds and the deep foundations suggest that the bath was built in Hellenistic times, but it is one of the rare examples whose long duration can be shown. Provided with additional rooms with bath tubs for immersion and with new high quality terrazzo floors, it was used at least until around AD 200 (Bergmann & Heinzelmann forthcoming). In its immediate vicinity, still in Area 3, the remains of a huge foundation are visible. Its dimensions seem to indicate that they belonged to a big public building, maybe a temple. Its facade was probably oriented westwards to the Canopic Nile. Unfortunately, however, this part is inaccessible because of modern constructions. Again all findings in this area, including the coins, indicate a (late?) Hellenistic date. Both buildings, the round baths and the possible temple foundation, consist of fired bricks, which is quite unusual for Ptolemaic architecture outside Alexandria.

Excavations at Kom el Giza One important aim of the field archaeological activities was to explore further the parts of the town found during the SCA’s rescue excavations in the 80s and early 90s (Fig. 4). Therefore, these areas were cleaned and all 110

M. BERGMANN, ET AL.: SCHEDIA ON THE CANOPIC NILE

Fig. 5: Kom el Giza, Area 3. Plan of Hellenistic baths. German Mission at Kom el Giza/Beheira. Fig. 6: Kom el Giza, Area 1. Plan of Roman villa, tombs and later basins. German Mission at Kom el Giza/Beheira.

111

LAKE MAREOTIS: RECONSTRUCTING THE PAST Fig. 7: Kom el Hamam. Plan with excavations areas 2003-2006. German Mission at Kom el Giza/Beheira.

Fig. 8: Kom el Hamam. Plan of Area 6 with Roman storage building and later phases. German Mission at Kom el Giza/Beheira.

112

M. BERGMANN, ET AL.: SCHEDIA ON THE CANOPIC NILE

Fig. 9: Kom el Hamam. Reconstruction of the Roman storage building (J. Schumann). German Mission at Kom el Giza/Beheira.

While these two structures of Area 3 seem to be in a certain sense public buildings, Areas 1-2 farther to the east reveal a completely different picture (Fig. 6). In both areas the earliest traces of use were numerous constructions that are possibly to be interpreted as funerary monuments dated on typological grounds to the Roman period. Especially one type of monument is very characteristic: a stepped pyramidal structure of ashlar blocks, probably with a pillar on it, which has close parallels in the necropoli of Alexandria and Marina el-Alamein (Daszewski 1998).

particular interest is a large building measuring 70 x 40 m, which is visible on the surface of Kom el Hamam. Three stratigraphic sondages were carried out down to the foundations of the building (Figs. 8 & 9). They show that it was constructed around AD 100 above earlier mud-brick structures, which were demolished for the purpose. The building was erected in an interesting building technique: the supporting structure consisted of 10 x 8 rows of massive brick pillars with cores of opus caementicium, which bore massive cross vaults of caementicium, collapsed fragments of which were found. It was an enormous fired brick building, particularly by Egyptian standards, that also demonstrates an in-depth technical knowledge of Roman buildings from central Italy of the period. However, two features indicate that the responsible architect was not really familiar with this building technique: contrary to the Roman prototypes, all pillars were constructed without foundations, instead of which huge mudbrick pylons were added already in the original phase of the building on the outside in order to absorb the pressure of the vaults. Furthermore, mud-brick walls running in a north-south direction were constructed between the pillars, which means that the building was separated into a series of nine parallel naves each ca. 5 m wide, 40 m long and 5.5 m high. All these naves opened on the south side through huge doors. The floor consisted of a simple mud-brick pavement. In the middle of the naves long low benches (ca. 0.25 m high, 1.50 m wide) made of mudbricks, seem to have been installed. Unfortunately, no specific finds were found that could give an indication of the original function of the building; obviously it was completely emptied before the changes of the second phase. However, judging by the typology of the building, it most probably formed a large storage facility for some kind of precious goods.

One of the less well preserved pyramidal structures has been investigated. Its core consisted of a cube of fired bricks. Around and underneath these bricks a stratum containing ashes, bones and pottery has been found. All bones belonged to animals of different types, such as at least four jaws of sheep as well as a complete skeleton of a young goat missing the head. They do not show signs of burning. However, a certain portion of the bones concentrated in one area was burnt and was in a very fragmentary state of conservation. They have yet to be analyzed, but it seems likely that they belong to a human cremation, while the unburnt animal bones might be connected to some funeral rites. Because of a Trajanic coin found within the ash layers this structure probably belongs to the 2nd century AD. The area must be regarded as suburban cemetery. At a somewhat later time, an isolated villa was built apparently in the midst of the sparse graves of Area 1 (Fig. 6). This villa of the middle Imperial period, was organised around a central atrium-like courtyard and was furnished with mosaics and opus sectile floors. It was in use for a long time and repeatedly rebuilt and renovated. During the 4th century AD the villa was abandoned, and numerous basins were installed over its ruins and the neighbouring graves, as well as in most of the other areas excavated up to now. Undoubtedly, these were production units that were used intensively and were repaired repeatedly up to the 6th century.

Later, the building underwent several different phases of use and renovation. Only a short time after its erection, additional mud-brick walls were added during the later 2nd and 3rd centuries AD in order to create simple apartments. These habitations show two main phases of use. In one of the sectors several rooms were examined, some of them containing cooking installations. They revealed a huge amount of pottery, mainly amphorae, as well as coarse and

Excavations at Kom el Hamam New archeological examinations were carried out at Kom el Hamam and other parts of the ancient city (Fig. 7). Of 113

LAKE MAREOTIS: RECONSTRUCTING THE PAST Fig. 10: Kom el Hamam, Area 9. Roman enclosure wall. German Mission at Kom el Giza/Beheira.

cooking wares, mostly of local production. In the late 4th or early 5th century AD these habitations were abandoned and levelled in order to install a granary. On a level ca. 1 m above the pavements of the prior habitations, a special brick floor was constructed supported by some dozens of small parallel vaulted chambers – a typical Roman system for keeping grain dry. This grain storage facility was in use until the late 5th or 6th centuries AD, when it was destroyed by fire. On its ruins simple dwellings and agrarian installations were installed, which were in use until the Arab conquest.

Once again in the 4th century, a radical change took place. As in the case of the villa suburbana at Kom el Giza, most of the earlier buildings at Kom el Hamam were destroyed or changed their function. Instead of the dwellings or monumental buildings, a whole series of simple working installations occur. Among these are several types of basins, which seem to have served various purposes. Some have the characteristic arrangement for wine production with a higher platform for treading and a lower catchment basin accompanied by further basins for fermentation (Fig. 11) (Rodziewicz 1988). Others are isolated or have furnaces nearby and appear to have served other productive processes that cannot be identified at present. Aside from their function, the basins testify to an intensive agricultural use of the territory in late antiquity. These agricultural installations were in use until the beginning of the 7th century, when Schedia was completely abandoned at the time of the Arab invasion.

Immediately south of this huge building, the remains of another monumental structure was uncovered (Figs. 7 & 10). At a distance of ca. 4.5 m a huge wall runs parallel to the southern front of the above-mentioned storage building. The wall has impressive dimensions: with a thickness of about 2 m its foundations consist of a massive foundation of opus caementicium and three layers of huge ashlar blocks, each with a height of 0.45 m, while the elevation consisted of two more layers of ashlar blocks and a massive mud-brick wall on top. The stratigraphic evidence shows that the wall was constructed shortly before the huge storage building, probably in the first half of the 1st century AD. Its function is not yet clear. The dimensions as well as the massive construction technique indicate that the wall must originally have reached an impressive height. Therefore, it seems probable that the wall actually formed a high enclosure belonging maybe to a temple or some other kind of public building. Further investigations showed that the wall actually belonged to a much bigger structure, consisting of at least two rectangular courts. Strangely enough the northern half of it was abandoned shortly after its construction and substituted by the huge pillar building, while the southern part was continuously in use until late antiquity.

Analysis of the Finds The detailed analysis of our excavated material, especially coins and pottery, will give further information in the future concerning Schedia’s economic development. However, because of the huge amount of finds we are just at the beginning. Coins Around 4,000 coins found at the ancient site of Schedia have been analyzed by Ch. Noeske. Among them are 1,400 stray and stratified finds from the recent excavations. On the whole they cover the period from ca. 320 BC (Ptolemy as a satrap) to the time of Arab conquest (Heraclius, AD 641). Hellenistic coins from the SCA excavations at Kom el Giza are numerous, while in the excavations on the northern side of the Kanubiye Canal (Kom el Hamam, Kom el Sherif) the Roman Imperial and late-antique coins 114

M. BERGMANN, ET AL.: SCHEDIA ON THE CANOPIC NILE prevail. The analysis of these coins has different aims. The dates they offer for the stratigraphic units are important, but they are also analyzed in their own right, as sources for minting policy, monetary circulation and its regions, and for economic change. Comparison to the few other places where coins have been noted, such as Elephantine, Abu Mina and parts of Alexandria, is informative in this respect.

phorae make up approximately 2/3 or 3/4 of the assemblages, which is a characteristic percentage for Mediterranean sites that engage actively in commercial exchange, while in the latter amphorae are attested at about 1/3. Consideration of some of the specific functional groups confirms the difference between the earlier and the later layers. The characteristics of the fine wares and amphorae vary markedly between the two periods. There are so few lamps that it is not possible to speak of trends in them. The utilitarian wares (coarse and cooking wares, which at Schedia appear in the same fabric) are essentially of local production throughout, and their typological development remains to be determined.

Pottery At Schedia the programme of pottery study aims at as complete a documentation as possible. The protocol used is predicated on the idea that all fragments, even typologically unidentifiable body sherds, have some information to give. Therefore, all the sherds per stratigraphic unit are attributed to a functional group and fabric, counted and weighed. For rim sherds the percentage of the diameter preserved is also registered, in order to calculate estimated vessel equivalents. Diagnostic fragments are, of course, classified typologically, and chronologically sensitive items are used for dating purposes. This degree of quantification permits our material to be compared with other assemblages throughout the Roman world, where various standards prevail. It also allows us to make statistical considerations at various levels (Martin 2008; Martin forthcoming).

Among the fine wares in the earlier layers it is noticeable that Schedia did not participate in the fine-ware boom of the Augustan period, when most parts of the Roman world wanted to import or produce red-gloss wares reflecting the repertoire of Italian sigillata, that is to present themselves as Roman at the table (Martin 2008: 268). There are some very occasional imports of Eastern Sigillata A and less frequently yet of Italian sigillata, but for the most part Schedia seems to continue to employ red and black-slipped wares in the Hellenistic tradition. In the late-antique layers a certain quantity of Cypriot and also of African Red-Slip Ware appears, and a larger amount of Egyptian products inspired by them is attested. It is only then that one can say that Schedia participates in the Mediterranean-wide trends in fine wares.

A first level concerns the overall composition of the assemblages by functional groups. There is some evidence that the percentage in particular of transport amphorae with respect to fine table wares, lamps, cooking and coarse wares, reflects the degree of integration of a site into trade networks. Here Schedia shows a decided difference between the late-antique layers up to the abandonment no earlier than the 7th century and the earlier ones. In the former, am-

In the earlier levels Egyptian amphorae greatly outnumber imported ones – the majority were the Amphore Egyptienne 3 type which were first developed at the beginning of the Empire. Other examples included the earlier Amphore

Fig. 11: Kom el Hamam. Plan of Area 8 with basins for wine-production. German Mission at Kom el Giza/Beheira.

115

LAKE MAREOTIS: RECONSTRUCTING THE PAST Egyptienne 2 and the contemporary Amphore Egyptienne 4 types (Empereur 1998: 79). They appear most frequently in the same fabric as the utilitarian vessels, decidedly more rarely in the northwest coastal fabric associated with production around Maryut, and only occasionally in the Nile silt fabric. Imported amphorae are represented by a diffuse scatter of various Aegean and Levantine pieces. In late antiquity, on the other hand, the provenience of the amphorae is much more focused and the percentage of imported vessels much higher. Egyptian vessels appear in the Nile silt fabric associated with Egloff 172 and the Carthage Late Roman Amphora 7, while the imports come almost exclusively from Cilicia with the Carthage Late Roman Amphora 1 and Gaza with the Carthage Late Roman Amphora 4.

During the 2nd and 3rd centuries the city reaches its maximum extension. That the brick-pillar building investigated here was transformed for residential purposes contrary to its original destination, can be considered a further indication of population pressure and of an increasing process of urbanization. Parts of the city reveal a strong Greek tradition (the Hellenistic bath and the stepped bases of grave monuments) and Roman influence (villa suburbana). Other monuments, like the pillar building with its contemporary use of fired and mud-bricks show an interesting mixture of imported and local building traditions. At the same time, Schedia appears to have been a city with interesting cultural contrasts: while a prosperous upper class lived in houses of clearly Roman influence, a large number of inhabitants lived in simple dwellings of local Egyptian character.

In short, Schedia in the Imperial period presents, paradoxically for the major river harbour of Alexandria of which the written sources inform us, the picture of a site little integrated into the exchange networks and cultural trends of the time, while late-antique Schedia appears to participate more actively in them. How to explain this situation remains to be seen. Could it be that in the earlier period little needed to be offered in exchange for Egyptian products (e.g. grain collected in tax, stone from the Imperial quarries), while late-antique Egypt was in a similar situation to other provinces, with imports and exports. In any case, merit is due to the pottery study that serves to highlight a line of research to be pursued that might not have been highlighted in any other circumstance.

In late antiquity the explored parts of the city underwent a fundamental change. They seem to lose a large part of their urban character with the transformation of the older buildings into productive installations, of rural and other determination. At the same time the new installation of a granary in the pillar building coincides with the undiminished export of grain to Constantinople. How these changes are to be interpreted against the background of new discussions about the flourishing agriculture and renewed trade and imports of late antique Egypt (Banaji 2001; Kingsley & Decker 2001) and against the situation in Alexandria itself, remains to be seen. Although the new transformation of the pillar building into a large granary points to the still uninterrupted importance of the city as a supply base, it changes into a centre for agricultural production that possibly still has regional importance in supplying Alexandria but appears to a large extent to lose its role as an emporium, perhaps because of the overall decrease in trade. This hypothesis, which for now can take only the late Ptolemaic and Roman settlement of Schedia into account, needs to be checked by further investigation and especially to be complemented by a greater consideration of the Ptolemaic phases.

Conclusions The following preliminary hypotheses concerning the development of Schedia seem possible. After the foundation of Alexandria and the construction of the Canopic Canal the early Ptolemaic settlement of Schedia existed probably around an artificial harbour basin at the starting point of the canal. Because of the artificial bridge, the ‘schedia’, and a very dynamic development, the settlement, extended on both sides of the Canopic Nile towards the southeast. Here, on its southern borders, monumental and public buildings (baths and temples) were erected not later than middle and late Ptolemaic times. The settlement, therefore, already had urban character, which is also documented by the inscriptions.

Bibliography Abd el Fattah, A., 2000-2001, Brief report on excavations carried on at Kom el Giza during the season 1989. Annales du Service des Antiquités de l’Égypte 76: 9-12. Abd el Fattah, A., 1998, Quelques nouveaux moules alexandrins à Kom Giza. In A. Abd el-Abd el Fattah & P. Gallo (eds.), Aegyptiaca Alexandrina. Monuments pharaoniques découverts récemment à Alexandrie 1: 65-73. Alexandrina. Abd el Fattah, A., 1988, Recent discoveries in Alexandria and the Chora. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athènes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole française d’Athènes, 11-12 décembre 1988: 47-53. Athens, Paris.

In the course of the Late Hellenistic and Early Roman periods a substantial necropolis with monumental tombs developed along the southeastern outskirts of the city, but because of the continuous growth of Schedia, luxurious residences for a prosperous upper class soon penetrated the periphery previously used for burials. At the same time on the southern embankment a monumental building, maybe a temple complex, as well as large storage facilities like the pillar building, arise, apparently as a consequence of massive investments. On one hand, they attest to the economic attractiveness of Schedia in the Imperial period, and, on the other, their distance from the harbour suggests that the more favourable zones were already built up. 116

M. BERGMANN, ET AL.: SCHEDIA ON THE CANOPIC NILE Banaji, J., 2001, Agrarian Change in Late Antiquity. Gold, Labour and Aristocratic Dominance. Oxford. Bergmann, B., & Heinzelmann, M., 2007, Schedia, Alexandrias Hafen am kanopischen Nil. Zwischenbericht zu den Arbeiten 2003-2007. Hefte des Archäologischen Seminars Bern 20: 65-77. Bergmann, B., & Heinzelmann, M., forthcoming, The tholos-bath at Schedia. In M.F. Boussac, T. Fournet & B. Redon (eds.), Le bain collectif en Egypte. Des balaneia antiques aux hamams contemporains: origine, évolution et actualité des pratiques. Actes du colloque Balnéorient d’Alexandrie 1-4 Décembre 2006. Alexandria. Bernand, A., 1966, Alexandrie la Grande. Paris. Bernand, A., 1970a, Alexandrian Canal. Le delta égyptien d’après les textes grecs I. Les confins libyques: 329-380. Cairo. Bernand, A., 1970b, Schedia/Menelais. Le delta égyptien d’après les textes grecs I. Les confins libyques: 381-442. Cairo. Botti, D.G., 1902, Studio sul III° nomo dell’Egitto inferiore. Bulletin de la Société d’Archéologique d’Alexandrie 4: 41-61. Boussac, M.F., Fournet, T., & Redon, B. (eds.), forthcoming, Le bain collectif en Egypte. Des balaneia antiques aux hamams contemporains: origine, évolution et actualité des pratiques. Actes du colloque Balnéorient d’Alexandrie 1-4 Décembre 2006. Alexandria. Daszewski, A.W., 1998, La nécropole de Marina el-Alamein. In S. Marchgay, M.-Th. Le Dinahet & J.-F. Salles (eds.), Nécropoles et pouvoir: idéologies, pratiques et interprétations. Proceedings of a conference Lyon 1995: 229-241. Lyon. Empereur, J.-Y., 1998, Les ateliers d’amphores du lac Mariout. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athènes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole

française d’Athènes, 11-12 décembre 1988: 75-91. Athens, Paris. Jomard, E.F., & Jacotin, P., 1818/25, Carte Topographique De L’Égypte Et De Plusieurs Parties Des Pays Limitrophes (1818/25), M 1: 100 000, Blatt Nr. 37. Kingsley, S., & Decker, M., 2001, New Rome, new theories on inter-regional exchange. In S. Kingsley & M. Decker (eds.), Economy and Exchange in the East Mediterranean during Late Antiquity. Proceedings of a conference at Somerville College. Oxford 29th May 1999: 1-27. Oxford. Martin, A., 2008, Pottery from Schedia near Alexandria (Egypt). Rei Cretariae Romanae Fautorum Acta 40: 263-269. Martin, A., forthcoming, The pottery from a late-antique settlement at Schedia (Western Delta, Egypt). In S. Menchelli, M. Pasquinucci & S. Santoro (eds.), The 3rd International Conference on Late Roman Coarse Ware, Cooking Ware and Amphorae in the Mediterranean: Archaeology and Archaeometry (LRCW3) (Parma – Pisa, 25-30 March 2008). Oxford. Procopius, De Aedificiis, Vol. VII, H.B. Dewing (transl. & ed.), 1961. Cambridge, MA. RE 1899, Wissowa, G. (ed.), Paulys Real-Encyclopädie der classischen Altertumswissenschaften, Vol. I 1, 2030 s.v. Χαιρέου (Sethe). Stuttgart. RE 1921, Wissowa, G., Kroll, W., &Witte, K. (eds.), Paulys Real-Encyclopädie der Classischen Altertumswissen-schaft, Vol. II A 2, 401-403 s.v. Schedia (Kees). Stuttgart. Rodziewicz, M.D., 1988, Classification of wineries from Mareotis. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athènes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole française d’Athènes, 11-12 décembre 1988: 27-36. Athens, Paris. Strabo, Geography, Vol. VIII, H.L. Jones (transl. & ed.), 1959. Cambridge, MA.

117

LAKE MAREOTIS: RECONSTRUCTING THE PAST

118

P. WILSON: SURVEY WORK IN MAREOTIS

Recent Survey Work in the Southern Mareotis Area Penelope Wilson

Introduction The ‘Sais and its Hinterland Project’ based at Sa el-Hagar, Sais in Gharbiyah Province, was designed partly to determine the impact of river distributary systems on settlement in the north and north-west of Egypt. In order to compare the Canopic with the Rosetta riverine systems, 70 sites were surveyed and visited in Beheira and Kafr esh-Sheikh Provinces. This paper describes and discusses sites in the southern Mareotis area, which formed an interesting group in themselves. They represent places influenced by another type of water-system, an inland lake located at sea-level, that is fed by a series of distributaries from the Canopic system. The location and extent of Lake Mareotis fluctuated over time and this may have been an influencing factor in the life-span recorded at some of the sites in the southern Mareotis area.

sides. The lower parts of its sides have been left largely untouched, but are used for animal husbandry. The village of Kom Ishu was formerly on the western, southern and eastern sides of the hill but has developed in linear streets radiating to the east. It is possible that the original heart of the village was sheltered from the north wind by the hill. The field boundaries beyond the village suggest that the kom may have covered a larger area at one time. The whole of the upper surface of the hill is now covered by a modern cemetery. Pottery gathered at the site is mostly of Late Roman date.

Kom Ishu and Kom el-Hagg The two sites share a similar geological nature as they both consist of rocky limestone outcrops at the edge of a limestone ridge formation and seem to have been the focus of human activity in Lake Mareotis (Embabi 2004: 260-2).

Kom el-Hagg (SCA 100155) is surrounded by a small ezbet, called Ezbet el-Bank. The mound has a sloping, rounded edge to the east and the other sides are slightly steeper at the edges. The area is used for stabling animals and as a manure store. The southern edge of the kom has been cut away, leaving a metre-high section above field level containing pottery fragments. The fields to the south of the main mound are covered in sherds and may have been part of the ancient site, perhaps representing a settlement against the rocky outcrop. It is also possible that the sherds to the south were washed or swept down off the rocky hill. It would be interesting, therefore, to define the exact nature of the site at this point because of its location and the apparent use of the local geological features as at Kom Ishu and also at Kom Lemsan, situated to the north.

Kom Ishu is approximately 12 km west-south-west of Kafr ad Dawar. The rocky outcrop has steep, almost sheer,

The pottery sample from the site gave a broadly consistent date range from the Ptolemaic through to the Early Roman

The Sites (Figs. 1-2) The sites are discussed in an anti-clockwise direction around Lake Mareotis. Although the pottery collection and survey are of a preliminary nature, some interesting observations can be made arising from a study of the results.

Fig. 1: Ancient sites in the Delta, highlighting the Mareotis region (drawn by P. Wilson).

119

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig. 2: Map of the sites surveyed, with 0 metre sea-level contour area highlighted in grey. By R. Dickinson, after 1:50,000 Survey of Egypt maps (1997).

120

P. WILSON: SURVEY WORK IN MAREOTIS periods. The different types of vessel from the site included both amphorae and domestic vessels and so did not indicate a particular function for the site, but there certainly seems to have been some kind of settlement here. Most of the pottery was Egyptian in origin, and presumably was produced locally, with only one of the amphorae of Early Ptolemaic date perhaps from Rhodes, possibly a reused wine or water container. Both Kom el-Hagg and Kom Ishu are situated on the western edge of the main body of Lake Mareotis and could have had strategic value in guarding the lake shores or in watching over the desert areas to the west, the river to the south and marshes or fields on all sides (Bernand 1970: 866). Although the two places are within signalling distance, the pottery represents two distinct phases, with that from Kom el-Hagg covering the Ptolemaic and Early Roman periods, while that at Kom Ishu is of the Later Roman period. There may be a number of reasons for this, mostly due to the nature of the collection of the pottery sample. But it could be that Kom Ishu replaced Kom el-Hagg as a watching place or simply that it continued for a longer period of time, perhaps because it was situated upon a more long-lived canal which later became the Tirat Abis. In such a case, Kom Ishu had a better strategic value in the later period because of its water channel communication.

Fig. 3: The main mound at Kom el-Mahar (photograph by P. Wilson). 2nd century. Most of the pottery sample consists of tablewares, with some imported finewares and glass, and may have come from domestic contexts. Kom el-Mahar could have been a small town or village focussed on a series of elite houses and villas on the southern shore of the lake, taking advantage of the elevated land (Botti 1902: 55).

Kom el-Mahar and Sidi Ghazi Kom el-Mahar (SCA 100167) lies isolated in farmland to the east of the Nubariyah Canal and south-west of Sidi Ghazi village. The main mound is covered in pottery, including finewares (Fig. 3). The name of the site in Arabic means ‘shell’ and shelly deposits are apparent all over the mound and in the sections. The shell derives from the fact that the mud-brick used to build the houses of the town was made from silt mixed with crushed shell from degraded, local oolitic limestone. When the mud-brick disintegrated, a large amount of shell was left lying upon the surface.

The modern village of Sidi Ghazi (SCA 100142) lies approximately 10 km south of Kafr ad Dawar and to the north of the Masraf al-Umum Canal which runs to the west of the Baslaqun-Luqin group of tells. There is now no tell at the location of Sidi Ghazi, but modern maps indicate that the village lies upon a small area of raised ground. It is likely that the original tell has been subsumed under the village. A site here may have formed one of the islands inside Lake Mareotis. South-western Sites Kom el-Gel (SCA 100217) is the site at Ezbet Abd el-Qadir at-Tawil and lies approximately 17 km north-west of Abu Matamir. The site consists of a mound with a flat top and gently sloping northern side and a secondary mound to the east. The top of the mound is covered with a modern cemetery, while the sides are used for animal husbandry. The houses on the western side of the tell are built on flat land at the base of the hill. The fields to the south of the mound contain noticeable quantities of pottery and some limestone fragments, suggesting that this was once a more sizeable mound. Satellite imagery suggests that the area to the south may have been part of the mound because of the shape of the fields.

The lower slopes of the eastern and south-eastern side of the mound are covered by a modern cemetery. The northern and north-western side of the mound has sheer sides, where they have been dug away and it is likely that the original site extended in all directions away from the mound. It is now surrounded by deep irrigation ditches and fields, which are encroaching on the outer parts of the site. One of the northern sections contains a substantial, stratified sequence, including ashlar limestone blocks, a wall made of sandy mud-bricks and shell-filled, mudbricks. There are also fragments of red granite on the mound and glass fragments lying upon the surface of the site. The southern area of the site has been flattened and there are building outlines visible on the surface. There are some lower, smaller mounds covered in scrub to the northwest and south-east sides of the site.

The pottery sherds from the site comprised fine wares, domestic material and amphorae. The pottery dated mainly to the Late Ptolemaic period, with some Early and Late Roman sherds included in the collection.

The main pottery component from the site ranges from Late Ptolemaic to Early Roman in date (Fig. 4), suggesting that it may be a Ptolemaic foundation continuing as a town into the Roman period, before being abandoned around the

The original site at Kom el-Farag lies under a modern town, approximately 13 km north-west of Abu Matamir. It may be the location of Kom el-Hanache (Toussoun 1922: 121

LAKE MAREOTIS: RECONSTRUCTING THE PAST Fig. 4: Late Ptolemaic to Early Roman pottery sample from Kom el-Mahar (drawings and analysis by D. Grigoropoulos).

lake next to the village which may be a relic of an ancient river distributary or ox-bow lake, extending for about 500 m in a north-west to south-east orientation. The lake may link with some ancient depressions in the landscape to the west or have formed one of the distributary channels from the Canopic Branch. If there was once an ancient settlement here, it may have been located upon the river, lying along a raised levee as with other tells in this part of Beheira.

Pl.II), but as the Ezbet Farag grew to the south, the tell seems to have been subsumed underneath the houses and removed by the Delta Light Railway. The cemetery part of Kom el-Farag lies to the north-west and may be a remnant of the older mound, being a maximum of 1.5 m above ground level. The old part of the village upon the mound has a less organised street plan than the more recent areas of the town with their orthogonal layout. The interest of this location is that it lies on the Tirat Abdel al-Fatah and Tirat al-Hajar drainage channels and, perhaps, on the channel of one of the earlier Canopic distributaries on the western side of Beheira. Kom Farag is also at the centre of a cluster of small villages and ezbets with the word kom in their name, such as Kom es-Siwayd to the south, Kom Kifri, as Sa’ayidah and Jaradah to the south-west and to the north-west Ezbet Abd el-Qadir atTawil (Kom Makhboura), which has a cemetery upon a kom. They may well be part of a town cluster focussed on an irrigation basin network dating at least to the medieval period.

Abu Guduur (SCA 100212) is 2.5 km north-west of Abu Matamir and lies to the south of the main road, upon a low ridge of silt along an ancient distributary course, at the edge of the Mareotis basin. It once covered an extensive area, but now consists of some low, sandy mounds covered in scrub and a larger flat area with scrub and bushes. Most of the site is now used as a municipal rubbish dump, having once also been a gravel extraction and processing area. The highest mound is only about 2 m above the level of the surrounding land and the whole area is very sandy. There was pottery lying on the tell as well as some bronze objects, including a stirrup ring and a three-flanged arrowhead. Both could date to the Late Dynastic period. A small diagnostic set of pottery fragments were collected from the site. As far as can be judged, they mostly date to the Late Ptolemaic period, but there were some Late Roman sherds and some Late Dynastic pottery as well.

South of the Lake Kom el-Adda (SCA 100230) or Kom Abu El-Eida lies 8 km north-west of Abu Matamir. Although there is a cemetery mound to the south-east of the main village, the main interest of the area is the curious crescent-shaped, enclosed 122

P. WILSON: SURVEY WORK IN MAREOTIS The place may have been a strategic Dynastic location for a small military establishment related to the other western delta fortified sites and to other mounds running from Abu Matamir to the south east. Southern Port on the Naukratis Canal Kom Trugi or Turuga (SCA 100251) lies 6 km due north of Abu Matamir, close to several other places with kom in their name and on the former edge of Lake Mareotis. Turuga is known to have been an important inland port on the lake for people going from Alexandria to the Wadi elNatrun and for goods and supplies coming out of the Western Desert and delta towards Alexandria for trade (Timm 1984-1992: 2545-6; De Cosson 1935: 79, 151). It is possible that the site had two halves on either side of the waterway into the lake, one known as Psenemphaia, according to a stela from the site (Bernand 1970: 899-913) and the other known as Theroge or Therange, which lends its name to the whole site after the Ptolemaic period. Roman period material was noted at the site in 1895 by D.G. Hogarth (Spencer 1999: 305, 308-11) and Botti noted that marble column capitals could been seen there and that a headless white marble statue was found at Kom Trugi (Botti 1902: 58). Kom Trugi suffered greatly from sebakhin digging (De Cosson 1936), with the authorities only intervening at the discovery of important objects or remains such as a Ptolemaic-Roman bath-house (El-Khashab 1957: 127-139).

Fig. 5: Red brick structures at Kom Trugi, excavated below foundation level. The scale is 2 metres high (photograph by P. Wilson). Fig. 6: Section through the cemetery mound at Kom elQadi (photograph by P. Wilson).

It is clear that this was once an extensive site and perhaps a high mound. It now consists of flat, sandy areas interspersed with several excavated red brick structures and limestone upon mounds. To the north of the area, there is a modern village with an adjoining cemetery and both may have been built upon part of the original site. Amongst the extant red brick structures there is part of a pavement left standing upon its sand-brick foundation and a pillar of earth, as well as red brick and plastered tanks, perhaps from a winery. There is also a fragment of a limestone wall consisting of five courses of ashlar blocks for a length of 4.75 m, with some plaster still adhering to one side (Fig. 5). All of this area seems to have been dug out to below the floor level of these buildings and the site seems very denuded. On one of the mounds to the south there are the foundation walls of a limestone building. The flat, sandy area between the mounds shows building plans in wet weather, according to the local people, but is now mostly used as a football pitch. There are also a number of red granite grindstones at the site, lying upon the ground. Although it is difficult to make sense of the standing remains, they do suggest that an impressive town with large red brick and limestone buildings once stood here.

the medieval period. The visible levels date perhaps to the Early and Middle Roman periods and information about the later levels may be lost. The height of the mound suggests that there is still buried material at the site, but that it is disappearing fast. North-east of the Lake Kom el-Qadi (SCA 100165), south of the main Cairo to Alexandria road, and Kafr ad Dawar may be part of a cluster of sites around Al Baslaqun, forming a township or irrigation basin group of sites. The sprawling area at Kom el-Qadi has several distinct zones and a main archaeological area to the south of Kom el-Qadi village. It consists of a high mound, around 8 m high, the southern and northern faces having been cut away to form sheer sections. The southern sections show significant stratified sequences, including sand-brick walls, pottery deposits and human burials in both pottery coffins and brick vaults. Skulls of adults and children can be seen in the sections and on top of the mound (Fig. 6). The remains may be interpreted as

The pottery sample collected from the site dates mainly from the Early Roman period, but there are also some Ptolemaic, Middle-Late Roman and Medieval Arab period sherds in the collected sample. Because of the disturbed nature of the site, however, it is difficult to associate particular pottery with specific areas of the site. The range of the pottery agrees well with the historical material, suggesting that the site was used from the Ptolemaic through to 123

LAKE MAREOTIS: RECONSTRUCTING THE PAST Summary (Table 1) Lake Mareotis created its own special topography and landscape, with the shallow flooded area bordered by reeds and rushes forcing people onto the high ground around the lake edge, islands in the lake and between the distributaries to the south and the Canopic and Alexandria Canal to the north. The strange, thin tongues of land, with their dykes and paths, must have made the desert routes seem attractive to travellers and so created opportunities on the Western Desert edges for monastic hostels and stopping places. The topography of the area shows that the presence of the lake had a profound impact upon social organisation and economic management in the North-West Delta. Pharaonic period attempts to control the Western Desert edges may always have led to limited settlement in the north-west area, as suggested by Abu Guduur and the fortress line in the Ramesside period (Habachi 1980). The foundation of Alexandria seems to have acted like a magnet, attracting service settlements and industries upon a huge scale, as well as enabling the development of agricultural lands around the city and bringing settlement in favourable locations. The preliminary dating evidence from the sites in Beheira shows that the earliest material at the surveyed sites in the majority of cases comes from the Ptolemaic period, contemporary with the development of Alexandria (Wilson & Grigoropoulos 2009: 42-43). The sites at Kom el-Mahar, Kom el-Gel and Kom Trugi may all be part of this wave of development, with the creation of farming towns, a process which continued into the Early Roman period. This seems to have been a busy time for the lake, with many settlements established either as internal trans-shipping ports for the movement of goods and people

a reuse of the abandoned town mound as a cemetery. This may be the Roman cemetery from which three inscriptions were published (Botti 1902: 57). On the eastern side of this mound a substantial mud-brick wall is visible in the section. In the village to the west, there is a Roman bath-house, which has been restored recently and includes part of the concrete floor of a tholos, some possible underfloor furnaces and rectangular tanks. More uncovered red brick chambers as well as surface building plans, can be seen on a sandy area to the east of the village and main mound. A further sandy area lying to the east of this part of the village and a patch of extensive low scrub to the north-east of the village may have been part of the original site. There were some larger fragments of limestone lying in the mounds and the village. The pottery collected was mostly from the mound. It consisted of Early and Middle Roman material, including some imported fine wares, many cooking pots which showed evidence of burning on the outside and some Egyptian amphorae. The material seems to indicate the domestic nature of the site in the Roman period, but the relationship of the cemetery remains to the settlements was not clear from a preliminary examination of the mound sections. The site seems to be a small town and its associated cemetery, perhaps dating from the height of Roman management of this area. Table 1: Summary table of sites surveyed by the ‘Sais and its Hinterland Project’ in the region of southern Lake Mareotis. Site & current dimensions

Early Roman Ptolemaic Early Late 1st c. BC rd st 3 -late 1 c. BC to 2nd c. AD

Middle Roman Late Roman 3rd-4th c. AD 5th-end 7th c. AD

Kom Ishu 150 x 200 x 10 m

Early medieval Arab 8th-16th c. AD

X

FUNCTION Watch post

Kom el-Hagg 120 x 200 x 4 m

X

X

Watch post

Kom el-Mahar 440 x 240 x 12 m

X

X

Main Ptol-Roman town and villa site

Sidi Ghazi Under town Kom el-Gel 270 x 220 x 4 m

Island village X late

X

Town/villa

Kom el-Farag Under town

Farming village

Kom el-Adda 220 x 300 x 3 m

Riverside town

Abu Guduur 300 x 250 x 2 m

X late

Kom Trugi 350 x 300 x 3 m

X

X late X

X

Kom Hassan Under fields Kom el-Qadi 370 x 250 x 8 m

Fort?, riverside town X

Inland harbour town Unclear

X

X

Main Roman town

124

P. WILSON: SURVEY WORK IN MAREOTIS Bibliography SCA numbers refer to site registration numbers of the Supreme Council of Antiquities, Egypt.

to and from Alexandria, or as points of control around the lake, such as at Kom Ishu, Kom el-Hagg and perhaps Sidi Ghazi. The apparent abandonment of places, such as Kom el-Mahar and Kom el-Gel, may be part of the ‘Antonine Plague’ effect with depopulation of some areas (Bagnall & Frier 1994: 173-8). In the later periods, the lack of Late Roman and Early Arab continuity at the sites suggests that, in common with other parts of Beheira province, once the Canopic Branch of the Nile and its distributaries had silted up, the sustainability of some of the settlements was not possible and that people moved away from earlier settled land. They may have gone to a few major centres such as Abu Matamir or Dilingat or Damanhur or Alexandria itself, rather than carrying on trying to eke out an existence on the shores of the Mareotis Lake.

Bagnall, R.S., & Frier, B.W., 1994, The Demography of Roman Egypt. Cambridge. Bernand, A., 1970, Le Delta égyptien d’après les textes grecques, I-III. Cairo. Botti, G., 1902, Studio sul IIIo nomo dell’egitto inferiore: e più specialmente sulla Regione Mareotica. Bulletin de la societé royale d’archéologie - Alexandrie 4: 4184. De Cosson, A., 1935, Mareotis: Being an Account of the History and Ancient Monuments of the North-West Desert of Egypt and Lake Mareotis. London. De Cosson, A., 1936, El Barnugi, Bulletin de la societé royale d’archaeologie - Alexandrie 31: 113-116. Daressy, G., 1929, Ménélais et l’embouchure de la branche Canopique. Revue de l’Égypte ancienne 2: 20-51, Pls. I-II. Embabi, N.S., 2004, The Geomorphology of Egypt. Landforms and Evolution. Vol. I. The Nile Valley and the Western Desert. Cairo. El-Khachab, A.M., 1957, Les Hammams du Kom Trougah. Annale du Service des Antiquités de l’ Egypte 54: 117-139. Habachi, L., 1980, The Military Posts of Ramesses II on the Coastal Road and the Western Part of the Delta. Bulletin de l’Institut français d’archéologie orientale 80: 13-30. Spencer, P., 1999, Hogarth’s 1895 Report to the Committee of the Egypt Exploration Fund. In W.J. Tait & A. Leahy (eds.), Studies on Ancient Egypt in Honour of H.S. Smith: 302-311. London. Survey of Egypt 1997. 1:50,000 Maps; “Abu al-Matamir”, “Kafr ad-Dawwar”, “Ikinij Maryut” & “al-Iskan dariyyah”. Survey of Egypt/Ministry of Finances, Cairo. Timm, S., 1984-1992, Das christlich-koptische Ägypten in arabischer Zeit: Eine Sammlung christlicher Syätten in Ägypten in arabischer Zeit, 6 Vols. Wiesbaden. Toussoun, O., 1922, Mémoire sur les anciennes branches du Nil. Mémoires de l’Institut d’Égypte 4, Premier fascicle, Cairo Wilson, P., & Grigoropolous, D., 2009, The West Nile Delta Regional Survey, Beheira and Kafr esh-Sheikh Provinces. London.

Of the sites surveyed here, Kom Trugi was the largest and most important, perhaps straddling the distributary of the Canopic Branch from Naukratis and providing a secondary direct link from the south to Alexandria and from the Western Delta area to Alexandria. Kom el-Qadi seems to have been a typical Early Roman town with a bath-house, similar to other new towns in the Roman delta, and a substantial mound used as a cemetery. It may have been founded in the Early Roman period at around the same time as Schedia and part of the cluster around BaslaqunLuqin, once suggested to be the location of Menelais (Daressy 1929). Further detailed work in any or all of the Mareotis catchment area would be highly desirable to explore riverine, lacustrine and human interaction during periods of profound cultural change. Acknowledgements I thank Dr Dimitrios Grigoropoulos of the Deutsches Archäologisches Institut – Athens who studied the pottery from the sites and Dr Roger Dickinson for the mapping. I am also grateful to the Supreme Council of Antiquities in Damanhur under Fawzy Fawzy el-Kholi and Naama Suleiman, as well as Zahi Hawass and Magdy Ghandour in Cairo. For the fieldwork, I thank Dr Joanne Rowland, Mohamed Osman, Ashraf abd el Rahman and Ahmed Bilal. The Department of Archaeology, Durham University and Egypt Exploration Society also supported the work. Funding from the Arts and Humanities Research Council, UK is acknowledged. The survey is now published by the Egypt Exploration Society’s Delta Survey Project under the title The West Nile Delta Regional Survey, Beheira and Kafr esh Sheikh Provinces.

125

LAKE MAREOTIS: RECONSTRUCTING THE PAST

126

D. DZIERZBICKA: WINERIES OF THE MAREOTIC REGION

Wineries of the Mareotic Region Dorota Dzierzbicka

Wineries or remains of wine making installations have been found in various parts of the Nile Valley: the coastal areas and the Delta, the Fayum and Bahariya Oasis, as well as Upper Egypt (Dzierzbicka 2005: 11-24). However, their concentration in the Mareotic region, especially along the southern shore of the lake and around Abu Mina, is unmatched in other parts of the country (Fig. 1).

Sea, it was by far the best in the Nile Valley. Sources from the Pharaonic and Persian periods testify to wine production in this area long before the coming of the Greeks (cf. Forbes 1955: 70-124). The wineries date from the Roman and Byzantine periods. Some are of considerable size and testify to extensive cultivation of grapes and large scale wine production in this area, especially in correlation with the local pottery kilns which indicate a mass production of amphorae (Empereur 1993: 39-47).

In addition to the identified wineries, there are also elements of wine-making installations that are no longer in situ – primarily lion headed spouts through which freshly pressed or trodden must flowed into the vat. Numerous examples, including those found around Lake Maryut, can be seen in Alexandria in the Greco-Roman Museum or at the Kom el-Shukaqa necropolis (Rodziewicz 1998: 28, n. 6).

Literary Sources In literary sources wine produced in the Mareotic region during the Greco-Roman Period is well attested. It was known to Horace (Odes I 37.14), who wrote that the mind of Cleopatra was made mad by Mareotic wine (mens lymphata Mareotico). According to Strabo (XVII 1,14) the Mareotic region produced good wine and in such quantity that it was racked to be matured. The white Mareotic grape variety (Mareotides albæ) is mentioned by Virgil in Georgica (Geor. ii. 91). Athenaeus (Deip. I 33 d-e), in turn

Many factors contributed to the large concentration of wineries in the vicinity of Lake Maryut. Although in antiquity the climate in the Delta may have not been as beneficial for vineyards as on the northern shores of the Mediterranean

Fig. 1: A map of the Mareotic region with marked locations of wineries. Plotted on the map are excavated and published structures, as well as unexcavated ones mentioned by scholars in other works or featured on SCA maps as protected archaeological sites. The wineries designated as NN were plotted by Rodziewicz (1998: 28, Fig. 1), but the name of their location was not provided (D. Dzierzbicka).

127

LAKE MAREOTIS: RECONSTRUCTING THE PAST claims that Mareotic wine was excellent; it was white and pleasant, fragrant, easily assimilated, thin, did not go to the head and was diuretic. According to this author better yet was the wine from Taenia, from the same region. This was pale yellow and had an oily quality, which disappeared after it had been diluted with water. Besides being sweet, this wine was rather aromatic and mildly astringent.

There are three other wine making installations in the vicinity of the modern village of Huwariya. One of them is a simple Early Roman Type I unit built of dressed stone blocks bound with mortar (unexplored, mentioned by Rodziewicz 1998: 31). A winery in southern Huwariya (reported by Rodziewicz 1998: 34), built of irregular stone blocks, clay and waterproof plaster, consisted of one press unit – a treading platform and vat connected by two flights of steps. The round base of a mechanical press was set in a niche in the wall of the treading platform. A channel in the floor of the platform led from the press directly to the vat. A third installation, located 3 km southeast of Huwariya (Rodziewicz 1998: 35), consisted of a treading platform, a vat and a mechanical press separated from the treading floor by a low, thin wall. A channel under the floor led from the mechanical press to the vat. The walls of the structure were built of irregular stone blocks bonded with mortar and covered with plaster reinforced with potsherds.

Wineries1 The South Coast of Lake Maryut The wineries in the immediate vicinity of Lake Mareotis are located on the strip of land between the lake and the Mediterranean Sea, and along the southern shore of the lake. The majority are found on the southern shore of the lake. A large winery near Burg el-Arab (30o55’37”N 29o 31’48.50”E),2 is located in the northeast corner of a large building (El-Ashmawi 1998: 55-73). A mechanical press was set up on the treading platform and closed off from the surrounding area by a thin wall. The must from the mechanical press enclosure and from the treading floor flowed to the adjacent collection basin through two separate channels.

A small, presumably Roman winery of George Nesim (30°57’18.30”N 29°36’15.20”E) was found by the Lake Mareotis Research Project (Blue & Khalil forthcoming). It appears to have consisted of a treading platform and vat (Type I; Rodziewicz 1998: 31), but the presence of a mechanical press unit with a separate, small collection basin within the same complex cannot be excluded, although it requires further investigation.

The winery located at Taher el-Masry (30o56’59.40”N 29o34’58.20”E)3 was built of stone blocks and stone rubble embedded in mortar. The uncovered press unit consisted of a square treading platform (3.75 m x 3.75 m) and a vat (2.37 m x 2.00 m), 1.40 m deep. Outlines of structures surrounding the unit suggest that the winery may have been part of some kind of a larger complex.

The North Coast of Lake Maryut To the north of the lake, at Abu Talaat (unpublished, mentioned by Rodziewicz 1998: 29), an Egyptian archaeological mission uncovered a winery with two connected treading platforms and one collecting vat. The wine-making unit was entirely hewn out of the bedrock except for one wall, which was built of stone blocks.

A restored wine-making unit near Marea (30o58’49.10”N 29o40’12.90”E) consisted of a large treading platform, a smaller room with a base for a mechanical press, and a collection vat (El-Fakharani 1983: 182-184). Two separate channels led from the two rooms to the vat. The rooms where the pressing took place were raised higher than the room of the vat and two flights of steps connected the two levels. Nearby, on the same site another unpublished winemaking unit was found.4

The winery at Site 215 recorded by the Lake Mareotis Research Project (30°53’30”N 29°25’22”E) has two parallel treading platforms visible on the surface. The channels in their western sides indicate the presence of two collection basins, which have not yet been uncovered. Thus, this winery represents a compound of two Type I wineries according to the typology established by Rodziewicz (1998: 31-35). Further investigation is needed to identify its other features.

1. This part of the paper includes elements previously published in Dzierzbicka 2005. 2. GPS coordinates of the wineries as well as some additional information – dimensions, architectural details – were collected on my study tour of the Mareotic wineries in November 2005 during my scholarship at the Polish Centre of Mediterranean Archaeology in Cairo, Egypt. The study tour included the wineries near Burg el-Arab (also called Abu Sir), at Taher el-Masry and Marea, two of the installations at Abu Mina, and the wineries at Karm el-Shewelhy and Karm el-Baraasi. I am thankful to Mr. Adli Rushdy from the West Delta Inspectorate and to Mr. Saber Selim from the Islamic Inspectorate of the Supreme Council of Antiquities for allowing me to see the sites. For GPS locations and photographs of the wineries at George Nesim, Kom Trouga and Site 215 I am greatly indebted to Dr. Lucy Blue and Dr. Emad Khalil of the Lake Mareotis Research Project. 3. Explored by Mr. Mustafa Rushdy of the Egyptian Antiquities Organisation in 1998/99 (as it was communicated to me by Mr. Maged Ahmed, Inspector of the SCA), unpublished. The survey I conducted in 2005 yielded the GPS coordinates of the site, as well as some preliminary observations based on the remains visible on the surface. 4. I am thankful to Prof. Mona Haggag for bringing it to my attention.

Area to the East of Lake Maryut Much fewer wineries have been located in the eastern part of the Mareotis region. A winery was uncovered in 1917 by Daninos Pasha near Abu Qir Bay east of Alexandria (published as a private bath in Breccia 1926: 47-49). The Ptolemaic dating established by Breccia on the basis of the quality of plaster used in the structure seems rather uncertain. A large treading platform (3.25 m x 5.80 m) covered with waterproof plaster was raised 0.30 m above the floor level of the building. Two channels led through a wall separating the platform from a collection basin 1.60 m deep, 4.20 m long and 2.20 m wide. Twin flights of steps connected the two parts of the winery. 128

D. DZIERZBICKA: WINERIES OF THE MAREOTIC REGION There is also a Roman Period winery at Kom Trouga (30°57’46.90”N, 30°10’27.10”E). This unpublished winery mentioned by El-Fakharani (1983: 184) and Rodziewicz (1998: 31) was a complex that consisted of several scattered wine making units, today badly damaged due to quarrying activity in the area. Remains of basins and adjacent raised treading platforms are visible. All of the units appear to have consisted of a single treading platform and vat, covered with waterproof plaster. The best preserved one was built of stone blocks, while others were brick structures. Possible wine-making installations have also been found at Schedia.5

two treading platforms, two vats, one mechanical press, and storage rooms. The structure was built of mud-brick and re-used stone blocks. The winery located at Karm el-Shewelhy (30o49’37.70”N, 29o38’65”E) consisted of a treading platform and vat, two storage rooms, and a courtyard. The wine-making unit was hewn in rock except for the eastern wall, which was built of limestone blocks. The floor of the treading platform and the walls of the vat were covered with waterproof plaster (Abd el-Aziz Negm 1998: 73). A sizeable winery at Karm el-Baraasi (30o50’41.40”N, 29o40’49.50”E) consisted of several large treading platforms with vats and mechanical presses. The excavations published by Abd el-Aziz Negm (1998: 63-70) brought to light a large treading platform and vat. In the same room as the vat there were two small enclosures (4 m2 each) for mechanical presses, each equipped with a small vat of its own. In November 2005 another treading platform was visible semicircular in shape, with an adjacent vat. This unit was located further to the west and separated from the first platform by a room. The floor level of this room was lower than the two treading platforms and connected to them by two short flights of steps. Further to the west traces of three more treading platforms were visible (not yet excavated). The complex was built of mudbrick with corners reinforced with limestone blocks. The entrance to the unit uncovered by the excavators led from a courtyard. The winery was part of a larger complex – across the courtyard from the winery there was a house, and the courtyard itself was surrounded by rooms. Around 300 m to the north-west of this complex (30o50’50.70”N, 29o40’40.20”E) there are probably more installations yet to be uncovered, as one can tell from the outlines of basins covered with waterproof plaster that are discernible on the surface.

The Region of Abu Mina A number of wineries were found in the vicinity of Abu Mina, not far from Lake Mareotis. The region was a thriving grape-growing and wine-making centre. There are three wineries within the town area. A large winery uncovered in 1907 by Kauffmann to the east of the double bath (Grossmann, et al. 1984: 148; Grossmann, et al. 1997: 87-90) behind the portico of the colonnaded street (30o50’52.90”N, 29o39’76.10”E) was expanded in five phases until it became a complex of five treading floors with four mechanical presses. It was built of mud-brick and stone blocks covered with waterproof plaster. Another winery was located behind the apse of the basilica (30o50’45.80”N 29o39’82.30”E). It consisted of two rooms, the first housing one large treading platform flanked by two mechanical presses, a collection vat, and one smaller raised platform. In the second room there was another treading floor and vat (Grossmann 1999: 82-83, fig. 9). The third winery (Müller-Wiener & Grossmann 1967: 468-473) within the area of Abu Mina was uncovered in the so-called Kumring A, in three rooms of Building 17, entered from a courtyard. Two of these rooms each contained a set of one large and one small treading platform and vat. The third room on the opposite side of the courtyard was used for storage.

Elements of Mareotic Wineries The Mareotic evidence provides an abundant and diverse range of evidence for wineries, a brief overview of the basic elements of a winery (the treading platform, the collection basin, and the mechanical press), and their function, are provided here.

The winery at Izbat Mohamed Farid, 1 km north of Abu Mina, had two phases of use (Grossmann 2002: 31). In the first phase, at the beginning or first half of the 5th century AD it was a simple pressing unit consisting of a treading platform and vat, with walls built of baked brick and covered with waterproof plaster. When it was no longer in use, another winery was built (later part of 6th century AD) partly on top of the old one. This new structure comprised a treading platform, a vat, and two mechanical presses. During both phases of use the winery stood next to a multistory country house.

Taking the winery at Marea as an example, it can be seen that a basic press unit was a compound consisting of a raised treading platform and a collection basin, built within an enclosed and most probably roofed, space (El-Fakharani 1983: 182-184). The treading floor was separated from the vat by a low wall. It was covered with waterproof plaster, preserved, for instance, in the Burg el-Arab winery (El-Ashmawi 1998: 62), and was often quite large, as in the case of the Karm el-Baraasi winery (Abd el-Aziz Negm 1998: 66). The room in which the press unit was located was sometimes entered from a courtyard, or had a large window giving easy access to the treading floor. Grapes were thrown through this window, or simply carried through the door, to the treading platform, where they were crushed.

Three wineries near Abu Mina were explored in 1986 by the Supreme Council of Antiquities (Abd el-Aziz Negm 1998: 65-73). The winery at Karm Gadoura consisted of

5. For this information I am indebted to Prof. Dr. Marianne Bergmann and Prof. Dr. Michael Heinzelmann of the Swiss-German mission excavating at Schedia.

129

LAKE MAREOTIS: RECONSTRUCTING THE PAST Fig. 2: A schematic drawing of a Type 4 winery with names of its elements used in Greek papyri from Egypt (D. Dzierzbicka).

Wineries in Papyri Archaeological finds are not the only available sources on wineries (for more details on wine production in papyri see Dzierzbicka 2005: 24-89). Greek papyri of the Hellenistic and Roman period allow for the identification of words used to call the elements of a winery. The reconstruction drawing of Type 4 winery (Fig. 2) shows the basic parts of a winery and gives the terms by which they were called in papyrus documents. Matching the architectural features with their Greek names in papyri makes it possible for the researcher to use hundreds of texts that provide invaluable information on the function of the winery as a whole and on its elements, permitting a better understanding of the way these structures were used and provoking more questions to ask oneself in the field.

Most wineries were also equipped with a mechanical press – either a movable bag press, or a more sophisticated, fixed screw press. Their preserved remains were found in Abu Mina (Grossmann, et al. 1997: 87-90) and in Burg el-Arab (El-Ashmawi 1998: 60-64), where one can still clearly see a raised, circular base of the screw press. These devices were used to squeeze the remaining liquid out of the trodden grape pulp either directly on the treading floor, or in an area separated from it, as in Burg el-Arab. In the latter case, a channel connected the place of the mechanical press with the main collection basin, or with a smaller, separate vat. The must flowed into a plastered vat or vats dug into the ground. It remained there through the first turbulent phase of fermentation, during which the basins were covered with wooden planks for protection against contaminants.

One of these questions, especially important in the case of the Mareotic wineries, concerns their surroundings. Papyrus documents containing references to buildings and other immovable property found in vineyards, provide interesting information on other installations forming part of a wine-making complex (Table 1). The table below is a comparative chart of the facilities mentioned in extant texts, showing how frequently they appear and in what combinations. The documents mention elements already discussed: the treading platform (lênos) and collection basin (pithos), in a later period referred to by a common term – lênopithos, a mechanical press (stemphylourgikon organon or mêchanê), as well as a sunning yard (hêliastêrion), a storeroom (thesauros) and a cellar (oinothêkê). In fact, spaces that may have been storerooms were uncovered adjacent to wineries, for instance in Abu Mina (Grossmann, et al. 1997: 89). Large wineries where mass production took place even had a potter’s workshop (keramikon ergastêrion) on the premises. There was indeed a large pottery kiln in the vicinity of the winery near Burg elArab, suggesting this may have been such a more developed production estate (El-Ashmawi 1998: 62).

Generally speaking, wineries exhibit a rather uniform structure, dictated by practical needs, that hardly changed regardless of whether the winery was located on the Mareotis lakeshore, in the Bahariya Oasis (unpublished wine press; photo in Hawass 2000: 164), in Nubia (Gardberg 1970: 41-44), or even on the Crimean Peninsula (Michalowski 1958: 49). Mareotic wineries provide an excellent chance to study the characteristic traits of these structures. What is striking about the typology elaborated by Rodziewicz (1998: 31-36) is the diversity of wine making installations which led to the distinguishing of as many as six simple and two composite types in just over a dozen wineries. However, the distinguished types are in fact very useful as they have strong similarities with wineries in other parts of Egypt and beyond (e.g. Michalowski 1958: 49). As it was noted by Rodziewicz (1998), the wineries consist of two or three essential elements: the treading platform, the collection basin, and, in types 3-8, a mechanical press. The convenient uniformity of this category of structures makes them fairly easy to recognise in an archaeological excavation. 130

P. Tebt. III.1 814, Sale, after 227 BC, Tebtynis

X

P. Lond. II 401, Petition, 115–110 BC, Thebaid

X

P. dem. Gieben 2, Sale, 107–30 BC, Sebennytos

X

PSI VIII 918, Sale, AD 38/39, Tebtynis

X

X

X

P. Ross. Georg. II 28, Lease, after AD 163/164, Arsinoites

X

P. Flor. III 385, 2nd–3rd century AD, Hermopolites

Oikopedon

Pyrgos

Kella

Epoikion

Epaulis

X

X X X

PSI XIII 1328, Petition, AD 201, Oxyrhynchos

X

P. Oxy. LI 3638, Cession, AD 220, Sinary, Oxyrhynchites

X

P. Flor. I 50, Division of property, AD 269, Hermopolis

X

P. Oxy. XXXIV 2723, Sale, 3rd century AD, Oxyrhynchos

X

X

X

SPP XX 58 (7) Col. II, 3rd century AD, Hermopolis

X

X

X

SB XX 14291, Lease, 3rd century AD, Oxyrhynchos

X

X X

X

X

X

X

X

X

X

X

X

X

X

X

X X

X

P. Vind. Sal. 12, Lease, AD 334/335, Hermopolis

X X

X

X X

P. Vind. Tand. 28, Lease, AD 576/577; Hermopolis

SPP XX 218, Lease, early 7th century AD, Hermopolites

X

X X

P. Cairo Masp. I 67097 (r), Contract, AD 571/572 (?), Aphrodites Kome (Antaiopolites)

Thesauros

X

P. Oxy. XLIX 3491, Marriage contract, AD 157/158, Oxyrhynchos

P. Vat. Aphrod. 25 fr. A, Division of property, 6th century AD, Aphrodito

Oinothêkê

Hêliastêrion

Keramikon ergastêrion

Pithos

Stem. organon

Mêchanê

Lênopithos

DOCUMENT

Lênôn; lênos

D. DZIERZBICKA: WINERIES OF THE MAREOTIC REGION

X X

X X

X

X

X X

X

Table 1: Installations forming part of a wine-making complex mentioned in Greek papyri containing references to immovable property found in vineyards or to the appurtenances of a winery. 131

LAKE MAREOTIS: RECONSTRUCTING THE PAST P. dem. Gieben 2 = Vandorpe, K., & Clarysse, W., 1998, A Greek Winery for Sale in a Fayum Demotic Papyrus. In A.M.F.W. Verhoogt & S.P. Vleeming (eds.), Two Faces of Graeco-Roman Egypt. Greece and Demotic and Greek-Demotic Texts and Studies Presented to P.W. Pestman: 127-139. Leiden, Boston, Köln. P. Flor. I 50 = Vitelli, G. (ed.), 1906, Papiri greco-egizii, Papiri Fiorentini I. Documenti pubblici e privati dell’età romana e bizantina. Milan. P. Flor. III 385 = Vitelli, G. (ed.), 1915, Papiri grecoegizii, Papiri Fiorentini III. Documenti e testi letterari dell‘età romana e bizantina. Milan. P. Lond. II 401 = Kenyon, F.G. (ed.), 1898, Greek Papyri in the British Museum II, 13-14, no. 401. London. P. Oxy. XXXIV 2723 = Ingrams, L., Kingston, P., Parsons, P.J., & Rea, J.R. (eds.), 1968, The Oxyrhynchus Papyri XXXIV, no. 2723. London. P. Oxy. XLIX 3491 = Bülow-Jacobsen, A., & Whitehorne, J.E.G. (eds.), 1982, The Oxyrhynchus Papyri XLIX, no. 3491. London. P. Oxy. LI 3638 = Rea, J.R. (ed.), 1984, The Oxyrhynchus Papyri LI, no. 3638. London. P. Ross. Georg. II 28 = Krüger, O. (ed.), 1929, Papyri russischer und georgischer Sammlungen II. Ptolemäische und frührömische Texte, no. 28. Tiflis. PSI VIII 918 = Vitelli, G., & Norsa, M. (eds.), 1927, Papiri greci e latini VIII, no. 918. Florence. PSI XIII 1328 = Vitelli, G., & Norsa, M. (eds), 1949-1953, Papiri greci e latini XIII, no. 1328. Florence. P. Tebt. III.1 814 = Hunt, A.S., & Smyly, J.G. (eds.), 1933, The Tebtunis Papyri III.1, no. 814. London. P. Vat. Aphrod. 25 = Pintaudi, R. (ed.), 1980, I Papiri Vaticani di Aphrodito, no. 25. Rome. P. Vind. Sal. 12 = Salomons, R.P. (ed.), 1976, Einige Wiener Papyri, no. 12. Amsterdam. P. Vind. Tand. 28 = Sijpesteijn, P.J., & Worp, K.A. (eds.), 1976, Fünfunddreissig Wiener Papyri, no. 28. Zutphen. SB XX 14291 = Bilabel, F., Kiessling, E., & Rupprecht, H.-A. (eds.), 1997, Sammelbuch griechischer Urkunden aus Aegypten XX, no. 14291. Wiesbaden. SPP XX 58 = Wessely, C. (ed.), 1921, Studien zur Palaeographie und Papyruskunde XX. Catalogus Papyrorum Raineri. Series Graeca. Pars I. Textus Graeci papyrorum, qui in libro „Papyrus Erzherzog Rainer– Führer durch die Austellung Wien 1894“ descripti sunt, no. 58. Leipzig.

The winery was not just a press unit, but a building or a set of buildings, located either in the urban area or in the vicinity of a vineyard. According to some documents the complex was also equipped with farm buildings (epoikion, pyrgos, epaulis), storehouses (kella) and building plots (oikopedon). Indeed, the treading platform and vat at Abu Qir Bay was part of a sizeable square structure built of limestone blocks. The layout of rooms in the building is only partly preserved (Breccia 1926: 47-49). It has not been determined where the entrance to the winery was located and whether there were passages leading to the adjacent rooms. The wine-making complex in this structure may have not been limited to only one unit, consisting of a treading platform and vat, especially that the other rooms also had an industrial function, perhaps related to wine and oil production. As we learn from papyri (Dzierzbicka 2005: 24-89), after the turbulent phase of fermentation was over, the wine was poured into vessels that had been obtained from a pottery workshop located within the complex or acquired from outside the winery. Wine that was paid for in advance was distributed straight from the vat and carried off from the winery in vessels provided by the buyer. The filled jars that remained in the winery were placed in storage rooms or in the sunning yard (hêliastêrion) for maturing. Both of these facilities were usually spaces or buildings within the wine-making complex and they were equipped with doors and locks. Maturing wine was occasionally controlled to check if it had not gone sour or spoiled. After the process of fermentation had ended, the vessels were sealed. There are still many uncertainties about the broader context of wineries. The immediate surroundings of a winemaking structure can provide information on the building that housed the installation and determine whether it was located in a rural setting or in a larger industrial or even residential area. Further archaeological investigations, as well as the study of literary and documentary sources, may provide insight into industrial buildings and structures found in the vicinity of a winery and improve our understanding of the character of wine production in the Mareotic region. Acknowledgments I am grateful to the Polish Centre of Mediterranean Archaeology of the University of Warsaw for assistance granted to me during my stay in Egypt during the conference.

Ancient Sources Athenaeus, The Learned Banqueters, books I-III.106e, S.D. Olson (transl. & ed.), 2006. Cambridge. Horace, Odes I, Carpe Diem, West, D. (transl. & ed.), 1995. Oxford. Strabo, Geography, H.L. Jones (ed.), 1996. London, New York. Virgil, Eclogues. Georgics. Aeneid 1-6, Vol. 1, H.R. Fairclough (ed.), 1974. Cambridge.

Bibliography Papyri P. Cairo Masp. I 67097 (r) = Maspero, J. (ed.), 1911, Papyrus grecs d’époque byzantine. Catalogue général des antiquités égyptiennes du Musée du Caire I, no. 67097. Cairo.

132

D. DZIERZBICKA: WINERIES OF THE MAREOTIC REGION Gardberg, C.J., 1970, Scandinavian Joint Expedition to Sudanese Nubia, Vol 7: Late Nubian Sites: Churches and Settlements. Astrom. Grossmann, P., Kosciuk, J., Severin, G., & Severin, H.-G., 1984, Abu Mina. Elfter Vorläufiger Bericht. Kampagnen 1982–1983. Mitteilungen des Deutchen Archäologischen Instituts Abteilung Kairo 40: 148. Grossmann, P., Arnold, F., & Kościuk, J., 1997, Excavations at Abu Mina 1995. Bulletin de la Société d’Archéologie Copte 36: 87-90. Grossmann, P., 1999, Report on the excavations at Abu Mina in spring 1998. Bulletin de la Société d’Archéologie Copte 38: 75-84. Grossmann, P., 2002, Report on the excavations at Abu Mina in spring 2001. Bulletin de la Société d’Archéologie Copte 41: 25-31. Hawass, Z., 2000, The Valley of the Golden Mummies. Cairo. Michalowski, K., 1958, Mirmeki I. Warszawa. Müller-Wiener, W., & Grossmann, P., 1967, Abu Mina. 6. Vorläufiger Bericht. Archäologischer Anzeiger 82.4: 468-473. Rodziewicz, M.D., 1998, Classification of wineries from Mareotis. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athènes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole française d’Athènes, 11-12 décembre 1988: 27-36. Athens, Paris.

Secondary Sources Abd el-Aziz Negm, M., 1998, Recent excavations around Abou Mina. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athènes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole française d’Athènes, 11-12 décembre 1988: 65-73. Athens, Paris. El-Ashmawi, F., 1998, Pottery kiln and wine-factory at Burg el-Arab. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athènes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole française d’Athènes, 11-12 décembre 1988: 55-64. Athens, Paris. Blue, L., & Khalil, E. (eds.), forthcoming, A Multidisciplinary Approach to Alexandria’s Economic Past: the Mareotis Case Study. Oxford Breccia, E., 1926, Le rovine e i monumenti di Canopo, Teadelfia e il tempio di Pneferôs. Bergamo. Dzierzbicka, D., 2005, Wineries in Graeco-Roman Egypt. Journal of Juristic Papyrology 35: 9-89. Empereur, J.-Y., 1993, La production viticole dans l’ Egypte ptolémaïque et romaine. In M.-C. Maouretti & J.P. Brun (eds.), La production du vin et de l’huile en Méditerranée: 39-47. Athens. El-Fakharani, F., 1983, Recent Excavations at Marea, Egypt. Aegyptiaca Treverensia 2: 175-186. Forbes, R.J., 1955, Studies in Ancient Technology, Vol 3. Leiden.

133

LAKE MAREOTIS: RECONSTRUCTING THE PAST

134

E. KHALIL: MAREOTIS’ WATERFRONT

Waterfront Installations and Maritime Activities in the Mareotic Region Emad Khalil

Introduction Lake Mareotis represents one of the most distinctive geomorphic features in the north-west coastal region of Egypt. In antiquity, it was fed by means of a number of canals, which bifurcated off the Nile’s defunct Canopic Branch, and flowed into the southern and eastern sides of the lake (Fig. 1). Some of these canals were navigable, which enabled merchandise to be transported to and from the hinterland. By the Greco-Roman period the lake was also connected to the sea through a navigable canal that debouched at Alexandria (Strabo 17.1.7). Its connection to both the Nile and the sea resulted in Mareotis becoming a vital conduit of communication in Egypt’s internal transport system. Moreover, it supported around its shores various agricultural activities and embraced major production centres for different industries, which contributed significantly to the economy of Alexandria and to Egypt as a whole. Accordingly, this paper will look at the role that Lake Mareotis played in the ‘maritime’1 transport system of GrecoRoman Egypt. It will also examine the types and nature of the maritime and waterfront installations that were recorded along the shores of the lake and the possible spatial and functional relationship between the different sites.

(Strabo 17.1.14; Pliny 5.11.63). It comprised a main rectangular body of water which merged to the east and south with the Nile Delta Plain, and a narrow arm that extended westwards parallel to the northern coast. However, during the past two millennia Lake Mareotis has undergone dramatic changes which significantly affected its size and nature. Nonetheless, the western extremities of the lake reflect the original extant remains, and form an arm, known as the Mareotic Arm, that extends some 40 km west of Alexandria and is 2-3 km wide and is separated from the lake’s main body by causeways and shallows. It also contains an island, Mareotis Island, which is 3.7 km long and about 680 m at its widest point (see both Blue and Hopkinson this volume).

The Lake Mareotis Research Project Much evidence indicates that Lake Mareotis extended in antiquity for about 50 km south and west of Alexandria

Since 2004 the Centre for Maritime Archaeology (CMA) of the University of Southampton, in collaboration with the Department of Underwater Antiquities (DUA) of the

Archaeological investigation of the western Mareotic Arm has been ongoing for several decades; however, previous research has been largely limited to specific areas and specific issues such as work on the Byzantine port of Marea/ Philoxenité (Petruso & Gabel 1982; El-Fakharani 1983; Rodziewicz 2003), and work on amphorae and wine production (Empereur & Picon 1986, 1998) and on the wineries of the Mareotic region (Rodziewicz 1998b).

Fig. 1: The approximate ancient and present limits of Lake Mareotis (E. Khalil).

1. The term ‘maritime’ is used in this sense to denote all aspects of waterborne activity and communication from the sea, across the lake, along the canals and on the Nile River.

135

LAKE MAREOTIS: RECONSTRUCTING THE PAST Fig. 2: The Lake Mareotis Research Project survey area along the shores of the western Mareotic Arm (Lake Mareotis Research Project).

Waterfront Sites in the Mareotic Region The waterfront sites that were recorded along the shores of the western Mareotic Arm and on Mareotis Island can be classified into four categories:

Egyptian Supreme Council of Antiquities (SCA), has conducted a comprehensive archaeological survey along the shores of western arm of Lake Mareotis (see Blue this volume). The survey revealed a wealth of archaeological sites including numerous settlements of a maritime and industrial nature, which reflect the economic activities that took place in the Mareotic region (Fig. 2). Most of the archaeological sites recorded along the shores of the Mareotic Arm are located between Marea/Philoxenité and Taposiris Magna, with a concentration on Mareotis Island. The sites identified include maritime structures, such as harbours, jetties and quays, in addition to what appear to be waterfront warehouses and storage facilities. Sites pertaining to industrial activities are also evident and include amphora kilns, ceramic slag and kiln wasters, as well as a number of structures relating to water management, such as cisterns, wells and water wheels (sakkia). This correlates with much archaeological and textual evidence for viniculture and wine production in the region (Rodziewicz 1998b).

1 - Harbours; in the form of harbour complexes of significant design and constitute substantial structures. However, the only two sites that fit this description are Marea/Philoxenité and Taposiris Magna. The complexity and magnitude of their structures are unrepresented elsewhere in the entire Mareotic region. The two harbours date, however, to quite different periods. Taposiris Magna is essentially Hellenistic in date (see Boussac & El Amouri this volume), while Marea/Philoxenité mostly dates to the Late Roman period (see Haggag, Pichot, Rodziewicz and Babraj & Szymańska this volume). However, the two harbours are associated with relatively large towns, and much historical and archaeological evidence indicates that these two towns where probably among the largest and most active along the shores of Lake Mareotis in antiquity.

The dating of the sites discovered relies primarily on the ceramic assemblages collected during the survey. Accordingly it was realised that the majority of sites date from the Hellenistic period until the 7th century.

Taposiris Magna has extensive archaeological remains that date from the Hellenistic to the Late Roman period, including evidence for thriving maritime and commercial 136

E. KHALIL: MAREOTIS’ WATERFRONT activities (El-Fakharani 1974; Empereur 1998: 225-7; Rodziewicz 1998a; Boussac & El Amouri this volume). The town which is located on the northern shore of the western sub-basin of the lake has one of the best preserved harbours in Lake Mareotis (see Boussac & El Amouri this volume). Commercial activities in Taposiris Magna were mainly focused on handling products transported across the lake as well as receiving goods arriving from the west through overland routes, and shipping them to Alexandria (Empereur 1998: 225). Under the Romans, Taposiris Magna was a customs station where duties were levied on products coming from the Mareotic region and from Cyrenaica heading east towards Alexandria or to the Nile Delta (Empereur 1998: 225-7; Vörös 2001: 15-6). Alternatively, river vessels could have travelled on the lake through the harbour of Taposiris Magna to the west as far as the lake extended.

ings to allow water to flush away the silt and sediments that might accumulate in the harbour basin. This arrangement was supplemented by the construction of a 1,700 m long wall that extends southwards from the artificial ridge to the southern shore of the lake. It was also supplemented by the construction of a solid limestone wall that extended from the northern shores of the lake to the seashore. This wall ensured that even caravans travelling overland had to go through the town of Taposiris Magna (see Fig. 1 in Boussac & El Amouri this volume) (De Cosson 1935: 111; Rodziewicz 1990: 72-4). Another prominent structure in Taposiris Magna is a 17 m high tower that stands on the coastal ridge to the north of the harbour over looking the Mediterranean coast to the north and the lake to the south (Kadous 2001: 457-60). The function of the tower and its relation with the town and harbour of Taposiris Magna is disputed (El-Fakharani 1974; Vörös 2001: 37). Nonetheless, it is generally accepted that the tower, which represents a 1:4 or 1:5 replica of the Pharos lighthouse of Alexandria, was in fact a funerary monument for a Hellenistic necropolis that occupied the area around and below the tower (Empereur 1998: 225). The utilisation of such a structure as a landmark by navigators on both the sea and the lake is a possibility that cannot be overlooked (Fig. 3)

The harbour of Taposiris Magna was constructed to control the movement of vessels travelling through it. This was achieved by digging a channel c. 1,700 m long and 50 m wide parallel to the northern shore of the lake. The spoil resulting from the digging was piled up to form an artificial ridge which delimits the channel from the south. At the western end of the channel stood a limestone structure that took the form of a double-opening gate or bridge, through which all boats wishing to go through Taposiris Magna had to pass (Empereur 1998: 225-7; Rodziewicz 1998a: 102, n. 32; Vörös 2001: 15-6) (see Figs. 5-7 in Boussac & El Amouri, this volume). The total width of the gate is about 8.3 m, however, it is divided by a 1.2 m thick wall into two openings; one is 4.1 m wide and the other one is 3 m wide, thus indicative of the maximum possible width of the vessels that passed through. The eastern entrance of the channel is partially obstructed by a quay which is c. 230 m long extending from north to south perpendicular to the shoreline. The distance between the southern end of the quay and the eastern end of the artificial ridge, c. 100 m, forms the eastern entrance of this semi-closed harbour basin of Taposiris Magna. The eastern quay of the harbour includes at least two de-silting open-

As a result of the recent excavation of the area, it is now believed that the digging of the channel as well as the construction of this harbour system, took place the during the Early Roman period rather than during the Hellenistic period, as was previously believed (El-Fakharani 1974; Boussac & El-Amouri this volume). Although the northern shoreline at Taposiris Magna was occupied during the Hellenistic period, as evidenced by houses and shops from the 2nd and 1st century BC, it seems that it was abandoned by the end of the Hellenistic period as a result of a rise in the lake level. The flooded area was then excavated in the Roman period to create the closed harbour system (see Boussac & El Amouri this volume). Marea/Philoxenité is located about 15 km east of Taposiris Magna, on the southern shore of the lake. El-Falaki (1966: 96) identified this settlement and its associated harbour as the town of Marea, the capital of the Mareotic region. According to Herodotus (2.31), Marea was a post of Egyptian soldiers guarding the Libyan border during the time of King Psammetichus of the 26th Dynasty (Rodziewicz 2003: 27). In the Ptolemaic and Roman periods, Marea functioned as a major trade centre, second only to Alexandria. In the Byzantine period, in addition to its involvement in commercial activities and internal trade, Marea flourished as a stopover for pilgrims heading to the holy Byzantine shrine of St. Minas (Abu Mina), 20 km south of the lake (Gabel & Petruso 1980; Kucharczyk 2002; Rodziewicz 2003). Until recently, most archaeological research carried out in the area has revealed no evidence earlier than the 5th century AD. However, recent archaeological investigation at Marea has revealed material

Fig. 3: Taposiris Magna tower (photo: E. Khalil).

137

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig. 4: The middle quay of the harbour of Marea (photo: E. Khalil). from the Hellenistic and Roman periods (see both Pichot and Babraj & Szymańska this volume).

Excavations on the peninsula at the easternmost part of Marea, resulted in the discovery of a large Early Roman building which consists of a courtyard surrounded by numerous rooms of relatively similar size. The building is connected to a quay to the north through stairways cut into the rock. Therefore, it has been suggested that the building could have been used for storage and trade. The remains of 1st century BC to 1st century AD workshops for metalwork were also discovered on the peninsula (see Pichot this volume). The recent archaeological discoveries at Marea, particularly of pottery and coins, would indicate that the area was thriving before the 5th century and possibly as early as the Hellenistic period

Amongst the most significant archaeological remains in Marea are four quays that extend into the lake and divide the 1.5 km long shoreline into eastern, central and western harbour basins (Figs. 4 & 5). The dimensions of the quays from west to east are: 41 m x 6.5 m, 111 m x 5 m, 125 m x 7 m, and 35 m x 4 m (Szymańska & Babraj 2008: 1115). Judging from the construction technique of the quays at Marea, in which large regular limestone blocks (1 m x 0.5 m x 0.03 m) were used for their construction, it seems reasonable to suggest that the four quays were constructed earlier than the Byzantine city. However, they were probably subject to several building phases in subsequent periods since evidence of Byzantine hydraulic mortar (opus signinum) can still be seen between many of the building blocks of the quays.

2 - The second category of waterfront sites that was recorded along the shores of the western Mareotic Arm consists of different types of anchorage facilities such as quays and jetties, which form the majority of maritime Fig. 5: The quay of Marea harbour was made of large regular limestone blocks without evidence of the use of mortar (photo: E. Khalil).

138

E. KHALIL: MAREOTIS’ WATERFRONT

Fig. 6: The box-shaped harbour on the southern shore of Lake Mareotis (photo: Lake Mareotis Research Project). installations in the region. More than ten different anchorage facilities were recorded on the northern and southern shores of the lake and on the northern shore of Mareotis Island. Possibly the most substantial of them is a Kibotos or box-shaped harbour (Site 09) which is located at the end of a promontory on the southern shore of the lake, about 2 km to the west of the Sidi Kerir-Borg El Arab road. The harbour, which is constructed of large limestone blocks, consists of two parallel moles enclosing an area some 60 m long (N to S) and 36 m (E to W) wide (Fig. 6). The eastern mole is 60 m long and at its northern extremity it returns to the west for a distance of 12 m, while the western mole, which is less well preserved, extends for 40 m and returns to the east at the northern end for a distance of some 6 m. A gap of 18 m between the two ends of the two moles equates to the entrance of the harbour on the north side. The moles are constructed of up to three courses of single and double breadth limestone blocks. Although the dating for this harbour is still uncertain, judging from its construction technique and from the large size of blocks used (c. 1.10 x 0.7 x 0.5 m), it seems that the harbour is pre-Roman in date. However, the existence of lumps of coarse red mortar opus signinum with lime inclusions between some of the blocks indicate that it remained in use at least until the Byzantine period. One carved mooring ring was noted on the upper course of one of the blocks, which would have helped facilitate the mooring of vessels to the outside of the harbour (El-Fakharani 1984; Blue & Ramses 2006).

They are Sites 204 (Gamal) and 208 (Quseir). The jetties at each of these sites, which are located approximately 2 km apart, are about 50 m long and 8 m wide. Evidence of red mortar opus signinum is noted between some of the blocks. Also the jetty at Site 208 (Quseir) had mooring stones extending from the upper course of blocks at the eastern side (the lee side) of the quay. Other jetty-like features, although not as substantial, are located along the southern shore of the lake and the northern shore of the Mareotis Island. However, it was realised during the survey that the anchorage facilities along the northern shore of the lake are mostly associated with civic and residential sites, while those along the southern shore of the lake, particularly on Mareotis Island, are associated with sites of a commercial nature. The dating of these sites based on ceramic collections is quite problematic since the jetties are continuously washed by water in the winter, which, in many cases, does not leave any ceramics to be dated. However, judging from the ceramics dated from adjacent sites, it was realised that most of the sites could have been used for quite a long period of time, probably from the Hellenistic to the Late Roman or Byzantine period. 3 - The third type of maritime installation identified can be described as seawalls or more accurately lake walls. Unlike the jetties, which are perpendicular to the shore, lake walls parallel the shore and they were intended to define the shores and protect them from the effects of silting and sedimentation (Fig. 8). At least five lake walls were discovered in the survey region. These kinds of structures are mainly found along the southern shore of the lake and the northern shore of Mareotis Island, those shorelines most subject to silting and the deposition of sediments as a result of the prevailing northwest winds that would carry sediment from the coastal ridges and deposit it into the lake. Besides acting as a form of protection against silting, the lake walls could also have been utilised as docking facilities for merchant vessels. Another possible function for such structures was to retain rainwater for use in agricultural purposes.

With the exception of this square harbour on the southern shore of the lake, all the other anchorage facilities along the shores of the lake take the form of jetties and quays that extend into the water perpendicular to the shore (Fig. 7). The technique used for the construction of most of the quays was building two parallel single or double breadth piers of limestone and filling the distance between them with rubble. With the exception of the substantial structures already noted at Marea/Philoxenité and Taposiris Magna the two most substantial anchorages are located on the northern shore of the lake opposite Mareotis Island. 139

LAKE MAREOTIS: RECONSTRUCTING THE PAST Fig. 7: One of the quays extending into the lake at the northern shore of Mareotis western arm (photo: Lake Mareotis Research Project).

Fig. 8: (below) A lake wall extending parallel to the southern shore of the lake. Evidence for red mortar (opus signinum) can still be seen between the blocks (photo: Lake Mareotis Research Project).

The longest of these lake walls is located on the north shore at the western end of Mareotis Island (Site 21). It is c. 245 m long and 1 m wide and is constructed of a series of limestone blocks laid as stretchers along the lake edge (see Hopkinson this volume). Other examples include a wall located in the middle of the island on the north shore, the extant remains of which extended some 70 m in length and was made of one course of large limestone blocks of 0.60 m x 0.30 m x 0.25 m dimension, that were arranged as headers facing the shoreline (Site 123). Similarly, on the southern shore of the lake further substantial walls of over 250 m in length were identified at both Sites 109 and 44. 4 - The fourth and final type of waterfront structures does not necessarily have a maritime function. At a number of sites in the survey region the remains of several multi room buildings were recorded very close to the present waterline. Some structures even extend into the water (Fig. 9). Examples of this type of structure are found in Sites 117, 118 and 119 which are located at the north-eastern shore of the island. At Site 117 there are the remains of a multi-roomed building that measures 12 m EW x 17 m NS which is divided on the inside into at least four smaller rooms. Site 118, about 25 m west of Site 117, contains the remains of at least two multi-roomed structures which measure 18 m EW x 15 m NS and 20 m EW x 20 m NS. Each of them contains the remains of numerous walls which belonged to a number of internal rooms of different shapes and sizes. About 40 m to the west of Site 118, a further Site 119 contains the remains of a rectangular building that extends from the shoreline southwards for about 40 m and measures about 25 m EW. Limited excavation carried out in the middle section of this building revealed the remains of two 140

E. KHALIL: MAREOTIS’ WATERFRONT

Fig. 9: Some of the multi-room square buildings located at Site 118 along the northern shore of the Mareotis Island. The buildings could have been used for storage purposes (image: Lake Mareotis Research Project). significant structures. The first structure is a rectangular enclosure that measures c. 9 m x 4.5 m that contained two rooms. Both rooms have almost the same dimension c. 3 m EW x 2.7 m NS. The second structure constructed on the same alignment as the first and adjacent to it to the west, is a rectangular building that measures c. 2.5 m NS and at least 5 m EW, in which remains of imported Hellenistic amphorae were discovered.

Marea and Taposiris Magna, only a distance of some 15 km, but out of a total of more than ninety archaeological sites recorded along the shores of the Mareotic Arm west of Alexandria, only four substantial sites were recorded to the west of Taposiris Magna. Thus, judging from the nature and extent of these sites, it is evident that the navigable limits of Lake Mareotis in antiquity extended west of Taposiris Magna for at least 12 km.

Accordingly, it seems reasonable to suggest that this type of waterfront structure were used as storage facilities for different merchandise and products that were traded along the Mareotic Arm.

Likewise, it is noticeable that maritime installations located on the southern shore of the lake from Marea to the eastern end of Mareotis Island, are in fact located on a ridge that extends for about 800 m from the present southern shoreline to the west. It is noteworthy that no sites were recorded on the southern shore of the lake opposite this ridge. This actually raises a question about the nature of the area between the ridge and the southern shore of the lake, and whether or not it was actually land in antiquity, that has subsequently been artificially excavated or subject to inundation due to the changes in ground water level (see Flaux forthcoming). Similarly, all the archaeological sites of a maritime nature that were recorded on Mareotis Island, where located along its northern shore, with essentially no evidence for sites either on the southern shore of the island or on the northern shore of the lake opposite. This also raises a question about the nature of this island and whether or not it was actually an island in antiquity.

By examining the remains of these structures, particularly on Maerotis Island, it becomes evident that they have undergone construction phases over successive periods. Moreover, it seems that the sections of the structures closest to the waterline were subject to accumulated sediments, and hence had to be rebuilt. In other words the different phases of building and modification of structures could be the result of adapting to changes in the waterline. Relations and Significance By looking at the distribution of archaeological sites in general and maritime sites in particular along the shores of the Mareotic Arm, it becomes evident that not only is there an apparent concentration of sites in the area between 141

LAKE MAREOTIS: RECONSTRUCTING THE PAST Navigation in Lake Mareotis In the 1st century BC, when speaking about the water supply for Lake Mareotis, Strabo (17.1.7) mentions that it is “…filled by many canals from the Nile, both from above and on the sides, and through these canals the imports are much larger than those from the sea, so that the harbour on the lake was in fact richer than that on the sea”. On another occasion (Strabo 17.1.22) speaks of “…several canals, which empty into Lake Mareotis”. As a result, it has been assumed that there was intense maritime traffic passing through the lake carrying various products and cargoes to Alexandria. Merchandise which would have been transported to Alexandria for local consumption and for transshipment to other Mediterranean harbours would have included Egyptian products such as papyrus, textile and grain (Rickman 1971: 300-6, 1980: 231-5; Lewis 1983: 165-7), as well as quarried stones from the Eastern Desert (Peacock 1992: 5-7; 2002: 426-7). It would have also included products imported via the Red Sea from India, Arabia and East Africa such as spices, tortoiseshell, frankincense, ivory, cotton, silk and gems (Strabo 2.5.12; Casson 1991: 200-212; Peacock 2002: 432-3). At the same time, Alexandria was receiving from the Mediterranean, for local consumption and for transshipment to the south, various products such as wine, oil, fine pottery, glass, timber, copper and tin. Yet, the role that Lake Mareotis played in this internal transport system is some what unclear.

The different nature of maritime installations located along the northern shore of the lake and those along the southern shore and on Mareotis Island, is also noteworthy. As mentioned earlier, maritime installations such as quays and jetties located the northern shore of the lake are mostly associated with large tell sites of a civic and residential nature. These tells were up to 60,000 m2 in area, and were densely occupied. They are mainly covered with building stones and the foundations of buildings, as well as the remains of several wells, cisterns and red brick basins lined with opus signinum, which could have been used in baths, houses or other urban structures. However, the situation on the southern shore is quite different since the southern shore is where most industrial and commercial sites were recorded, and hence maritime installations were mostly associates with those sites. In antiquity, almost all the amphora and wine production sites in the region, as well as sakkia installations (Empereur & Picon 1998; Rodziewicz 1998b; see also the individual contributions of Blue, Hopkinson and Dzierzbicka this volume), were located on the southern shore of the lake, the focus of agricultural and industrial activities. The reason that agricultural and industrial activities where concentrated on the southern shores of the Mareotic Arm is mainly due to the difference in the topography between the northern and southern shores of the lake. The western arm of Lake Mareotis is delimited from the north and the south by two limestone carbonate ridges, of average elevation 25-35 m and average width 300 m (Said 1990: 499; Warne & Stanley 1993; El-Raey, et al. 1995: 191; Frihy, et al. 1996: 282). The northern ridge is known as the Abusir Ridge, and to the south a longitudinal depression 3-4 km wide known as Al-Alamein-Maryut Depression, extends roughly E-W, partly occupied by the western arm of Lake Mareotis. This depression is delimited to the south by another coastal ridge, known as Gebel Maryut Ridge, which is located 5-9 km south of the Abusir Ridge.

Although it is well known that Lake Mareotis was fed by means of a number of canals, which branched off the Canopic Branch, and flowed into the southern and eastern reaches of the lake, there is a considerable degree of uncertainty about the exact number, location and the routes of these canals. However, the most important of these canals was Schedia Canal (see Bergmann, Heinzelmann & Martin this volume) It bifurcated off the Canopic Branch of the Nile at the town of Schedia, originally a Hellenistic foundation that was later known as Chaereu, currently located in the region of the villages of Kom El-Giza, Kom El-Nashw and Kom El-Hamam, some 30 km south-east of Alexandria (Bergmann & Heinzelmann 2004). While the Canopic Branch continued north to debouch into the Canopic Bay (Abukir Bay), the Schedia Canal turned northwest towards Alexandria and followed a course close to the present course of the Al-Mahmoudeyah Canal. In a statement by Strabo (17.1.16) in which he describes the town of Schedia, he mentions that it has “…the station for paying duty on the goods brought down from above it and brought up from below it; and for this purpose, also, a schedia (float) has been laid across the river, from which the place has its name”. Accordingly, Schedia was the main Nile emporium, customs harbour and checkpoint east of Alexandria, where custom duties were imposed on imported and exported goods (Empereur 1998: 225; Bergmann & Heinzelmann 2004). Moreover, it seems that the Canopic Branch at Schedia was obstructed by some kind of a pontoon that prevented boats from sailing past it until duties were paid on merchandise travelling both ways. Additionally, it was at Schedia where exported commodities

Accordingly, the distance between the Gebel Maryut Ridge and the southern shore of the lake is far greater than the distance between the Abusir Ridge and the northern shore of the lake. In antiquity, the area south of the lake was a fertile plain that flourished with agricultural activities and was known for the quality of its vines, olives and fruits as well as for the cultivation of flax and papyrus (Athenaeus 1.33.d-e; Pliny 13.22.71; Strabo 17.1.14; Empereur & Picon 1998; Horden & Purcell 2002: 353; McGovern 2003: 121-3). Moreover, the abundance of calcareous clay, particularly suitable for amphora production, resulted in a thriving large-scale amphora industry along the southern shore of the lake (Empereur & Picon 1986: 103-9, 1992, 1998; Rodziewicz 1998; Blue & Ramses 2007). Thus, the focus of wine and amphora production in Hellenistic and Roman times was the southern shores of the Mareotic Arm. Thus, it is reasonable to suggest that those maritime installations located along the southern shore of the lake were very much involved in commercial activities including the transport of Mareotic products to Alexandria and possibly also to the southern limits of the lake. 142

E. KHALIL: MAREOTIS’ WATERFRONT brought from upriver were transferred from large Nile boats to smaller boats that could travel easily through the canals to Alexandria (Procopius 6.1.3; Hassan 1997: 365 n. 13).

sequently, the lake’s southern and eastern shoreline was unstable and subject to constant change, and was therefore unsuitable for the establishment of substantial harbours and waterfront installations. A recent survey conducted along the ancient southern and eastern limits of the lake (Wilson 2007; see Wilson this volume) revealed that most settlements established in this area during the Ptolemaic and Roman periods, were located on high ground around the lake’s edge. Also it revealed that many settlements were involved in agriculture and industrial activities mainly in the service of Alexandria. However, as yet there is no evidence for substantial maritime structures or significant waterfront installations.

As it approached Alexandria, the Schedia Canal bifurcated into two branches in the Alexandrian suburb of Eleusis (El-Nozha). The first branch turned towards the north-east leading to Canopus, east of Alexandria, while the other branch continued south of Alexandria and parallel to the lake’s northern shore, until it debouched into Lake Mareotis south-east of Alexandria. According to Strabo (17.1.7), boats also sailed from the Nile to the Canopic Branch and through the network of canals that fed the lake from the south and east, then across the lake northwards to Alexandria. This indicates that navigation on Lake Mareotis was intense and operated in many directions. It also raises a point about the practicalities of sailing in Lake Mareotis from south to north against the prevailing north-westerly wind. The predominant winds along the north coast of Egypt are north-westerly and they prevail more than 40% of the time throughout the year and more than 70% of the time during the summer sailing season (El-Zouka 1979: 125-7; El-Gindy 1999: 17). Accordingly, merchant vessels sailing in Lake Mareotis from south to north would have faced a direct headwind, which meant that the boats had to tack in order to reach Alexandria. Tacking in Lake Mareotis was possible considering the large area of the water body; however, tacking from the southern limits of the lake to Alexandria would have meant that boats would have to travel several times the direct distance across a water body full of shallows and marshlands and against prevailing winds. In the 5th century St. Palladius (7.1) mentioned that he sailed across Lake Mareotis from north to south, from Alexandria to the monastic settlement of Mount Nitria, a distance of about 50 km, in a day and half. Accordingly, sailing in the lake in the opposite direction would have taken much longer, possibly as long as four to five days.

Conclusion There were two ways for river vessels to travel to and from Alexandria, either across the lake, or along the Schedia Canal. Considering the arguments outlined above, particularly in relation to the prevailing winds, it seems reasonable to suggest that the main northbound traffic probably went via the Canopic Branch and the Schedia Canal, rather than across the lake. However, sailing south across the lake would have been conducted with considerable ease. Along those stretches of the canal where boats had to maneouver against the wind, they could have been towed along from the shore, a standard procedure for moving river boats in rivers and canals around the world. In that respect, Strabo’s statement (17.1.7) about the lake harbour south of Alexandria being richer than the seaport of Alexandria, would still be valid. At the time of Strabo, the Schedia Canal debouched into Lake Mareotis, so all the canal traffic had to pass through the lake. Moreover, boats coming from the western arm of the lake also arrived at the lake harbour. Therefore, it is possible that the lake harbour was quite busy receiving river vessels from the south as well as from the west. At the same time, it is not unreasonable to suggest that east–west commercial traffic along the western Mareotic Arm was probably more intense and more regular than the north–south traffic that passed through the main body of the lake. Recent archaeological investigation in the Mareotic region have revealed that the number, nature and extent of archaeological sites along the shores of the Mareotic Arm, is unparalleled any where else in the Western Deltaic Region (see Blue this volume). Settlements in this region were located far from the silting effects of the Nile sediments, the coastline was more stable, prevailing winds were more favourable for east-west movement, and settlements were in close proximity to Alexandria. Therefore, it is understandable why so many shoreline settlements and associated maritime installations were established along its shores. Thus, the contribution of the western Mareotic Arm to the economy of ancient Alexandria and hence of Egypt as a whole, was probably far more significant than any other area along the shores of Lake Mareotis. Thus, the shores of the western arm of Lake Mareotis appear to have been one of the most active areas of economic activity in the Western Deltaic region during the Hellenistic, Roman and Byzantine periods.

Furthermore, the extended period of travel across the lake would have laid boats venerable to another challenge that prevailed on Lake Mareotis in antiquity. Achilles Tatius (4.12) in the 2nd century and Heliodorus (1:14) in the 3rd century, spoke of piracy and bandits on Lake Mareotis. The marshes and islands of the lake provided excellent hideouts for groups of bandits and their vessels. Also, the large size of the lake made it quite difficult to guard and control, therefore, it is possible that sailing across the lake with valuable commodities was quite risky. Moreover, settlements located on the southern and eastern shores of Lake Mareotis were far more susceptible to sedimentation from silts deposited via the nearby Canopic Branch of the Nile, particularly during flood seasons, as well as sediment which had been carried by the prevailing winds across the lake from the north-west to the southeast. All this would have contributed to the build up of sediments against the southern and eastern shores of the lake, thus preventing settlements in this region from being as actively involved in across lake transportation. Con143

LAKE MAREOTIS: RECONSTRUCTING THE PAST Bibliography Achilles Tatius, The Adventures of Leucippe and Clitophon, S. Gaselee (transl.), 1969. London. Athenaeus, Deipnosophistae, C.B. Gulick (transl.), 1953. London. Bergmann, M., & Heinzelmann, M., 2004, Schedia (Kom el-Gizah & Kom el-Hamam, Department of Beheira) Report on the documentation and excavation season. 18 March – 18 April 2003. Paper presented at conference on Alexandria in Antiquity, University of Oxford, December 2004. Blue, L., & Ramses, S., 2006, Lake Mareotis Research Project. Report submitted to the Egyptian Supreme Council for Antiquities on the fieldwork and results of the September 2006 field season. Blue, L., & Ramses, S., 2007, Lake Mareotis Research Project. Report submitted to the Egyptian Supreme Council for Antiquities on the fieldwork and results of the May & July/August 2007 field seasons. Casson, L., 1991, The Ancient Mariners. Princeton. De Cosson, A., 1935, Mareotis: Being an Account of the History and Ancient Monuments of the North-West Desert of Egypt and Lake Mareotis. London El-Ashmawi, F., 1998, Pottery Kiln and Wine Factory at Burg El-Arab. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et ro maine. Actes du colloque d’Athènes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole française d’Athènes, 11-12 décembre 1988: 55-64. Athens, Paris. El-Fakharani, F., 1974, The Lighthouse of Abusir in Egypt. Harvard Studies in Classical Philology 78: 257-272. El-Fakharani, F., 1983, Recent Excavation at Marea in Egypt. In G. Grimm, H. Heinen & E. Winter (eds.), Aegyptiaca Trevernsia II, Das Römisch-Byzantinische Ägypten. Akten des internationalen Symposions 26.-30. September 1978 in Trier, 175-186. Mainz. El-Falaki, M.S., 1966, Ancient Alexandria. Alexandria: Dar Nashr Al-Thaquafa ( ), M.B. Astronome (transl.), 1872. Copenhague. El-Gindy, A., 1999, Meteorological and Hydrodynamic Conditions in the Eastern Harbour of Alexandria and its Vicinity. In Y. Halim (ed.), Proceedings of the Workshop on the Status of Pilot Project for the Sustainable Development of the Submarine Archaeological Sites at Qayetbey Citadel and Eastern Harbour of Alexandria 20-21 November 1999, Reports on Hydro dynamics, Geophysics, Morphology and Geology, V. II: 1-82. Alexandria. El-Raey, M., Nasr, S., Frihy, O., Desouki, S., & Dewidar, Kh., 1995, Potential Impact of Accelerated Sea-Level Rise on Alexandria Governorate, Egypt. Journal of Coastal Research 14: 190-204. El-Zouka, M.K., 1979, Irrigation Areas in the Western Delta: A Geographical Study ( 8´ ), Alexandria (in Arabic). Empereur, J.-Y., & Picon, M., 1986, A La Recherche des Fours d’Amphores. In J-Y. Empereur & Y. Garlan (eds.), Recherches Sur Les Amphores Grecques. Actes 144

du colloque international organisé par le Centre national de la recherche scientifique, l’Université de Rennes II et l’École française d’Athènes (Athènes, 10-12 Septembre 1984): 103-126. Athens, Paris. Empereur, J.-Y., & Picon, M., 1998. Les Atelier d’Amphores du Lac Mariout. In J.-Y. Empereur (ed.), Com merce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athènes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole française d’Athènes, 11-12 décembre 1988: 75-88. Athens, Paris. Empereur, J.-Y., & Picon, M., 1992, La Reconnaissance Maréotide. Extrait des Cahiers de la Céramique Égyptien 3: 145-152. Empereur, J.-Y., 1998, Alexandria Rediscovered. New York. Flaux, C., forthcoming, Geomorphological survey of the south-west shores of Lake Mareotis. In L. Blue & E. Khalil (eds.), A Multidisciplinary Approach to Alex andria’s Economic Past: the Mareotis Case Study. Oxford. Frihy, O.E., Dewidar, Kh.M., & El-Raey, M.M., 1996, Evaluation of Coastal Problems at Alexandria, Egypt. Ocean & Coastal Management 30.2-3: 281-295. Gabel, C., & Petruso, K., 1980, An Environmental and Cultural Study at Lake Maryut, Lower Egypt. Boston University, African Studies Center Working Papers No. 25. Boston. Petruso, K., & Gabel, C., 1982, Marea: A Byzantine Port in Northern Egypt. Boston University, African Studies Center Working Papers No. 62. Boston. Hassan, F.A., 1997, The Dynamics of Riverine Civilization: A Geoarchaeological Perspective on the Nile Valley, Egypt. World Archaeology 29.1: 51-74. Heliodorus. Aethiopica, T. Underdowne (transl.), 1895. London. Herodotus. The Histories, A. de Sélincourt (transl.), 1971. Middlesex. Horden, P., & Purcell, N., 2002, The Corrupting Sea. Oxford. Kadous, E., 2001, Monuments of Ancient Alexandria. ( ) Alexandria (in Arabic). Kucharczyk, R., 2002, Marea 2001: Windowpanes and other Glass Finds. In M. Gawlikowski & W. A. Daszewski (eds.), Polish Archaeology in the Mediterranean XIII: 65-71. Lewis, N., 1983, Life in Egypt under Roman Rule. Oxford. McGovern, P.E., 2003, Ancient Wine: The Search for the Origins of Viniculture. Princeton. Palladius, The Lausiac History. The Monks of Nitria. VII. I. http://www.fordham.edu/halsall/basis/palladiuslausiac.html (accessed February 2010). Peacock, D., 1992, Rome in the Desert: A Symbol of Power. An Inaugural Lecture delivered at the University of Southampton. Peacock, D., 2002, The Roman Period. In I. Shaw (ed.), The Oxford History of Ancient Egypt: 422-445. Oxford. Pliny the Elder, The Natural History, J. Bostock & B.A. Riley (transl.), 1855. London.

E. KHALIL: MAREOTIS’ WATERFRONT Procopius, The Buildings, H.B. Dewing (transl.), 1940. London. Rickman, G., 1971, Roman Granaries and Store Buildings. Cambridge. Rickman, G., 1980, The Corn Supply of Ancient Rome. Oxford. Rodziewicz, M.D., 1990, Taenia and Mareotis: Archaeological Research West of Alexandria. Acta of the First International Colloquium of the Egyptian Society of Greek and Roman Studies 1: 62-80. Cairo. Rodziewicz, M.D., 1998a, From Alexandria to the West by Land and by Waterways. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellén istique et romaine. Actes du colloque d’Athènes, organisé par le CNRS, le Laboratoire de céram ologie de Lyon et l’Ecole française d’Athènes, 11-12 décembre 1988: 93-103. Athens, Paris. Rodziewicz, M.D., 1998b, Classification of Wineries from Mareotis. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athènes, organisé par le CNRS,

145



le Laboratoire de céramologie de Lyon et l’Ecole française d’Athènes, 11-12 décembre 1988: 27-36. Athens, Paris. Rodziewicz, M.D., 2003, Philoxenité, Pilgrimage Harbour of Abu Mina. Bulletin de la Société Archéolo gique d’Alexandrie 47: 27-47. Said, R., 1990, Quaternary. In R. Said (ed.), The Geology of Egypt: 487-507. Rotterdam. Strabo, Geography, H.L. Jones (transl.), 2001. London. Szymańska, H., & Babraj, K., 2008, Marea V.1: Byzantine Marea – Excavations in 2000-2003 and 2006. Krakow. Vörös, G., 2001, Taposiris Magna: Port of Isis. Budapest. Warne, A.G., & Stanley, D.J., 1993, Late Quaternary Evolution of the Northwest Nile Delta and Adjacent Coast in the Alexandria Region, Egypt. Journal of Coastal Research 9.1: 26-64. Wilson, P., 2007, Western Delta Regional Survey. http://www.dur.ac.uk/penelope.wilson/Delta/AbuMatamir.html (accessed February 2010).

LAKE MAREOTIS: RECONSTRUCTING THE PAST

146

S. RAMSES & A. OMAR: OUTRAGE AND DESTRUCTION

Lake Mareotis Research Project. Phases of Outrage and Destruction Sameh Ramses & Ahmed Omar

Interest and Aim of the Project The historical importance of the region of Mareotis during the Ptolemaic and Roman periods, either as an important wine production center, in support of other Egyptian industries (Strabo 17.1.7), or as a resort of the elite (Buttler 1989: 8-12), has attracted a great deal of archaeological attention, as this volume illustrates (El-Fakharani 1983; Empereur & Picon 1998; Rodziewicz 1998). Nevertheless, the region of Mareotis is threatened by modern urban development and many historical sites are subjected to destruction, not only due to its unique location in the vicinity of Alexandria, but also its moderate climate which has established the area as one of the most popular resorts in the region (Abd-Elhady 2008: 177-184). Therefore, the aim of the Lake Mareotis Research Project, conducted between the summer of 2004 until the summer of 2008, was to survey, record and map the platforms, harbours, and any other waterfront constructions that could be found along the northern and southern shores of the western part of Lake Mareotis, and to create a very detailed database for each site, as well as to determine its present state of preservation.

Phases of Outrage and Destruction According to the last survey of 2008 more than 70 archaeological sites have already been noted in the region (Fig. 2). Of course, not all of them had previously been registered, but some had been surveyed or documented.1 However, despite efforts by the Archaeological Department of Alexandria University to discover and protect some of these sites, the full extent of the area, and the variety of remains, both on land and in the water, still remains to be determined and much work needs to be done. This is particularly urgent considering the very immediate threats of modern urban development and various industrial and agricultural activities that are threatening the shores of the lake. Consequently, during the course of the survey, three main factors were identified that are believed to present a particular danger to the archaeology of the region. Fish Farms As a consequence of being inundated by water largely of agriculture runoff and drainage water, the lake water quality has deteriorated making it more difficult for a variety of fish to survive. Simultaneously, many fish-farmers began to appropriate quite large portions from the lake basin and along the shores. In the process of creating fish producing basins, a great deal of dredging and destruction of some of the archaeological sites, both offshore and under water, has taken place. Accordingly, the original shores topography has been deformed, and the dredging leads to

The survey project covers an area of about 40 km long and up to 3 km wide within the western arm of Lake Mareotis, to the west of Borg El-Arab airport, on the Sidi Kerir road, and extends towards El-Hammam city, it also includes Mareotis Island some 3.7 km long and 680 m wide (Fig. 1; see Blue, Hopkinson and Khalil this volume). Fig . 1: General plan from Alexandria to Mareotis (Lake Mareotis Research Project).

1. It is noteworthy that only about nine sites had already been registered by the Supreme Council of Antiquities (SCA), and only six are considered to be SCA properties. Atlas Almoqea Alathariaya, SCA, 2002: No. 3: El-Beheira Governorate; No. 4: The Rest of the Lower Egypt Governorates. Egypt.

147

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig. 2: Map of sites investigated during the Lake Mareotis Research Project (Lake Mareotis Research Project). Industrial Activities Although not as polluted as some parts of Lake Mareotis, the western arm determines a substantial area of the current lake littoral, and is considerably affected by a huge amount of untreated industrial waste and polluted water which flows into the lake on a daily basis. The industrial activities around the shores of the lake not only affect the water quality but in some cases have a direct impact on the

mis-interpretations of its original form. These activities particularly characterise the southern shore of the lake (Fig. 3). It is worth mentioning that some farms have appropriated quite large areas, up to nearly 4 hectares, especially in the vicinity of Mareotis Island, which embraces a great deal of archaeological remains, making access to the island increasingly difficult by cutting away the roads and destroying the routes that lead to the island. 148

S. RAMSES & A. OMAR: OUTRAGE AND DESTRUCTION

Fig. 3: Two examples of fish farm deformation along the southern shore of the lake (Lake Mareotis Research Project). Fig. 4: Lime storing and collecting at Site 44 (Lake Mareotis Research Project).

Fig. 5: Lake wall affected at Site 44 (Lake Mareotis Research Project).

archaeological sites located in the region. Site 44, located on the southwest shore of the lake, provides clear evidence for this negligence. It is directly adjacent to a cement factory, and is used to store the lime for the cement (Fig. 4). The cement mound is huge and it no doubt, amongst other things, covers the western end of a lake wall that extends along the lake shore (Fig. 5), and disregards the existence of additional archaeological features including a circular structure at the far west of the site near the water’s edge, and four square probable wine-basins, that were previously excavated by the SCA. Similar scenes are witnessed further west, for example at the site of Naga El-Mawalik (Site 109). This is an example of an endangered site that still displays aspects of its ancient harbour and enclosure walls, together with many other hills and mounds of pottery. Its location close to the cement factory, means that it is being used to supply the factory with limestone. Parts of this site are now totally inaccessible, with unrecorded archaeological features being surrounded by barbed wire.

On the north shore of the lake, at the western extent of the survey area, Site 214, immediately opposite Naga El-Mawalik, has recently been purchased by the El-Nasr Salt Company. They have acquired two already excavated sites (Sites 214-215) which contain two wineries and a big house of the late Roman period and numerous mounds of pottery (Figs. 6-7). These sites are due to be dredged and leveled to be used as evaporations basins to produce salt. They have only been partially recorded. There are many other places in the area free of archaeological remains that could be used for such purposes. Other sites, particularly those on the northern shore of the survey area, are being used as rubbish tips totally swamped by rubbish from the city and isolated from the main road by means of barbed wire. These sites are distinctive by their bad smell. Urban Development Urban development is considered to be one of the most dangerous factors that affect the archaeological sites of the 149

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig. 6: (above) Site 214: the partially excavated winery (Lake Mareotis Research Project).

Fig. 7: (right) Site 214: part of the fragmented paved floor of a Roman period waterfront building (Lake Mareotis Research Project). region. Vast portions of land have been acquired, many divided into smaller units to be sold and used for housing. Both the local Bedouin and big companies that wield profit and power are responsible for destroying and leveling vast areas at a remarkably rapid rate. The Lake Mareotis Research Project has had direct experience of these actions over the last four years of survey. This can be summarised in the following four case studies:

documentation of the site, we returned to find that the three hills were totally destroyed by the locals and the Hellenistic tomb flattened (Fig. 12). It is worth mentioning that this site had not been previously registered or recorded by the SCA. El-Gamal (Site 204) The situation at the site of El-Gamal was a little different as the site was quite impressive and important, to the extant that it had already been partially recorded by the SCA, and a site guard had been assigned. The destruction of this site was undertaken in stages – first a soccer field was cleared to the east of the tell; the following year another one was cleared on the prohibited area of the archaeological site towards the limit of the lake shores, destroying part of the tell and site boundaries, and so the destruction continued (Fig. 13).

Site 09 A Kibotos harbour, the only well preserved example either on the sea, or on the lake. Nevertheless, the governorate has dedicated an area of about 8 hectares around the site to build a private university without leaving any access to the archaeological site (Figs. 8-9). Accordingly, over time, the site will be destroyed. Ez-Zeraa Elbahry (Site 201) When visited in September 2004, there was a hill of about 5 m in height, which contained several mounds of pottery with traces of building plots everywhere. When we returned in May 2007, the 5 m high hill had disappeared and there were only private property signs and modern partitioning walls visible (Figs. 10-11).

Even those sites renowned for their archaeological importance are also under threat. To the southeast of Taposiris Magna, the extensive seawall and jetty associated with the site are slowly being encroached by small concrete building plot walls, purchased for urban development (see Boussac & El Amouri this volume).

Nagea Ez-Zohorat (Site 202) A similar pattern of destruction was noted at Nagea EzZohorat (Site 202) as was witnessed at Ez-Zeraa Elbahry (Site 201). The site comprised three hills containing amongst other remains, a cistern, a quay, and a rock-cut Hellenistic tomb. We revisited the site in August 2007 and just few days after we had commenced the recording and

Further destruction is caused at many sites by the introduction of agriculture, particularly the cultivation of figs and olives. Cattle are also allowed to graze freely around the whole region, including on established archaeological sites. Moreover, seasonal extraction of reeds in marshy lake side areas permits the access of heavy loaders and trucks to a number of the archaeological sites. 150

S. RAMSES & A. OMAR: OUTRAGE AND DESTRUCTION

Fig. 8: Kibotos harbour extending from the southern shore of the lake (Lake Mareotis Research Project).

Fig. 9: One of the university properties on the land surrounding the Kibotos (Lake Mareotis Research Project).

Fig. 10: Archeological tell divided for private possession (Lake Mareotis Research Project).

Fig. 11: Private property sign and wall established on Site 201 (Lake Mareotis Research Project).

Fig. 12: Flattened Hellenistic tomb at Site 202 during bulldozer work in 2007 (Lake Mareotis Research Project). 151

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig. 13: Two soccer fields erected at the tell boundaries around Site 204 (Lake Mareotis Research Project). Conclusion and Recommendation From the above mentioned, the slow destruction of archaeological sites in Mareotis can be specified as follows: • Cultivating the boundaries of archaeological tell sites and in some cases, on the tells themselves. • Setting up barbed wire around many of the sites to give the impression that they are private property, and when the authorities are notified, the signs are not removed. • Using bulldozers and tractors to rapidly destroy and level the archaeological hills. • Selling archaeological sites, especially the unregistered sites, to private individuals, by employing unorthodox means that appropriate the names of established people, to ensure that legally the sites cannot be retrieved, thus preventing anyone, even the governmental organizations, from taking any action.

Generally, the aim of these suggestions would be to create a project which promotes a particular concept for both the short- and long-term management of the region, which considers its archaeological and historical importance. Such a plan should also aim to promote and preserve Mareotis for the coming generations. This is a great challenge that we all face. Bibliography Abd-Elhady, D., 2008, Derasat fi Asasyat Al-sayaha. Alexandria. Atlas Almoqea Alathariaya 3: El-Beheira Governorate, Egypt. Supreme Council of Antiquities, August 2002. Cairo. Atlas Almoqea Alathariaya 4: The Rest of Lower Egypt Governorates, Egypt. Supreme Council of Antiquities, August 2002. Cairo. Buttler, A.G., Fateh Elarab Lemesr, M.F. Abo Hadid & T. El-Massrein (transl.), 1989. Cairo. El-Fakharani, F., 1983, Recent Excavation at Marea in Egypt. In G. Grimm, H. Heinen & E. Winter (eds.), Aegyptiaca Trevernsia II, Das Römisch-Byzantinische Ägypten. Akten des internationalen Symposions 26.-30. September 1978 in Trier, 175-186. Mainz. Empereur, J.-Y., & Picon, M., 1998, Les Ateliers d’Amphores du Lac Mariout. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athènes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole française d’Athènes, 11-12 décembre 1988: 75-88. Athens, Paris. Rodziewicz, M.D., 1998, Fom Alexandria to the West by Land and by Waterways. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athènes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole française d’Athènes, 11-12 décembre 1988: 93-103. Athens, Paris. Strabo, Geography, H.L. Jones (transl & ed.), 1996. London, New York.

Accordingly, it has been established that the problems of the region are extensive and complex, but once identified can be resolved by the means of establishing a compromise between protecting Mareotis’ archaeological sites, whilst at the same time meeting the increasing needs of development and modernization. Therefore, the following is suggested to overcome the problems: • Coordination between the different regional authorities including those that operate in the interest of archaeology, investment, irrigation, and the governorate, to establish an integrated long term plan, which recognises a satisfactory compromise that identifies the needs of all parties. • Unifying the mapping system of each organization, in order to get acquainted with each others properties, in order to stop the illegal acquisition of land. • Support the Supreme Council of Antiquities, both to provide the facilities needed to control and protect such a vast area, and to implement archaeological legislation. • By encouraging local interest in the importance of the sites both as a source of income, and as an important aspect of cultural heritage. • By promoting the whole region to become an open archeological park in order not to be neglected and misused, otherwise it will soon disappear. 152

S. RAMSES & A. OMAR: OUTRAGE AND DESTRUCTION

153

LAKE MAREOTIS: RECONSTRUCTING THE PAST

154

INDEX

Select Index Places/topography Abu Mina, 41, 54, 67-73, 77-78, 81, 115, 127-130, 137 Abuqir Lake, 13, 15-16, 18 Abusir, 15, 35 Abusir Ridge, 36, 43, 142 Alexandria, 1-4, 7-8, 12-13, 15-16, 18-20, 25-28, 32-33, 35-36, 47, 49, 54-55, 57, 66-67, 69, 71-72, 75-78, 8283, 87, 89, 100, 103, 107-108, 110, 113, 115-116, 123125, 127-128, 135, 137, 141-143, 147 Beheira, 37, 107-115, 119, 122, 124-125, 147 Cairo, 13, 15, 67, 107, 123, 125, 128 Canopic Branch/Canopic Nile, 4, 7, 13, 18, 25, 36, 75, 107-108, 110, 116, 119, 122, 124-125, 135, 142-143 Chaereou, 7 Fayum, 44, 110, 127 Huwaryia/Hauwariya, 68-70, 75, 78 Kanubiye Canal, 108, 114 Kom el Giza (see also Beheira), 107-115 Lake Mareotis/Lake Mariout/Maryut Lake, 1-4, 7-8, 1119, 21-23, 25-28, 30-33, 35, 37, 44, 47-48, 57, 66-67, 72-73, 75-77, 83, 87, 95-96, 119, 121, 123-124, 128129, 135-137, 139-143, 147-152 Libya, 1, 3-4, 8, 75-77, 157 Marea/Mareia (see also Philoxenité), 1-4, 7-8, 14-15, 17, 27-28, 30-32, 35, 37, 44, 47-52, 54-55, 57-61, 66-67, 69-73, 75, 77-83, 91, 100, 128-129, 135-139, 141 Mareotic Arm, 32-33, 135-136, 138, 141-143 Mareotis Island, 28-32, 35, 37, 41-42, 44, 135-136, 139142, 147-148 Maryut Depression, 25, 35, 142 Mediterranean Sea, 11-12, 15-17, 19, 25, 27, 127-128 Naukratis Canal, 123, 125 Nile Delta, 36-37, 44, 83, 107, 110, 135, 137 Nile River, 3-4, 7-8, 11-13, 15-16, 18, 25, 27, 35-36, 39, 44, 50, 54, 66, 75, 83, 95-96, 103, 107-108, 110, 116, 125, 127, 135, 142-143 Nile Valley, 3, 27, 35, 41, 44, 54, 103, 127 Nitria/Nitraria, 8, 13-15, 143 Philoxenité/Philoxenite, 27, 35, 47, 54-55, 67, 69, 71-73, 75, 77-78, 83, 135-137, 139 Plinthine, 4, 14, 67, 87-88 Sais, 119, 124 Schedia, 7, 13, 107-110, 114-116, 125, 129, 142 Schedia Canal, 18, 142-143 Taposiris Magna/Taphosiris, 7-8, 27, 67, 71, 87-88, 92, 101, 136-137, 139, 141, 150 Western Desert, 3, 5, 123-124

El-Fakharani, F., 1-2, 27, 31, 37, 47, 49-53, 69-70, 76, 78, 129 El-Falaki, M., 1, 17-18, 37, 47, 57, 67, 75, 137 Herodotus, 3, 7, 54, 73, 75, 137 Justinian, 7-8, 76 Pliny, 4 Procopius, 7-8, 55 Sophronius, 8 St Menas, 51-52, 54, 78 Strabo, 3-4, 7-8, 18, 35-36, 39, 47, 54, 66, 75, 127, 142-143 Theodosius, 8 General amphora/amphorae, 27-29, 32, 37, 43-44, 58, 62, 72-73, 75, 78, 81-83, 87, 90, 100-103, 113, 115-116, 121, 124, 127, 135-136, 141-142 auger cores/auger survey, 30, 37-38 basilica, 51-53, 75-78, 81-83, 129 bath complex/bath facility, 28, 58, 70-72, 75, 110, 116, 128-129 bridge, 42, 89-92, 100, 107-108, 116, 137 camel, 4-5, 55 canal, 3-4, 7-8, 14-15, 18-19, 21, 27, 35-36, 50, 52, 66, 75, 91, 95-96, 107-108, 110, 114, 116, 121, 123-124, 135, 142-143 causeway, 25, 58, 60, 71, 77, 87, 89-91, 93, 95, 100, 135 cistern, 2, 28, 36-37, 42, 53, 58, 69-70, 72, 80, 136, 142, 150 coins, 51-53, 70, 79, 82, 91, 96, 100-101, 103, 107, 110, 113-115, 128 Coptic Encomium of St Menas, 54-55, 69, 72, 77 evaporation, 37, 149 finewares, 28, 121 fish farms, 19-22, 96, 100, 147, 149 fish-tank, 96-100, 103 floods/seasonal flooding, 4, 16-17, 22, 35-37, 39, 43, 54, 87, 90-91, 95, 124, 137, 143 forts, 49, 124 glass, 5, 25, 27, 52, 66, 75, 121, 142 harbours, 4, 7, 25, 27, 31, 38-39, 47-51, 53-55, 57, 66-73, 75-78, 81, 83, 87, 90-91, 93, 95-96, 98, 100-101, 103, 107-108, 110, 116, 124, 136-139, 142-143, 147, 149151 hypogea/hypogeum, 54, 69, 71 jetty/jetties, 28-29, 32, 37, 39, 41-43, 47, 49, 53-53, 58, 60, 67, 71-72, 75, 77-78, 81, 83, 89, 91, 93-98, 100-101, 103, 136, 138-139, 142, 150 karum, 36, 43, 71 Kibotos/Kibotos harbour, 31, 47-48, 71, 75, 91, 139, 150151 kiln, 31-32, 36-37, 43-44, 70-73, 75, 77-78, 82-83, 127, 130, 136 lake port, 67, 70, 72, 75 lake wall, 30, 32, 39, 43, 139-140, 149

People Athanasios, 77 Christodorus, 8 Claudius Ptolemy/Ptolemy the Geographer, 4, 36, 47, 54 Coste, P., 87, 89-91 De Cosson, A., 1, 35, 67, 75, 87, 89

155

LAKE MAREOTIS: RECONSTRUCTING THE PAST lighthouse, 49, 58, 137 mole, 28, 31, 41, 67, 139 natron, 3, 5, 15 papyri/papyrological documents, 4-5, 8, 96, 110, 130-132 papyrus, 7, 54, 75, 142 Pharos, 3, 137 pilgrim, 54-55, 69, 71-73, 77-78, 82, 137 pilgrimage, 66, 68-72, 78 pilgrimage centre, 67, 71-72 pollution, 27-28 port, 14, 25, 27, 42, 47, 49-52, 66-67, 69-73, 75, 78, 83, 96, 123-124, 135 qadus, 28 quarry, 27, 32, 36, 43, 67, 71, 77, 91, 107, 116, 129, 143

quay, 28-29, 32, 39, 41, 43, 47, 49-50, 66-67, 72, 77, 136140, 142, 150 sakkia/sakia/saqiya/sāqiyah, 28, 30, 32, 37, 41, 76, 79-81, 83, 136, 142 sebakh, 107, 110, 123 sewage, 25, 27, 77, 83 tombs, 2, 17, 32, 49, 54, 69, 71, 111, 116, 150-151 villa, 35, 43, 58, 77-78, 111, 113-114, 116, 124 vineyards, 5, 43, 75, 78, 127, 130-132 wine, 4-5, 25, 27, 43, 53-54, 70, 75, 77-78, 83, 101, 103, 121, 127-132, 142, 149 wine production, 30, 32, 70, 114, 127, 130, 132, 135-136, 142, 147 winery, 32, 43-44, 53, 69-70, 123, 127-132, 135, 149-150

156

ISBN 9781407306544 paperback ISBN 9781407336527 e-format DOI https://doi.org/10.30861/9781407306544 A catalogue record for this book is available from the British Library

BAR

PUBLISHING

Contents CONTENTS LIST OF FIGURES, TABLES AND MAPS INTRODUCTION CONTRIBUTORS

i ii vii viii

1. FAWZ I EL-FAKHARANI: PIONEER EXCAVATOR AT MAREOTIS. Mona Haggag

1

2. THE MAREOTIC REGION IN ANCIENT SOURCES. Mohamed S. Abd-el-Ghani

3

3. A NOTE ON LAKE MAREOTIS IN BY Z ANTINE TIMES. Mostafa El Abbadi

7

4. A STUDY OF THE EVOLUTION OF THE MARY UT LAKE THROUGH MAPS. Ismaeel Awad

11

5. LAKE MAREOTIS RESEARCH PROJECT. Lucy Blue

25

6. THE RESULTS OF A PRELIMINARY SURVEY AT MAREOTIS ISLAND. Dylan Hopkinson

35

7. THE CITY OF MAREA/PHILOXENITÉ . REFLECTIONS ON THE ALEXANDRIA

47

UNIVERSITY EXCAVATIONS, 1977-1981. Mona Haggag 8. MAREA PENINSULA: OCCUPATION AND WORKSHOP ACTIVITIES ON THE

57

SHORES OF LAKE MARIOUT IN THE WORK OF THE CENTER D’ É TUDES ALEXANDRINES (CEAlex, CNRS USR 3134). Valérie Pichot 9. ON INTERPRETATIONS OF ARCHAEOLOGICAL EVIDENCE CONCERNING

67

MAREA AND PHILOXENITE. Mieczyslaw D. Rodziewicz 10. MAREA OR PHILOXENITE? POLISH EXCAVATIONS IN THE MAREOTIC REGION 2000-2007. r s tof a ra

anna

75

ma ska

11. THE LAKE STRUCTURES AT TAPOSIRIS. Marie-Franç oise Boussac & Mourad El Amouri

87

12. SCHEDIA, ALEXANDRIA’ S HARBOUR ON THE CANOPIC NILE. INTERIM REPORT ON

107

THE GERMAN MISSION AT KOM EL GIZ A/BEHEIRA (2003-2008).

Marianne Bergmann, Michael Heinzelmann & Archer Martin

13. RECENT SURVEY WORK IN THE SOUTHERN MAREOTIS AREA. Penelope Wilson

119

14. WINERIES OF THE MAREOTIC REGION. Dorota Dzierzbicka

127

15. WATERFRONT INSTALLATIONS AND MARITIME ACTIVITIES IN THE

135

MAREOTIC REGION. Emad Khalil 16. LAKE MAREOTIS RESEARCH PROJECT. PHASES OF OUTRAGE AND DESTRUCTION.

147

Sameh Ramses & Ahmed Omar

SELECT INDEX

155 i

LAKE MAREOTIS: RECONSTRUCTING THE PAST

List of Figures, Tables and Maps F igures Chapter 4. Ismaeel Awad 1: Maryut Lake region. 2: 1548, V ray portraict de l a v il l e d’ A l ex andrie en Egypte, Pierre Belon du Mans. 3: 1570, detail from A egyptus antiqua, including a focused map of the region of Alexandria, Abraham Ortelius. 4: 1588, detail from A fricae T abul a V I I , Livio Sanuto. 5: 1655, detail from R oyaume et dé sert de B arca et l ’ Egypte, Nicolas Sanson D’ Abbeville. 6: 1717, detail from Carte de l a B asse Egypte et du cours du N il , Paul Lucas. 7: 1722, detail from Carte de l ’ Egypte A ncienne, Claude Sicard. 8: 1743, detail from Carte d’ Egypte, Richard Pococke. 9: 1753, detail from Carte de l ’ Egypte ancienne et moderne, Robert de Vaugondy. 10: 1764, detail from Carte des Embouchures du N il , e t partie de son cours, Jacques Nicolas Bellin. 11: 1801, Carte d’ A l ex andrie, French campaign. 12: 1801, Carte d’ A l ex andrie, French campaign, original scale 1:100,000. 13: 1866, Carte des Env irons d’ A l ex andrie, Mahmoud Bey El-Falaki. 14: 1866, Carte des Env irons d’ A l ex andrie, Mahmoud Bey El-Falaki, original scale 1:200,000. 15: 1911 (1914), A tl as of Egypt, V ol . I , L ow er Egypt, Sur v ey of Egypt (G. Daressy), original scale 1:50,000. 16: 1949, map series 1:100.000. Surv ey of Egypt, original scale 1:100,000. 17: 1970s, map series 1:25.000, Egyptian General Survey Authority, original scale 1:25,000. 18: 1991, map series 1:50.000, Egyptian General Survey Authority. 19: 2001, Landsat satellite image, resolution 15m/pxl. 20: Evolution diagram of Maryut Lake’ s surface areas (1801-2001). 21: Comparison between Maryut Lake 1801, French campaign map series 1:100,000 and 0 m level in the Mareotic region. 22: Comparison between Maryut Lake 1801-2001. Chapter 5. Lucy Blue 1: Location of Alexandria and Lake Mareotis. 2: An approximation of the ancient and present limits of Lake Mareotis. 3: Sites located along the northern shoreline, the southern shoreline and on Mareotis Island. 4: Topographic, ceramic and auger survey plan of Site 44 on Mareotis Island. 5: An example of a ‘ lake wall’ feature (Site 44). 6: The Kibotos site (Site 09). 7: The ‘ complex building’ (Site 13).

11 12 12 13 13 14 14 14 14 15 16 16 17 17 18 18 19 20 20 21 22 23 26 27 29 30 30 31 31

Chapter 6. Dylan Hopkinson 1 he lo ation of areotis sland and the sites mentioned in te t, modified from e osson 1935 2: Schematic of the main features of Mareotis Island. 3: The littoral features mentioned in this paper and topographic survey of the eastern settlement of the island. 4: The littoral features mentioned in this paper and topographic survey of the western settlement of the island. 5: Longitudinal and transverse schematic sections through the ‘ water tower’ on Mareotis Island.

42

Chapter 7. Mona Haggag 1: Kibotos harbour structure on the south shore of Lake Mareotis (Site 9 – L ake Mareotis Research Project). 2: Kibotos harbour structure, mooring ring. 3: Rock-cut Kibotos harbour structure at Ikingi Mariut. 4: Byzantine harbour of Marea. 5: Plan of the eastern harbour of Marea, after El-Fakharani 1977. 6: Decumanus between the eastern and middle ports of Marea. 7: Plan of shops off the main decumanus. 8: Basilica, Marea.

47 48 48 50 51 51 51 52

ii

35 36 38 40

9: Enigma Buliding, Marea. 10: Winery excavated by El-Fakharani 1983.

52 53

Chapter 8. Valérie Pichot 1: Map of Egypt: location of Lake Mariout and the site of Marea. 2 eneral lan of the site identified as area 3: The peninsula of Marea. 4: A) pier to the north of Marea peninsula, view from the north-west. B) ancient causeway connecting the peninsula to the mainland, view from the west. 5: Ancient causeway under excavation of Sector 1, view from the east. 6: Ancient causeway under excavation of Sector 2, view from the north. 7: Landmark and jetty to the south of the peninsula, view from the south-west. 8: Results of the geophysical survey undertaken on the Marea peninsula. 9: Sector 4 - Hearths in FS4024, under excavation. 10: Overall plan of the western zone of Sector 3. 11: Overall plan of the eastern zone of Sector 3. 12: Eastern zone of Sector 3 under excavation. 13: Sector 3 - One of the two limestone sphinxes discovered in the tower building ST300. 14: Sector 3 - Bronze candelabra after restoration in the CEAlex laboratory. 15: Sector 3 - Hoe chalk burner. Chapter 9. Mieczyslaw D. Rodziewicz 1: Central part of the plan of Mareotis drawn by Mül ler-Wiener in 1966, with the pilgrimage road from the lake harbour to Abu Mina and settlements, among them settlement “ M” in Huwaryia village. 2: Abu Mina with surrounding settlements/karms, and pilgrimage road towards the lake harbour. 3: Byzantine House/Hospitium at the pilgrimage road from the lake harbour to Abu Mina. 4: Byzantine pilgrimage harbour/Philoxenite (Nos. 5-17) and western/coastal rural remains (Nos. 1-4). On the eastern side, structured causeway (No. 18). Chapter 10. r s tof a ra anna ma ska 1. Remains of structures along the southern shore of Lake Mareotis. 2. Marea. Byzantine baths. 3. Marea. Baths and western courtyard. View from the south-west. 4 area ell o erated with s i ah iew from the so th 5. Marea. Funerary chapel. View from the west. 6. Marea. Plan of the basilica. 7. Marea. Apse of the basilica. View from the west. ‘ A’ and ‘ b’ locate the two burials. 8. Marea. Basilica. Amphora kiln. View from the west. Chapter 11. Marie-Franç oise Boussac & Mourad El Amouri 1. General map of the site and location of areas under study at Taposiris Magna and Plinthine. 2. Drawing by Pascal Coste, made in 1820. 3. Map of the remains of the ancient city of Taposiris by A. De Cosson, 1935. 4. Location of the different sectors of the site of Taposiris that were studied along the lake shore. 5. View of the Taposiris bridge from north. 6. Plan and elevation of the north pile of the bridge, November 2004. 7. Model and reconstruction of the bridge in Taposiris Magna. 8. Overall plan and detailed outline of the eastern jetty. 9. The southern end of the eastern jetty. 10. Jetty at Gamal. View from north-west. 11. Jetty at Q useir. View from south-east. 12 lan and ele ation of the inner so th fa e of the north fl sh and ttress 13 iew of the north fl sh d ring e a ation, from east 14. Lay out, elevations and sections of north and south walls of the water supply channel through the eastern jetty. 15. Water supply channel through the eastern jetty opening into the hoop-shaped basin. 16. Grooves and notches in the water supply chanel. 17 iew of the hoo -sha ed asin, the water s l hannel and the fish-tank iii

57 58 59 60 60 60 60 61 62 63 64 65 65 65 66 68 68 70 71

76 78 79 80 80 81 82 82 88 89 89 90 92 92 92 93 93 94 94 94 95 97 98 98 99

LAKE MAREOTIS: RECONSTRUCTING THE PAST 18. 19 20. 21

The pavement in the hoop-shaped basin. he o ening etween the hoo -sha ed asin and the fish-tank Main types of ceramics found in Taposiris Magna (Sector 9) during the Late Roman era. erami s from the filling of the hoo -sha ed asin

Chapter 12. Marianne Bergmann, Michael Heinzelmann & Archer Martin 1: Alexandria and its hinterland in 1925, Surv ey of Egypt, sheet 4. 2: Schedia. Topographical map based on quickbird satellite image (2005) indicating excavated areas (A) and corings (H). 3: Schedia. Map with reconstruction of the ancient topography. 4. Schedia. Plan of the southeastern settlement (Kom el Giza, Kom el Hamam) with excavation areas 1980-92 and 2003-06. 5: Kom el Giza, Area 3. Plan of Hellenistic baths. 6: Kom el Giza, Area 1. Plan of Roman villa, tombs and later basins. 7: Kom el Hamam. Plan with excavations areas 2003-2006. German Mission at Kom el Giza/Beheira. 8: Kom el Hamam. Plan of Area 6 with Roman storage building and later phases. 9: Kom el Hamam. Reconstruction of the Roman storage building. 10: Kom el Hamam, Area 9. Roman enclosure wall. 11: Kom el Hamam. Plan of Area 8 with basins for wine-production.

99 99 101 102 108 109 109 110 111 111 112 112 113 114 115

Chapter 13. Penelope Wilson 1: Ancient sites in the Delta, highlighting the Mareotis region. 2: Map of the sites surveyed, with 0 metre sea-level contour area highlighted in grey. 3: The main mound at Kom el-Mahar. 4: Late Ptolemaic to Early Roman pottery sample from Kom el-Mahar. 5: Red brick structures at Kom Trugi, excavated below foundation level. 6: Section through the cemetery mound at Kom el-Q adi.

119 120 121 122 123 123

Chapter 14. Dorota Dzierzbicka 1: A map of the Mareotic region with marked locations of wineries. 2: A schematic drawing of a Type 4 winery with names of its elements used in Greek papyri from Egypt.

127 130

Chapter 15. Emad Khalil 1: The approximate ancient and present limits of Lake Mareotis. 2: The Lake Mareotis Research Project survey area along the shores of the western Mareotic Arm. 3: Taposiris Magna tower. 4: The middle quay of the harbour of Marea. 5: The quay of Marea harbour was made of large regular limestone blocks without evidence of the use of mortar. 6: The box-shaped harbour on the southern shore of Lake Mareotis. 7: One of the quays extending into the lake at the northern shore of Mareotis western arm. 8: A lake wall extending parallel to the southern shore of the lake. Evidence for red mortar (opus signinum) can still be seen between the blocks. 9: Some of the multi-room square buildings located at Site 118 along the northern shore of the Mareotis Island. The buildings could have been used for storage purposes. Chapter 16. Sameh Ramses & Ahmed Omar 1: General plan from Alexandria to Mareotis. 2: Map of sites investigated during the Lake Mareotis Research Project. 3 wo e am les of fish farm deformation along the so thern shore of the lake 4: Lime storing and collecting at Site 44. 5: Lake wall affected at Site 44. 6: Site 214: the partially excavated winery. 7 ite 214 art of the fragmented a ed floor of a oman eriod waterfront ilding 8: Kibotos harbour extending from the southern shore of the lake. 9: One of the university properties on the land surrounding the Kibotos. 10: Archeological tell divided for private possession. 11: Private property sign and wall established on Site 201. 12: Flattened Hellenistic tomb at Site 202 during bulldozer work in 2007. 13 wo so er fields ere ted at the tell o ndaries aro nd ite 204 iv

135 136 137 138 138 139 140 140 141

147 148 149 149 149 150 150 151 151 151 151 151 152

T abl es Chapter 4. Ismaeel Awad 1: Comparison table between different periods of Maryut Lake.

21

Chapter 13. Penelope Wilson 1: Summary table of sites surveyed by the ‘ Sais and its Hinterland Project’ in the region of southern Lake Mareotis.

124

Chapter 14. Dorota Dzierzbicka 1: Installations forming part of a wine-making complex mentioned in Greek papyri containing references to immovable property found in vineyards or to the appurtenances of a winery.

131

Maps Chapter 7. Mona Haggag 1: Marea in Lake Mareotis. 2: Marea’ s public buildings on the lake shore.

48 50

v

LAKE MAREOTIS: RECONSTRUCTING THE PAST

vi

Introduction he following olle tion of a ers re resents the final s nthesis of a onferen e entitled T he I nternational Conference on the A rchaeol ogy of the Mareotic R egion. L ake Mareotis: R econstructing the P ast hosted by the University of Alexandria, Egypt between 5th & 6th April 2008. The objective of the conference was to bring together scholars that work in the Lake areotis region of le andria to a for m where the o ld resent and dis ss their ongoing resear h and field ro e ts The two-day conference was a great success as it provided an opportunity to share data and tackle themes and issues of common concern. It was also an opportunity for the Lake Mareotis Research Project, a collaboration between the Centre for Maritime Archaeology, at the University of Southampton, and the Department of Underwater Antiquities (DUA) of the Supreme Council of Antiquities (SCA), Alexandria, in conjunction with the Centre for Maritime Archaeology and Underwater Cultural Heritage, University of Alexandria, to show case the Lake Mareotis survey project and present recent results to colleagues. Lake Mareotis is a shallow body of brackish water currently about 90 km2 in area that extends south of Alexandria for about 12 km and to the west in the form of an arm some 40 km and is 3 km wide. Lake Mareotis was much larger in antiquity having been subject to siltation and reclamation over the last two thousand years. Due to its location and chara teristi s, ake areotis it is known to ha e ontri ted signifi antl to the an ient e onom of le andria and to the internal and international trans ort s stem of re o- oman g t he signifi an e of this region has een a knowledged since at least the 1930’ s (De Cosson 1935), and subsequently highlighted by the pioneering work of scholars such as ElFakharani (1983). This volume is published in honour of these early pioneers, in particular El-Fakharani who not only initiated s holarl in estigation of a n m er of sites in the ake areotis region t who also engaged for the first time, future Egyptian scholars by including and encouraging students of Archaeology from the University of Alexandria to participate in excavations at Marea (see Haggag this volume). Thus, the conference and this publication attempts to follow in the footsteps of El-Fakharani by seeking to encourage an inclusive dialogue between all scholars currently active in the region. The following papers represent the most recent attempt to contextualise and interpret Lake Mareotis. Current research, not only that generated as a product of the Lake Mareotis Research Project (Blue & Khalil, forthcoming), but also exciting new discoveries from other sites around the shores of the lake, are beginning to provide a more comprehensive understanding of the breadth and scale of activities conducted around the shores of Lake Mareotis and highlight the substantial contribution the region made to the ancient economy of Alexandria. There is still much to be discovered but with continued dialogue we can begin to tackle unanswered questions and resolve sometimes contradictory, interpretations. This volume therefore presents a collection of thoughts and perspectives that are often variable in their focus and approach. A degree of editorial consistency has been em lo ed t in order to retain the arti lar and di erse ers e ti es of the indi id al a thors, s e ifi nomen lat re has been maintained. To that end, a variety of spellings for Lake Mareotis (Maryut, Mariout) and sites in the Mareotic region such as Marea, Mareia etc., are supported. It is hoped that this volume will encourage future discussion and publications that continue to highlight the importance of the Lake Mareotis region in antiquity. Lucy Blue Ack now ledgements he editor wo ld like to a knowledge oth the ritish adem and the e erh lme r st who ro ided finan ial s ort to undertake the Lake Mareotis Research Project of which this volume is just one product. In addition, the support of Dr Emad Khalil and the University of Alexandria is acknowledged for hosting and organising the conference. I would also like to thank the anonymous reviewer of this volume for their helpful comments. Bibliography El-Fakharani, F.A., 1983, Recent Excavation at Marea in Egypt. In G. Grimm, H. Heinen & E. Winter (eds.), A egyptiaca T rev ernsia I I , D as R öm isch- B yz antinische Ä gypten. A kten des international en Symposions 26.30. September 1978 i n T rier, 175-186. Mainz. Blue, L., & Khalil, E. (eds.), forthcoming, A mul tidiscipl inary approach to A l ex andria’ s economic past: T he Mareotis case study. Southampton Monograph Series. Oxford De Cosson, A., 1935, Mareotis: B eing an A ccount of the H istory and A ncient Monuments of the N orth- W est D esert of Egypt and L ake Mareotis. London vii

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Contributors M arie- F ranç oise Boussac Professor in Greek History Paris Ouest Nanterre University 92000 Nanterre, France marie-francoise.boussac@ mom.fr

M ostafa E l Abbadi Professor Emeritus of Greco-Roman Studies Faculty of Arts University of Alexandria Hussein Sobhi St. El-Shatbi 21526 Alexandria, Egypt mostafaabbadi@ yahoo.com

Delphine Dixn euf Institut Franç ais d’ Archéologie Orientale 37, rue al-Cheikh Aly Y oussef B.P. Q asr al-Ayni 11562 11441 Cairo, Egypt ddixneuf@ ifao.egnet.net

M ohamed S . Abd- el- G hani Professor of Ancient History Dept. of Archaeology and Greco-Roman Stiudies Faculty of Arts University of Alexandria Hussein Sobhi St. El-Shatbi 21526 Alexandria, Egypt

Dorota Dz ierz bick a Department of Papyrology Institute of Archaeology University of Warsaw rakowskie r edmie ie 26/28 00-927 Warsaw 64 Poland d.dzierzbicka@ gmail.com

M ourad E l Amouri Ipso Facto Bureau d’ É tude et de Recherche, Archéologie et Océanographie 4 rue de Tilsit 13006 Marseille, France mourad.el-amouri@ ips-o.fr

T homas F aucher ANR-Nomisma Université Paris-Sorbonne-Paris IV 1, rue Chevalier de la Barre 75018 Paris, France thomas.faucher@ paris-sorbonne.fr

Ismaeel Aw ad Topography Department Centre d’ É tudes Alexandrines 50, rue Soliman Y ousri 21131 Alexandria, Egpyt ismaeel_a wad@ hotmail.com

M ona H aggag Bibliotheca Alexandrina P.O. Box 138 El-Shatbi 21526 Alexandria, Egypt Mona.Haggag@ bibalex.org

K rz ysz tof Babraj Archaeological Museum in Kraków ul. Senacka 3 31-002 Kraków , Poland Krzysztof.Babraj@ ma.krakow.pl

M ichael H einz elmann University of Cologne Institute of Archaeology Albertus Magnus Platz D-50923 Cologne michael.heinzelmann@ uni-koeln.de

M arianne Bergmann Georg-August-University Goettingen Archaeological Institute Nikolausberger Weg 15 D-37073 Goettingen marianne.bergmann@ phil.uni-goettingen.de

Dylan H opk inson Centre for Maritime Archaeology University of Southampton ighfield Southampton, England SO17 1BJ dylan.hopkinson@ btinternet.com

L ucy Blue Director, Centre for Maritime Archaeology University of Southampton ighfield Southampton, England SO17 1BJ L.Blue@ soton.ac.uk

viii

S ameh Ramses Supervisor of Egyptian Excavations & Surveying Team Department of Underwater Archaeology Supreme Council of Antiquities (SCA) 19 Abdel Moety el Gazaly, Stanley Alexandria, Egypt samehramses@ yahoo.com

E mad K halil Centre for Maritime Archaeology & Underwater Cultural Heritage Faculty of Arts University of Alexandria Hussein Sobhi St. El-Shatbi 21526 Alexandria, Egypt emadkhalil@ foa.edu.eg

M iecz yslaw D. R odz iew icz Ul. Literacka 25 m.10 01-864 Warsaw, Poland m.rodziewicz@ gmail.com

Archer M artin American Academy in Rome Via Angelo Masina, 5 I-00153 Roma archer.martin@ alice.it

anna S a s a Research Centre for Mediterranean Archaeology University of Warsaw 11, Sharia Mahalla, Heliopolis Cairo, Egypt leukaspis1@ interia.pl

Ahmed Omar Inspector, Department of Underwater Archaeology Supreme Council of Antiquities (SCA) 19 Abdel Moety el Gazaly, Stanley Alexandria, Egypt

Penelope W ilson Department of Archaeology Durham University South Road Durham, England DH1 3LE Penelope.Wilson@ durham.ac.uk

V alé rie Pichot Centre d’ É tudes Alexandrines 50, rue Soliman Y ousri 21131 Alexandria, Egpyt valerie.pichot@ cea.com.eg

ix

LAKE MAREOTIS: RECONSTRUCTING THE PAST

x

Fawzi El-Fakharani: Pioneer Excavator at Mareotis Mona Haggag

This volume is dedicated to the memory of Fawzi El-Fakharani who was, in many ways, a pioneer investigator of the area of Mareotis.

certain or clear that it was a temple dedicated to Hercules (Kanellopoulos 1994: 60-65). In 1970 Fakharani was delegated to Libya, where he established a department of Classical Archaeology. In Libya, Fakharani excavated the area of the Byzantine palace of the City of Tocra, ancient Taucheira. The patriarch, who presented the city at the council of Nicea in 325 CE, must have resided in this palace. Fakharani uncovered parts of the Eastern Church to which the palace was attached. He also nearthed the h r h s onfirmatori m and a tister , as well as two beautiful mosaic panels (pers. comm.).

Fawzi El-Fakharani was born in Alexandria in 1921. He graduated from the Faculty of Arts, University of Alexandria in 1946 and directly after his graduation he went to the United Kingdom for higher studies. In 1953 he obtained a Masters degree from the University of Liverpool. His undergraduate dissertation was entitled “ The origins of stucco reliefs in Classical Art” . Afterwards he moved to University College London where he acquired a doctoral degree in 1957 on “ Stucco Reliefs in Roman Art” .

In 1973, back in Alexandria, Fakharani began his studies of the city of Marea, thus realising his life’ s dream, which was to discover and document the ruins of this city. After a long and e ha sti e eriod of re aration, he finall o tained the ermission to e a ate the site identified reviously as the city of Marea, to the south of the western extremity of Lake Mareotis, some 45 km distance from le andria ith the e e tion of re io s identifi ations of the site as that of Marea made by Mahmoud Pasha ElFalaki (1872: 85, 86, 91, 96, 97, 101), and the valuable monograph on Mareotis by De Cosson (1935: 131-135), the academic community of archaeologists and researchers had paid no particular interest in the area of Mareotis. Fakharani’ s determined enthusiasm and efforts to identify the city’ s area proved successful, as the results of his excavations shed light on the various stages of life within the city and on the importance of the area as a whole (El-Fakharani 1983: 175-204). Some of the articles of this volume are devoted to Marea (see Haggag, Pichot, Rodziewicz, Babraj and ma ska thro gh whi h the reader an fig re o t that the discovery of Marea was and is, of great importance to researchers, as well as to the community at large.

Fakharani returned to Egypt and was appointed lecturer in the University of Alexandria in 1958. In 1964 he was promoted to an assistant professor, and won the chair of Professor of Classical Archaeology in 1974. He was appointed Chair of the Department of Classical Civilization between 1976 and 1980, and the director of the Institute for Mediterranean Studies between 1986 and 1989. In 1990 he became Chair of the Department of Archaeology at the University of Tanta, Egypt. In 1966 Prof. Fakharani was delegated to the University of mman in ordan here, he egan his first endea o rs in ar haeologi al fieldwork is e a ation target was to ontinue uncovering the grand Roman theatre of Amman. He published the results of his investigations in the J ahrbuch D es D eutschen A rchäol ogischen I nstituts A rchäol ogischen A nz eiger, in 1975 (El-Fakharani 1975a: 377-403). In 1967, his interests shifted towards the site of Q al’ at AlJabal on which stands the so-called Temple of Hercules (El-Fakharani 1975b: 533-554). There he reinvestigated the remains and uncovered most of the building’ s foundations and annexed colonnades. Scholars differed with regard to which deity the temple was dedicated, yet Fakharani’ s hypothesis was that the building had characteristic features unknown to temples of the East Roman Provinces in the time of Marcus Aurelius, although common in libraries of the same epoch in the neighbouring countries (El-Fakharani 1975b: 554). Fakharani reached the conclusion that the building was not a temple but the library built by Marcus Aurelius at Philadelphia, the city that he described as Alexandria’ s daughter city. Despite his reasonable arguments and study published in the periodical of the University of Rostock, the building was still termed as the “ Temple of Hercules” . This remained the case until a study conducted in 1994 of an inscription carved on the architrave of the building showed that it was not at all

With only ten thousand Egyptian pounds as a budget, Fakharani was given the daunting task of locating a suitable area to undertake a bold and pioneering step in Egyptian archaeology, his goal being to accommodate both male and female g tian st dents in the field his was something that had never been undertaken before and had to be managed with great care and respect for social traditions. Not only did he have to secure separate sleeping and study areas but he also had to supply electricity to the camp, food and sustenance for the students, and transportation to the desert area of Mareotis. All of this he accomplished on a meagre dget esides the diffi lties of o taining ermission from the ni ersit to take the st dents to the field, he had also to convince their parents, especially those of the female students, that their daughters would be safe and 1

LAKE MAREOTIS: RECONSTRUCTING THE PAST secure. It was not easy for Egyptian families during the 1970s to send their da ghters to the desert for field training, but Fakharani managed to allay any of the families fears and worries.

German team. They all emphasised that the remains from the Ptolemaic era lay only in the area of the nursery which then belonged to the Faculty of Agriculture, the University of Alexandria. Again, it took determination and patience to acquire the permission to excavate this site. During this time, Fakharani excavated a few trial trenches in the area around the remains of an alabaster tomb. Although the trenches yielded nothing of conspicuous importance, a Byzantine well just next to the tomb shed new light on the use of this area during this period. As the level of the monolithic pavement of the early Ptolemaic tomb was higher than that of the Byzantine rim of the uncovered well, it became evident that the tomb was not in situ as Adriani had assumed (Adriani 1940: 15-23), and that it had been moved to its present location during the Byzantine or early Islamic periods (author’ s interpretation).

Soon after the ancient buildings emerged from the desert sands, articles and interviews in local and international media started to appear. These articles helped win the support of the families of the students who had been allowed to camp in the desert with Fakharani on this pioneering expedition. The student families soon realised from these articles and media reports just how important the work was that their young daughters and sons were accomplishing. Thus, Prof. Fakharani overcame all obstacles and all the students, including myself, who participated in the e a ations, were, for the first time in the histor of the department, to become the lucky generation of students who had the o ort nit to e trained in the field We are extremely grateful to Prof. Fakharani for that experience. The directors of the Lake Mareotis Research Project, Lucy Blue, Sameh Ramses and Emad Khalil (to whom this volume owes its inspiration), are now reminding us of those wonderf l da s in ol ing st dents in their rrent fieldwork s r e ing the shores aro nd ake areotis, the first time this has happened since the pioneering days of Fakharani.

At the same time that Prof. Fakharani received the permission for his excavation of the nursery, Parkinson’ s disease had already begun digging its deadly roots into his aging body. Accordingly, he entrusted the project to another great scholar who has also made considerable contributions to the archaeology of Alexandria and its environs, namely Jean-Y ves Empereur. In March 2004, Fakharani passed away but he left behind a generation of empowered, highly motivated, committed and devoted students who strive to continue their work to the standards and example set by Fawzi El-Fakharani.

Prof. Fakharani never once lost sight of the idea that Marea was the “ Pompey of the East” , to use his own expression (El-Fakharani 1977: 5). It deserves more care and diligence as it represents a unique archaeological site that gives us a complete picture of the everyday life of ordinary people, in contrast to other archaeological projects in Egypt which have a tendency to focus on discovering how our once noble Pharaonic families lived and prepared for the afterlife. Marea, together with the results of survey and excavations from other settlements around the shores of the Lake (see Blue, Hopkinson, Boussac & El Amouri, Bergmann, et al., Wilson, Dzierzbicka, Khalil, and Ramses & Omar this volume), illustrate how people in a postPharaonic era lived, what their beliefs were, what their baths looked like and how they were used, the shape and layout of their shops, cisterns, and how they buried their dead. Marea necessitates more preservation, conservation, restoration and preparation efforts to emphasise the site’ s otential, in order to amongst other things, la e it firml on the tourist map of Egypt.

Bibliography Adriani, A., 1940, Fouilles et decouvertes. Alexandrie. I. Tombeau en alabaster du cimetiere latin. A nnuaire 3, 1935193 : 15-23. Alexandria. De Cosson, A., 1935, Mareotis: B eing an A ccount of the H istory and A ncient Monuments of the N orth- W est D esert of Egypt and L ake Mareotis. London El-Fakharani, F.A., 1975a, Das Theater von Amman in Jordanien. J ahrbuch des D eutschen A rchä ol ogischen I nstituts A rchaol ogischer A nz eiger 3: 377-403. El-Fakharani, F.A., 1975b, The Library of Philadelphia (? ) Or The So-Called Temple on the Citadel Hill in Amman. W issenchaftl ische Z eitschrift der U niv ersität R ostok XXIV.6: 533-554. El-Fakharani, F. A., 1977, A P rel iminary R eport on the 1s t Season of Ex cav ation at Marea. Excavation Report presented to the Antiquities Department, Egypt. El-Fakharani, F.A., 1983, Recent Excavation at Marea in Egypt. In G. Grimm, H. Heinen & E. Winter (eds.), A egyptiaca T rev ernsia I I , D as R öm isch- B yz antinische Ä gypten. A kten des international en Symposions 26.30. Se ptember 1 978 i n T rier, 175-186. Mainz. El-Falaki, Mahmoud Bey, 1872, Memoire sur l ’ antique A l ex andrie, s es faubourgs, s es env irons. Copenhagen. Kanellopoulos, C., 1994, The Great Temple of Amman, Vol. I: Architecture. A merican Center of O riental R esearch – A mman: i-xii.

inall , in 1999, akharani again took s to the field his time akharani s field was in le andria itself is in estigations were of the Latin cemetery which he believed to have been the site of the royal Ptolemaic burial ground. A geophysical survey was conducted in three successive hases firstl , olleag es from le andria ni ersit , secondly, in collaboration with Mrs. Calliope Lemniou Pappacosta, a researcher from Greece in consultation with experts from the University of Patras, and thirdly, by a

2

M. ABD-EL-GHANI: ANCIENT SOURCES

The Mareotic Region in Ancient Sources Mohamed S. Abd-el-Ghani

and on the south Lake Marea, also called Mareotis: “ This is filled man anals from the ile, oth from a o e and on the sides …” (Strabo 17.1.7).

Before addressing the Mareotic region in antiquity it would be appropriate to explore how Egyptian frontiers are discussed in the classical sources. There is a crucial and decisive paragraph in Strabo (17.1.5) to this effect – one that makes particular reference to the western frontiers of Egypt – the primary focus of this paper – and is quoted “ verbatim” as follows:

What is particularly interesting to note about the inhabitants of the Mareotic region, and the Western Desert of Egypt in general, is the ethnic distinction that characterises them apart from the rest of the Egyptians, even after their identity as Egyptians was settled from an administrative viewpoint. It is a phenomenon that is observed throughout the ancient texts through to the present day. This contrast between the Mareotes and the Egyptians is explicitly expressed in a document dated from the middle of the 4th century AD about the smuggling of natron.1

“ Now the early writers gave the name Egypt to only the part of the country that was inhabited and watered by the Nile, beginning at the region of Syene and extending to the sea t the later writers down to the resent time have added on the eastern side approximately all the parts between the Nile and the Arabian Gulf (i.e. the present Red Sea), and on the western side the parts extending as far as the oases, and on the sea-coast the parts extending from the Canobic mouth to Catabathmus (the present el-Salloum) and the domain of the Cyrenaeans. For the kings after Ptolemy became so powerful that they took possession of Cyrenaea itself and even united Cyprus with Egypt. The Romans, who succeeded the Ptolemies, separated their three dominions and have kept Egypt within its formers limits.”

Marea is stated by Herodotus as being located at the key position with regard to the Egyptian frontiers to the west. He relates that king Psammetichus I of the 26th Dynasty posted garrisons at Elephantine on the side of Ethiopia, at Daphnae of Pelusium on the side of Arabia and Assyria, and at Marea on the side of Libya (Herodotus II.30). Marea was also mentioned by other classical authors as the battlefield of some it hed attles in anti it efore the onquest of Egypt by Alexander the Great. In his reference to the second Egyptian revolt against the Persian rule under King Artaxerxes (c. 460 BC), Thucydides pointed out that Inaros son of Psammctichus, “ setting out from Maria, the city just north of Pharos, caused the greater part of Egypt to revolt from king Artaxerxe” (Thucydides I.104).

Among the early writers referred to above by Strabo, it was Herodotus (II.17) who adopted the ancient theory that Egypt was the land inhabited by Egyptians and was watered by the Nile. Nevertheless, he considered that the Egyptians were not only the inhabitants of the Nile Valley and the Delta proper, particularly the area that was directly watered by the Nile River, but also those places to which the Nile waters extended. In order to illustrate this, he set o t to write a signifi ant a o nt a o t the inha itants of the cities of Marea and Apis. His account, in this concern, goes as follows:

Diodorus of Sicily reported that the decisive pitched battle between King Apries, the legitimate king (588-566 BC) of the Saite Dynasty, and Amasis II, one of the king’ s generals and claimant to the throne who became a renowned king, took place near the village of Maria (Diodorus Siculus I.68),2 after the former’ s defeat by the Greeks of Cyrenê and Barcê .

“ The men of the cities of Marea and Apis, in the part of Egypt bordering on Libya, thinking themselves to be not Egyptians but Libyans, and misliking the observance of the religio s law whi h for ade them to eat ow s flesh, sent to Ammon saying that they had no part or lot with Egypt: for they dwelt outside the Delta and did not consent to the ways of its people, and they wished to be allowed to eat of all food. But the god forbade them: all the land, he said, watered be the Nile in its course was Egypt and all who dwelt lower than the city Elephantine and drank of that river’ s water were Egyptians.” (Herodotus II.18)

he first s e ifi and detailed a o nt a o t ake area and the Mareotic region was that of Strabo. After his statement that ake areotis was filled man anals from

1. P. Lond. II.231, p. 285 - Hunt and Edgar, Select papyri II. 428 = P. Abinn. I.9, ll. 4-5: σ ε σα ί ρα α α α ει ε ε δια αρε ε ε δια υ εί α ερχ ε α 2 f erodot s 169 where he identified the site of that attle to e at the town of Momemphis “ ε ι ι ” which is roughly identified as om ill all 1942 18 o otte har et 1997 118, n 249 identified it as a o rd h i m el- isn in their ommentary on Strabo 17.1.22

That they drank of that river’ s water is well attested in tra o s a o nt of le ander s first isit to the la e of Alexandria, the advantages of the site, and his determination to found a city bearing his name on the site. Among the advantages of the (new) city’ s site was its location between two (seas), on the north the Egyptian sea, as it is called, 3

LAKE MAREOTIS: RECONSTRUCTING THE PAST the Nile, both from above and on the sides, he goes on to complete the picture of the whole region as follows:

sites of the “ Libyan coastal nome” from immediately before Catabathmus (es-Salloum) to Glaucus Promontory (El-Imâ yid) (V. 5. 4 - 7), he goes on to report the sites of the “ coastal Mareotic Nome” or “ ο ο αρε ου αρ ιο ” , i.e., namely, the village of “ Cheimw” (El-Bordâ n) and Plinthine (ruins 6 km NE of Sidi Kireir) (Ptolemaios IV.5.8). Finally, he comes further east to Chersonesos the Smaller (El-Dekheila) and then to Alexandria, the capital of the whole of Egypt, to quote Ptolemy (Ptolemaios IV.5.9).

“ ...and through these canals the imports are much larger than those from the sea, so that the harbour on the lake was in fact richer than that on the sea, and here the exports from le andria also are larger than the im orts and an one might judge, if he were at either Alexandria or Dicaearchia (Puteoli in Italy) and saw the merchant vessels both at their arrival and at their departure, how much heavier or lighter they sailed thither or therefrom. And in addition to the great value of the things brought down from both directions, both into the harbour on the sea and into that on the lake, the salubrity of the air is also worthy of remark. And this likewise results from the fact that the land is washed by water on both sides and because of the timeliness of the Nile’ s risings … at Alexandria, at the beginning of the s mmer, the ile, eing f ll, fills the lake also and leaves no marshy matter to corrupt the rising vapours. At that time, also, the Etesian winds blow from the north and from a vast sea, so that the Alexandrians pass their time most pleasantly in summer.” (Strabo 17.1.7.)

In another passage of Ptolemy (13-32) he gives a list of the illages inland of the ian ome from west to east the last village to the east being the one called “ Mareotis” . In another passage he provides another list of the “ cities and villages of the Mareotic Nome” (the part inland), further to the east on the onfines of the re io s ian ome the farthest village to the east being that called the “ village of ancient Marea” or α αι ρεια (Ptolemaios IV.5.34). According to Ptolemy’ s calculations the village of Mareotis, the farthest point to the east of the Libyan Nome, was situated on longitude 58 (27º .25’ ) and latitude 28º .20’ , however, the village of ancient Marea is located at longitude 60 (29º .25’ ) and latitude 30º .10’ . Thus, the village of “ Mareotis” was situated to the south-west of the village of “ Ancient Marea” , at a location yet to be identified with a s e ifi an ient site all 1942 114 o ld not identify the present location of either site.

From the above paragraph of Strabo the advantages of Lake Mareotis with its vivid harbour, as well as that of the Mareotic region in general, is quite obvious: intensive commercial activities of imports and exports from the interior of Egypt as well as editerranean argoes an e ellent and health climate which was also a tourist attraction in itself. Another passage Strabo (17.1.14) sheds more light on the size of the lake and the other activities of the inhabitants of its surrounding shores. He describes the situation as follows:

Some of these aspects referred to in the works of the classi al a thors on erning the areoti region are refle ted in papyrological documents, scarce though they are. In the light of such documents the region of Mareotis constituted an administrative “ nome” during the 1st (OGIS 669 = S.B. 8444 68 , l 48 and 53, ol 3, l 1 88 , 2nd l 78, l 8 157 slo 84 138161 ll 12-14 a n 22, l 3 and 3rd (P. Flor. III.338, l6 13, 289, l 2 ent ries

“ Lake Mareia, which extends even as far as this [ the Chersones s/ ara it a ording to o otte har et 1997 or el-’ Dekheila’ according to Ball 1942] , has a breadth of more than one h ndred and fift stadia and a length of less than three h ndred t ontains eight islands and all the shores ro nd it are well inha ited and the intages in this region are so good that the Mareotic wine is racked off with a view to ageing it.”

About a century after Strabo’ s visit to Egypt, Pliny the Elder in his H istoria N atural is reported about Lake Mareotis as follows:

The second important aspect highlighted in these few documents is the close relationship between the Mareotic Nome and Alexandria. Some individuals in these documents would come to Alexandria, through the Mareotic ome, for siness or offi ial or di ial r oses his is lear in some ri ate or siness letters a n 22 P. Flor. III.338). Sometimes they hint at some degree of o-ordination etween the high offi ials of areotis and Alexandria when necessity requires. In a poorly preserved document, the essence of which could be understood in its general lines owing to the existence of several key words, two persons domiciled in the Mareotis region are mentioned in a petition to the strategos of Alexandria regarding the theft of two camels belonging to them. According to the lisher s reading and filling of the la nae, whi h seems almost plausible, they ask the Alexandrian strategos to write to his Mareotic counterpart to summon some suspected person to present himself before him for investigation (P. Oslo III.84). Why was the Alexandrian strategos asked to interfere instead of resorting directly to the Mareotic strategos? Does this further illustrate a degree of collaboration between the two strategoi?

“ Lake Mareotis, which lies on the south side of the city, arries traffi from the interior means of a anal from the ano i mo th of the ile it also in l des a onsidera le n m er of islands eing 30 miles a ross and 250 miles in circumference, according to Claudius Caesar.” (Pliny V.XI.63)

Strabo and Pliny also note the variations in measurement of the lake particularly during the inundation season in the s mmer when the ile flooded rom their a o nts it is clear that the lake was thus much more extensive in anti it than toda o otte har et 1997 104, n 179 Strabo 17.1.14). Last but not least among the classical works concerning the description of the Mareotic region, is the Geography of Claudius Ptolemaios in the 2nd century AD. In his fourth book (Chapter 5 about Egypt) Ptolemy begins his description of the geography of Egypt from west to east stating the longitude and latitude of each given site. After covering the 4

M. ABD-EL-GHANI: ANCIENT SOURCES Bibliography

The theft of two camels in this document also leads us to comment on the repeated reference to camels in the Mareotic documents. This is not unexpected from an area situated on the edge of the Western Desert where the amel was the hief means of traffi and trans ort n one of the documents relating to business affairs, a sum of 900 drachmae is reported as a rent for some camels (P. Haun. II.22, ll. 9-10 and note on l.9.). In another document dated AD 289, two Mareotes sell an Arabian camel elonging to them to one of the hief offi ers rim ilaris of the Prefect of Egypt for 16½ talents of silver (BGU I.13, II. 1-6).

P apyri BGU = A egyptische U rkunden aus K öni gl ichen ( l ater Staatl ichen) Museen z u B erl in, G riechische U rkunden (13). H. Satzinger (ed.), 1904. Berlin. P. Abinn. = T he A binnaeus A rchiv e: P apers of a R oman i h i s a i s (1-9). H.I. Bell, artin, rner an er hem eds , 1962 Oxford. P. Fay. = F ayum T ow ns and their P apyri (134). B.P. Grenfell, nt, ogarth ilne eds , 1900 London. P. Flor. = Papiri greco-egizii, P apiri F iorentini D ocumenti e testi l etterari del l ‘ età romana e biz antina (III, 338). G. Vitelli (ed.), 1915. Milan. P. Fouad = s a y s a (11). A. Bataille (ed.), 1939. Cairo. P. Hamb. = G riechische P apyrusurkunden der H amburger staats- und U niv ersität sbibl iothek (I, 39). P.M. Meyer (ed.), 1924. Leipzig. P. Haun. = P apyri G raecae H aunienses (II, 53). L etters and y a s a y low- a o sen transl ed , 1981 onn P. Iand. = a y i a a a , 53 al fleis h, Schaefer, G. Rosenberger, D. Curschmann, L. Eisner, re , iess, ohr mmel eds , 1912 Leipzig. P. Lond. = G reek P apyri in the B ritish Museum (II, 231). F. en on ell eds , 1973 ilan P. Oslo = P apyri O sl oenses , 84 itrem Amundsen (eds.), 1925. Oslo. P. Ryl. = aa h ka ai a y ii h h y a s i a y a h s (II, 98, 227). C.H. o erts rner eds , 1952 an hester S.B. = Sammel buch griechischer U rkunden aus A egypten , 8444 ila el iessling eds , 1952 iesbaden.

Some of the Mareotic products are also mentioned in the few available documents. In spite of the excellence and distinction of the Mareotic vineyards and wine, there appears to be limited documentary sources that comment on their merits. A single document, a private letter dating to the 2nd century AD, in which the sender tells his friend and addressee that he dispatched him a variety of gifts, among whi h we find and a n m er of n hes of grapes, - 6 small clusters: 2 of them white, 2 of the ‘ royal’ variety, the Mareotic as it is called, and likewise 2 smoke coloured, - in a small basket through Harklides” (P. Fouad, 77, ll. 15-20, l. 17). In other documents dating to the 3rd century AD, there are occasional references to “ strips of linen” , αι ίδιο : some of them white, some “ Mareotic” (P. Ryl. II. 227, ll. 25-33). If we notice that the use of αι ίδιο is a diminutive of αι εία which means “ a head and worn es e iall as a sign of i tor and if we bear in mind that a strip of land near Lake Mareotis was named αι εία, mentioned in Athenaeus (I.33e) and C. Ptolemaios (4.5.14), could one infer some connection between this place and the production of “ linen headbands” whence “ it might have derived its name? There is another mention of “ a glassware of a fine Mareotic quality” from the 4th century AD (P. Fay. 134, ll. 4, 6). Another 4th century document makes a clear reference to the illegal smuggling of natron by Mareotes or other Egyptians across the desert. In this corresponden e etween two high offi ials in harge of the natron monopoly (AD 342-51), strict orders are issued to detain the camel drivers and their beasts, whether in Arsinoe or elsewhere, in order to safeguard the natron of the Fiscus (P. Abinn. I.9 = P. Lond. II.231).

O ther Sources Athenaeus, D eipnosophists, C. Burton (transl.), 1927. London. all, , 1942, y i h assi a a h s. Cairo Diodorus Siculus, D iodorus Sicul us. C.H. Oldfather (transl.), 1935. Cambridge, MA. Herodotus, T he H istories. A. de Sélincourt (transl.), 1971. Middlesex. Pliny the Elder, N atural H istories osto k Riley (eds.), Perseus Digital Library: http://old.perseus.tufts.edu/cgi-bin/ptext? doc= Perseus:text:1999.02 0138 to a essed an ar , 2010 Ptolemaios (Claudius Ptolemy), T he G eography. E.L. Ste enson transl ed , 1932 ew ork Strabo, G eography. ones transl , 2001 ondon Thucydides, T he P el oponnesian W ar ent transl , 1910 ew ork, ondon o otte, , har et, , 1997, Strabon, L e v oyage en Egypt. Paris.

Finally, there are some references to army troops stationed in the Mareotic region, together with the payments and extra payments to them, as well as the divisions of the army to which they were assigned (P. Iand. IV. 53 (AD 88), col. 3 am 39 179 63 Although the image of the Mareotic region, especially in the papyri, is somewhat fragmentary, mosaic-like, and far from complete, it does shed, however, some light on the region. It is clear that the material evidence of future archaeological excavations will add a great deal to our knowledge of this region.

5

LAKE MAREOTIS: RECONSTRUCTING THE PAST

6

M. EL ABBADI: LAKE MAREOTIS IN BY Z ANTINE TIMES

A Note on Lake Mareotis in Byzantine Times Mostafa El Abbadi

This paper will present readings of selected passages in the ancient sources that have a bearing on Lake Mareotis and its region. As will be seen, the name Mareotis was often exchanged for Mareia, the name of its capital city and harbour, which in Pharaonic times served also as a garrison city protecting the north-western approaches to Egypt, as riefl stated in erodot s 18 30 and h dides 104 1

m eror stinian ilt man onstr tions and in articular the residences of the magistrates and baths.”

Of special interest is his use of the Greek form Taphosiris, and not the Latin corruption Taposiris, which was used by tra o lso the onstr tion work of stinian re eals the ontin ed im ortan e of the town more than fi e ent ries after the time of Strabo, despite the total change of religion from paganism to Christianity.

he first detailed des ri tion of the ake areotis region is fo nd in tra o 17 1 71 in his well known des ri tion of Alexandria where reference is also made to the southern region of ake areotis n a f rther signifi ant assage, tra o 17 1 15 omments on an im ortant as e t of the economic value of the lake:

s regards the so thern onfines of le andria, ro o i s re orts 1 1-5 he ile i er does not flow all the wa to le andria, t after flowing to the town whi h is named haereo , it then t rns to the left, lea ing aside the onfines of Alexandria. Consequently the men of former times, in order that the city might not be entirely cut off from the river, dug a very deep canal from Chaereou and thus by means of a short branch made the river accessible to it. There also, as it chances, are the mouth of certain streams flowing in from ake aria

he l s i e a r s lant grows in the g tian marshes and lakes, as also the Egyptian cyamus believed to e the lot s lant ermer 1985 39 ff from whi h was made the i ori m a kind of essel for drinking it has stalks a ro imatel e al in height, a o t 10 feet But though the byblus is a bare stalk with a tuft on top, the am s rod es lea es and flowers in man arts, and also a fruit like our cyamus, differing only in size and taste ordingl the am s-fields afford a leasing sight, and also enjoyment to those who wish to hold feasts therein. They hold feasts in cabin-boats, in which they enter the thick of the cyami and the shade of the lea es for the lea es are so er large that the are sed both for drinking-cups and for bowls, for these even ha e a kind of on a it s ited to this r ose and in fact Alexandria is full of these in the work-shops, where the are sed as essels and the farms ha e also this as one source of their revenues, I mean the revenue from the leaves.”

In this canal, it is by no means possible for large vessels to sail, so at Chaereou they transfer the Egyptian grain to oats l emboi whi h the are wont to all diaremata, and thus convey it to the city, which they are enabled to reach by way of the canal-route and they deposit it in the quarter of the city, which the Alexandrians call P hial è . But since it often came about that the grain was destroyed in that place by the people rising in sedition, the m eror stinian s rro nded this distri t with a wall … .”

In this passage, the following points are of special interest: a he freshwater anal from Chaereou received additional s lies from ertain streams flowing in from ake Maria” . It is worthy of note that Procopius states that the freshwater canal to Alexandria branches off at a location he called Chaereou, whereas Strabo uses the name of Schedia r risingl , other so r es of the 6th century and later, use the name of Chaereou onl ignifi antl we an mention the Mosaic Map of Madaba which is also believed to date from the 6th ent r i- onah 1954 16-18 nother so r e is te han s anti s 677 18 also in the 6th century, who speaks of Chaereou as a city in Egypt. To all appearances, Schedia and Chaereou were two different locations close to one another, Schedia north of the canal and located on the Canopic Branch, whereas Chaereou is sit ated immediatel so th of the anal aas 1997 25, to of ma 2 and seems to ha e gained in im ortan e at the time of stinian when te han s alled it a pol is.

Furthermore, in connection with feasting, Strabo makes mention of Taposiris at the western end of Lake Mareotis and adds 17 1 14 that there the also hold a great lic festival …. and near it there is a rocky place on the sea where likewise crowds of people in the prime of life assemble during every season of the year.” With this classical description in mind, we move on to some Byzantine sources in order to identify any similarities or differences in the descriptions attributed to the lake in later periods. An interesting passage is to be found in ro o i s work on the ildings of stinian in the 6th ent r 1 12 t deals with the regions to the west and so th of le andria e states that to the west n that territor , is a it one da s o rne distant from Alexandria, Taphosiris by name, where they say the god of the Egyptians, Osiris, was buried. In this city the

7

LAKE MAREOTIS: R

PAST

n that anal the large ile essels o ld not sail so their argo of grain had to e transferred to diaremata light trans ort oats his word was known onl in g t hen e ro o i s e lained it the reek word l embos” . he same word with the hange of a owel dieremata” has een fo nd in a ri f ertel 1917 130 hese dieremata were probably also used to navigate across Lake Mareotis.

The above discussed passages provide evidence that with the founding of Alexandria, Lake Mareotis acquired a li el and ital signifi an e that ontin ed into late anti it ith its ariet of flowering lants, the lake provided a pleasurable recreation resort for the Alexandrians who were, according to Strabo, in the habit of holding feasts in cabin boats. Even as far west as Taposiris, Strabo adds, a great public festival was held, obvio sl in onne tion with the lt of siris ore than fi e ent ries later, ro o i s re orts that stinian showed interest in that town a hosiris adding new onstr tion work which is an indication of its continued importance.

he grain was e ent all arried light trans ort oats and deposited in the quarter of Alexandria called P hial è . This P hial è is not attested anywhere else and we have the impression that it was close to a landing-place near the anal owe er, ro o i s s eaks of it as alread e isting and how it sed to e l ndered in times of tro le hen e stinian s rro nded it with a wall for rote tion

It is of interest to know that the freshwater canal that links to le andria re ei ed additional s lies from ertain streams flowing in from ake aria his assing mention by Procopius provides additional evidence regarding the intricate hydraulic system of the freshwater canal. Furthermore, this anal ro ided the final fa ilit for transferring grain from large Nile vessels – which could not sail in the anal - to light trans ort oats dieremata

inall , we ome to a te t o hroni s earl 7th ent r in his account of the Miracl es of Saints Cyrus and J ohannes. The miracle happened to a certain Christodorus, deaon of the h r h of the two saints at the time of ohn the lmsgi er, then isho of le andria 610-619 , whom o hroni s des ri es as the elo ed of the oor n his a o nt, o hroni s has this to sa mira le no 8

Finally, the lake and the regions surrounding it were of vital economic value to Alexandria, since the waters of the lake teemed with fish and water-fowl and the s rro nding fertile lands flo rished with large lantations ntil the early 7th century. Accordingly, Alexandria depended upon this neighbouring hinterland for the regular supply of its daily needs of meat and vegetables.

ake areia o le andrians elo ed of hrist, o all witho t e e tion know it well i e the mira le , so also do a good number of Egyptians and most of the Libyans and as well as you, who are curious to know about Alexandria. Christodorus once sailed across Mareotis with the intention of inspecting the domains belonging to the church of the saints. A storm suddenly blew violentl the time was winter , agitating the waters of the lake and raising great waves. Christodorus was in great danger and thought he was going to perish. Nevertheless, he prayed to the saints who listened to his prayer … . And the wind at once dropped and he was saved.”

Bibliography i- ona, , 1954, T he Madaba Mosaic Map er salem. alderini, , 1935-1987, D iz ionario. Cairo, Madrid. C.Theod. = Codex T heodosianus, ommsen e er eds , 1904-5, 2 ols erlin onner, , 1992, T he Mosaic Map of Madaba. Kampen. htt //198 62 75 1/www1/ofm/mad/legends/legends156 html essed to er 2009 o rtes rof aggag as o , transl ed , 2006, Sophrone de J é rusal em, Miracl es de Saints Cyr et J ean. Paris. ermer, , 1985, F l ora des P haraoischen A egypten, onders hrift 14 airo with thanks to rof ir at eif aas, , 1997, A l ex andria in L ate A ntiquity. Baltimore, London. erodot s, H istoriae, de transl , 1966 ford eonti s ea oleos, ife of ohn the lmsgi er n awes a nes transl eds , T hree B yz antine Saints: Contemporary B iographies of St. D aniel the Styl ite, St. T heodore of Sykeon, and St. J ohn the A l msgiv er 1948 ford ertel, , 1917, L iturgie. Leipzig. Palladius, T he L ausiac H istory tler transl , 1904 Cambridge. Palladius, H istoria L ausiaca. Migne, Patrologia Graecia 34 1860, re rint 1998 ottingen

For our purpose however, the main point of interest in this assage is the mention of the domains of the h r h n the Byzantine period, just as important temples of earlier times disposed of great wealth and property, so now the Church of Alexandria possessed considerable property. In 415 m eror heodosi s, in a law on atronage, g aranteed the property of the Church of Alexandria which had een a ired efore 397 heod 23 6 he material wealth of the Church of Alexandria is also asserted by the fa t that it owned a mer hant fleet in the editerranean. Early in the 7th ent r the same ohn the lmsgi er, sed that fleet to im ort orn from i il d ring a famine at le andria eonti s ea oleos, 13 owe er, thro gho t anti it , the lake s waters teemed with fish and water-fowl e t 3 867 eonti s ea oleos, 8 fin s, 27 10 le andria de ended on its neighbourhood for the supply of its daily needs of meat and vegeta les e ma also re all that there were some 2,000 monks near Alexandria at Nitria, south of Lake Mareotis and that however frugal their diet might have been, they were am l s stained rod ts of the region alladi s , 1-2 o omen, , 29 8

M. EL ABBADI: LAKE MAREOTIS IN BY Z ANTINE TIMES Procopius, B uil dings ewing transl , 1961 ondon P.Tebt. = T ebtunis P apyri, ol 2, nt m l transl , 1933 ondon fin s, H istoria Monachorum. Migne, Patrologia Latina 21, 1849 ottingen o hrone de r salem, Miracl es de Saints Cyr et J ean, as o transl , 2006 aris

Sozomen, H istoria Eccl esiastica. Migne, Patrologia Graeia 67 1864, re rint 2002 ottingen Stephanus Byzantius, O pera, 4 ols , indorf ed , 1825 ei ig Strabo, G eography, ones transl , 1917 ondon Thucydides, H istoriae, t art ones transl , 1897 Oxford.

9

LAKE MAREOTIS: R

PAST

10

I. AWAD: MARY UT LAKE THROUGH MAPS

A Study of the Evolution of the Maryut Lake through Maps Ismaeel Awad

This study concerns the evolution of the shape of Maryut Lake and its connections with the Nile River and the Mediterranean Sea from antiquity until the modern day by studying and observing maps of the Maryut Lake region, lassified hronologi all from the oldest to the most recent.

The aim of this study is to use maps to measure the evolution of the Maryut Lake, to test the value of using the map as a tool in the study, to evaluate the type of results obtained, and lastly, to try to forecast the shape and state of the Maryut Lake in the future. he ar t ake region onsists geogra hi all of fi e different parts (Fig. 1). In the north, a sandy strip extends along the Mediterranean coastline. To the south there is a calcareous ridge known as Abusir-Mexx that has an average altitude of 30 m high. The center of the Maryut region is di ided into two s -di isions, the first is the ar t Basin which is situated towards the east with an average altitude attaining sea level; the second sub-division is the Maryut Valley situated towards the west of the region. And finall , the ar t idge lies towards the so th-west and has an average altitude of 50 m high.

In the following pages, two different types of available ma s are onsidered he first t e of ma re resents edited illustrations of Maryut Lake by different historians and ancient cartographers who built up their illustrations through visual observation or by drawing. The study of these illustrations using the superimposition method, essentially overlapping one map on another, cannot be applied, as any attempt to do so would result in the production of inaccurate results based on the fact that these illustrations display approximate distances and inexact locations. The second type of map utilised topographic maps, which when edited using accurate topographic methods, can produce more accurate results. The application of topographic methods began during the French campaign at the end of the 18th century in which levelling and goniograph tools are used to construct maps in Egypt. Topographic methods have developed over time through to the current use of satellite images in mapping. The data extracted from these maps are accurate enough to be simply compared and overlapped.

About 20 maps and illustrations are used in this study, of which the oldest map was edited at the end of the 16th century and the most recent was extracted from a Landsat satellite image produced in 2001.There are two different ways of studying these different types of maps, as will be revealed during the course of this paper. F ig. 1: Maryut L ake region. D igitised by I . A w ad. A rchiv es CEA l ex .

11

LAKE MAREOTIS: RECONSTRUCTING THE PAST Illustrations The illustrations utilised in this study date from the end of the 16th century until the end of the 18th century. Some of them were constructed ‘ indoors’ without ever stepping foot in Egypt. They were done by using other maps edited by other publishers, used as base maps for adding new information or this reason it is diffi lt to se these illustrations to derive any form of accurate reference or in application to an overlapping study.

the e ol tion of names and lo ations of settlements such as towns or villages existing all around the Mareotic region. Illustrations edited before the end of the 16th century used to ignore the Maryut Lake as it was of little importance at that time, certainly when compared with the town of Alexandria. Ancient map makers rarely mentioned Maryut Lake, locating it at the margin of the map and illustrating its location in small letters. They portrayed an ignorance of the lake’ s shape, such as for example the illustration edited in 1548 by Pierre Belon du Mans (Fig. 2).

However, some useful information could be extracted from examining these illustrations, such as: the e ol tion of the onne tions etween ar t Lake and the Nile River, Nile branches and the Mediterranean Sea,

he ill strations sed in this st d are lassified hronologically from the oldest to the most recent: F ig. 2: 1548, Vray portraict de la ville d’ Alexandrie en Egypte, P ierre B el on du Mans. J ondet G ., 192, pl . I I . A rchiv es CEA l ex .

F ig. 3: 1570, de tail from Aegyptus antiqua, incl uding a focused map of the region of A l ex andria, A braham O rtel ius. J ondet G ., 192. A rchiv es CEA l ex .

12

I. AWAD: MARY UT LAKE THROUGH MAPS 1.0:

1 570,1 A braham O rtel ius (Fig. 3)2 It can be observed from the general map that the Maria, et Mareotis l acus qui et A ra potes (Maryut Lake’ s) shape took a direction from north-east to south-west incorporating eight islands as shown in the fig re

River), which was also connected to the Nile River at its south-eastern end. The boats depicted in the Moeris Swamp show that there was extensive navigation during this period between Moeris Swamp, Maryut Lake and the Nile River.

ar t ake had two onne tions, the first was the onnection with the Myris quod et Moeris stagnum (Moeris Swamp) to the south through the y s i s (Lycus

The second connection related directly to the Nile River in three places:3 the first one to the north related to the Canopica fossa4 (Canopic Branch), the second connected the lake with the Canopic Branch at the north-east near the town of Schedia, and finall , the third onne ted the lake directly with the Nile River at its eastern side. .02:1 158, L iv io Sanuto (Fig. 4)5 In this map, the B ycheira l acus (Maryut Lake) has deformed in its shape forming a triangle enclosing a large number of islands (21 islands). Just one connection linked the lake to the Canopic Branch at its eastern end while no connection was illustrated with the Moeri dis l acus (Moeris Lake). 1.03:

65,1 N icol as Sanson D ’ A bbev il l e (Fig. 5)6 The map of D’ Abbeville has similar limitations to the previous map. The shape of Maryut Lake is similar although more rectangular and adopting more of a northwest to southeast orientation. However, most of the islands that had been previously observed have disappeared. 1.04:

71, P aul L ucas (Fig. 6)7 No connections are shown with the L ac de l a Mareote (Maryut Lake) in this map while some islands reappeared (nine in total). The Moeris Lake has totally disappeared, and at the same time the L ac du Madie (Madieh Lake/Abuir ake is ill strated for the first time

F ig. 4: 158, detail from Africae Tabula VII, L iv io Sanuto. A rchiv es B ibl ioteca N az ional e Central e D i R oma. 1.05:

72,1 C l aude Sicard (Fig. 7)8 In Sicard’ s illustration the Mareotis P al us (Maryut Swamp’ s) form has changed again to resemble the shape of a bird or a duck. Some connections are apparent between Maryut Lake and the Nile to the east that joined together efore flowing into ar t wam small lake named N itraria P al us9 itraria wam was identified to

F ig. 5: 165, detail from Royaume et désert de Barca et l’ Egypte, N icol as Sanson D ’ A bbev il l e. C ommercial reproduction.

1. There is another illustration of Ortelius dated from 1595, found in a late commercial reproduction from the beginning of the 21th century. 2. A egyptus antiqua, map of the Antique Egypt. 3. There is a fourth connection shown in the focused map of Alexandria and not in the general ma t is sit ated etween the first and the se ond connections at the north passing near the village of Chabrie pagus. 4. Mentioned in the focused map of Alexandria. 5. A fricae T abul a V I I , map of Africa, sheet no. 7. 6. R oyaume et dé sert de B arca et l ’ Egypte, Kingdom of the desert of Barca and Egypt. 7. Carte de l a B asse Egypte et du cours du N il , map of Lower Egypt and the Nile River. 8. Carte de l ’ Egypte A ncienne, map of Ancient Egypt. The original map edited by Claude Sicard was not found. This is a synthesis of two copies extracted from the original map done by D’ Anville and Danlisle and the place name’ s list of Sicard. Sauneron & Martin 1982. 9. There are two lakes named N itraria in the map: one near the Maryut Lake which is mentioned in the study and the other located in Wadi Natrun between Alexandria and Cairo.

13

LAKE MAREOTIS: RECONSTRUCTING THE PAST

F ig. 6: 17, detail from Carte de la Basse Egypte et du cours du Nil, P aul L ucas. A rchiv es I F A O .

F ig. 7: 172, de tail from Carte de l’ Egypte Ancienne, Cl aude Sicard. Saune ron & Martin 1982, A rchiv es I F A O .

F ig. 8: 1743, detail from Carte d’ Egypte, R ichard P ococke. A rchiv es I F A O .

F ig. 9: 1753, de tail from Carte de l’ Egypte ancienne et moderne, R obert de V augondy. A rchiv es J .- Y . E mpereur.

the south-east of Maryut Lake, connected with it through A cheron Canal is (Acheron Canal). The town of Marea is illustrated, but strangely to the north-west of Maryut Swamp and not on the southern shores of the lake as located nowadays.

that used to connect Moeris Lake with Maryut Lake. The towns of P l inthine.S. and T aposiris.P . 11 were indicated but very far from the lake towards the west. 1.07:

.06:1 1743, R ichard P ococke (Fig. 8)10 The shape of B irk Mariou.s ./ L acus Mareia.f ./ Mariotis.S. (Maryut Lake) changed again in this map taking a triangular form. A lake is mentioned at the place of the Nitraria Lake but located further south than before, and the Moeris Lake appeared as B irk El K aroon/L acus Moeris feu Moerius (Karoun Lake).

ca 1753, R obert de V augondy (Fig. 9)12 th This 18 century map indicates yet a further change to the shape of B irk Mariou (Maryut Lake) now depicted as being aligned from north-east to south-west. There is still

10. Carte d’ Egypte, map of Egypt. 11. Known as ancient navigation ports situated actually at the western arm of the Maryut Lake. 12. Carte de l ’ Egypte ancienne et moderne, map of Ancient and Modern Egypt. The base map is inspired from Claude Sicard’ s map.

B ahr B el l omah i.e Mare seu L acus Sine A qua (lake or sea without water) was mentioned at the place of the old canal 14

I. AWAD: MARY UT LAKE THROUGH MAPS one easterly connection between Maryut Lake and the Nile River. A lake is mentioned and located at the same place of the Nitraria Lake as in the previous map. 1.08:

762,1 R igobert B onne13 The Bonne map has many similarities with the previous map, with just one change which is the appearance of the town of A busir located at the middle of the northern shoreline of the Maryut Lake. 1.09:

764,1 J acques N icol as B el l in (Fig. 10)14 A connection between L ac Mareotis (Maryut Lake) and the editerranean ea is shown for the first time to the west of Alexandria, with the disappearance of Lake Nitraria. Two connections appeared between Maryut Lake and the Nile (both meet at the same point, southeast of the lake). One of them approached from the east and the other from the B ahr- el - H el ame ou F l euv e seché (canal or sea without water) from the south-east, and passed through a L ac de Sel (Salty Lake)15. 1.0:

179, E dme Mentel l e & P ierre- G regoire Chanl aire16 In this map of 1799 Sebaga ou B irk Marioul 17 (Maryut Lake) reappeared in the form of a triangle with what appears to be a sea connection to the west of Alexandria. The connection with the Nile River appeared through the Canal de B ahiré ahire anal flowing into the ar t ake, and L ac de N atron (Natron Lake) appeared at the place of the ex. Nitraria Lake southeast of Maryut Lake. The B ahhr- bé - l â- mâ ou F l euv e sans eau (canal without water) was also shown approaching from the B irket- il K erun (Karoun Lake) which is located to the south in the F eï um region, extending towards the southwest of Maryut Lake but without any intervening connections.

F ig. 10: 1764, de tail from Carte des Embouchures du Nil, et partie de son cours, J acques N icol as B el l in. A rchiv es J .- Y . E mpereur. geo-referen ing the ma s in the same oordinate system, digitising ea h ma in to rogram, o erla ing ea h digitised ma one on the other, anal sing the digitised feat res in r rogram

Towns are named around the Maryut Lake such as the town of Mariout (Marea) located on the northern shoreline of the lake and T our des A rabes (town of Abusir) near to the Mediterranean Sea. 2.01:

180, F rench campaign, s cal e: 1: 10. (Figs. 11 & 12)18 his is the first a rate to ogra hi ma on erning the Mareotic region. It was edited by the engineers of the French campaign who came to Egypt between 1798 and 1801. The map was constructed by accurate topographic instruments including a level and a goniograph.

T opographic M aps In order to extract precise locations and establish an accurate surface area, four steps were undertaken when studying the second type of map:

In this map, a cut is shown in the dike separating the B ohé ï ré h Marî oût ou l ac Mareotis (Maryut Lake) from the B ohé ï ré h Maadî é h, l ac Madié ou d’ A boûqî r (Madieh or Abuqir Lake). The dike was breached by the English army in April 180119 during the war between the French

13. Carte de l ’ Egypte ancienne et moderne, map of Ancient and Modern Egypt. 14. Carte des Embouchures du N il , e t partie de son cours, map of the Nile river mouths, and a part of its stream. 15. The Salty Lake is at the place of the Nitraria Lake situated in Wadi Natrun. 16. Focused map Carte particul iè re et dé tail l é e du D el ta (Detailed map of the Delta) in the Carte physique et pol itique de l ’ Egypte (Physical and political map of Egypt). There is another similar map Carte d’ Egypte (map of Egypt) edited by Jean-Baptiste Bourguignon D’ Anville in 1765. Charles-Roux 1910. 17. Sebaga is near to Sabkha in ra i lang age whi h means marsh, and B irk comes from B irka in ra i , too, whi h means pond.

18. Carte d’ A l ex andrie, sheet no. 37 extracted from the Atlas of the French Expedition Carte topographique de l ’ Egypte et de pl usieurs parties des pays l imitrophes. 19. The dike was breached 19 April, 1801, by the English army under the orders of General Sir John Hely-Hutchinson, in order to cut off the freshwater supply brought by the Alexandria canal from the French garrison located in Alexandria, and to cut the road relating them to the main French army situated in Cairo.

15

LAKE MAREOTIS: RECONSTRUCTING THE PAST

F ig. 1 1: F ig. 12:

180 180,

, Carte d’ Alexandrie, F rench campaign. A rchiv es J .- Y . E mpereur. Carte d’ Alexandrie, F rench campaign, or iginal scal e 1: 10,. D

and English. As a result, the water in Abuqir Lake (which was situated at a higher altitude than the Maryut Lake and was directly connected with the Mediterranean Sea), was redire ted into ar t ake, a sing flooding in its eastern (125 km² ) and western (90 km² ) extremities.

igitised by I . A w ad. A rchiv es CEA l ex .

The depth of Maryut Lake was measured in units of feet using depth soundings all around the lake. We can observe from the depths shown on the map that the deepest zone in the Lake was the north-eastern zone that was an average depth of about 4.5 m. Maryut Lake becomes shallower

wo lake limits are re resented on the ma , the first referred to the limit of the flood on 9 g st, 1801,20 and the se ond showed the limit of the lake after the flood had returned to its preceding limit on 10 January, 1802.21

20 he flood limits were o tained from awiland s ma s ale 1:128.000), engineer in the English army. Jacotin 1809-1828: 83. 21. Due to the effect of the northerly wind and to the reduction of water in the Nile River branches. 22. Maybe it was the connection mentioned before in the ancient illustrations between the Mediterranean Sea and the Maryut Lake at the west of Alexandria.

Another dried up old cut22 is illustrated linking Maryut Lake with the Mediterranean Sea to the east of the A nse de Marabou (Gulf of Marabou), west of Alexandria. 16

I. AWAD: MARY UT LAKE THROUGH MAPS

F ig. 13:

186

, Carte des Environs d’ Alexandrie, M ahmoud B ey El - F al aki. A rchiv es CEA l ex .

F ig. 1 4 : 1 8 6 6 , Carte des Environs d’ Alexandrie, Mahmoud B ey El - F al aki, original scal e 1 : 2 0 0 , 0 0 0 . D igitised by I . A w ad. A rchiv es CEA l ex .

time efore the flood whi h was a o t 890 km 211 300 feddans). At this point, the lake’ s maximum width (from north to south) was about 45 km, and its maximum length (from east to west) reached about 56 km.

from the north-eastern zone towards the south (2 m in average) and the west (1.5 m in average) until the lake attends 0 m at its limits. n addition, two le elling rofiles23 were undertaken across Maryut Lake, one at the Maryut Valley24 and the other along the dried old cut mentioned previously.

23. Le Pè re 1809-1828: Pl. 19. 24. Between the ruins of the town of Marea and the Mediterranean Sea passing through the saint tomb of A boû e l K eï r.

hanks to digitising, we are a le to as ertain with s fficient accuracy the precise surface area of the lake at that 17

LAKE MAREOTIS: RECONSTRUCTING THE PAST

i Atlas of Egypt, Vol. I, Lower Egypt, Survey of Egypt (G. Daressy), original scal e 1: 50,. D igitised by I . A w ad. A rchiv es CEA l ex . F ig. 16:

194

, map series 1: 10.

Survey of Egypt, original scal e 1: 10,.

2.0:

186, M ahmoud B ey El - F al aki, s cal e: 1: 20. (Figs. 13 & 14)25 This map is interesting because El-Falaki (1872) mentioned that L ac Maré otis (Maryut Lake’ s) shape at that time (1866) was similar to the lake’ s shape when the town of Alexandria was founded in antiquity, according to Strabo’ s notes.26 he a thor also defined the an ient ran hes of the Nile as the Canopic Branch and Schedia Canal.

D igitised by I . A w ad. A rchiv es CEA l ex .

Lake was also reduced in area to 450 km² (which is about 50% of its area as indicated in 1801). This reduction in area occurred in just 65 years, and was especially apparent in the eastern part where its maximum length was about 24 km.

25. Carte des Env irons d’ A l ex andrie, map of the surroundings of Alexandria 26. El-Falaki 1872: 99.

From this map, it can be observed that the L ac d’ A boukir (Abuqir Lake) has reduced in area. Proportionally Maryut 18

I. AWAD: MARY UT LAKE THROUGH MAPS

F ig. 17: 1970 s, m ap series 1: 25.0, E A rchiv es CEA l ex .

gyptian G eneral Surv ey A uthority, or iginal scal e 1: 25,0. D

2.03:

igitised by I . A w ad.

19 1, Sur v ey of Egypt, s cal e: 1: 50. (Fig. 15) 27 The area of Maryut Lake continued to reduce particularly on its eastern margins. The Egyptian government began to have some interest in this zone by managing a system of irrigation canals and drains (for example: El-Omum Drain, whose waters flowed from the so th-eastern side of the Maryut Lake) in order to cultivate the region around Maryut Lake. The area of the lake was reduced to 340 km² (38% of 1801s area). However, the dimensions of the lake from east to west appear to have extended compared to previously (about 75 km). The ancient cut between the Mediterranean Sea and Maryut Lake to the west of Alexandria is again mentioned.

north-eastern shores, particularly with the construction of Nozha30 and the seaplane base. By now most of the western part of the lake had dried up, perhaps partly due to the fact that this map was surveyed in the dry season.31

2.04:

A great deal of development had also occurred in the lake:

2.05:

In the middle of the lake, the salt marshes became much larger (it has expanded ten fold in just 20 years). Big changes also took place in the north-eastern part of the lake he sea lane ase was transformed into a ig fish farm, and Nozha airport became a part of the land in order to se arate the new fish farm from ar t ake

2.06:

1970s , E gyptian G eneral Surv ey A uthority, scal e: 1: 25.0 (Fig. 17) 32 By the 1970s a big development had occurred at the eastern and southern parts of Maryut Lake with an expansion of irrigation canals and drains. The N ubareya Canal was completed in 1976 channeling waste water runoff from agriculture lands into the lake, and then into the Mediterranean Sea via the Mexx pump station.

28 1930, S urv ey of Egypt, s cal e: 1: 10. The area of Maryut Lake continued to reduce in size particularly towards the south-east as the lake became increasingl e loited for agri lt re r oses n artifi ial salt marsh a eared for the first time in the middle of the lake in order to produce salt.

194, S urv ey of Egypt, s cal e: 1: 0.1 (Fig. 16) 29 The lake was further reduced on its eastern shores. This map also indicates early interest in expansion along its

Maryut Lake’ s surface area has reduced further to just 210 km² (51.000 feddans) (which is about 1/4 of its area in 1801).

27. The maps of this series have been collected from: Survey of Egypt 1914. Eight maps have been used in digitising: sheet no. 009 R aml eh, sheet no. 010 A buqir, sheet no. 020 Mal l aha, sheet no. 021 A l ex andrie, sheet no. 022 K afr El D aw ar, sheet no. 046 H ammam, sheet no. 047 B ahig and sheet no. 048 K ing Mariout. 28. German edition, four maps have been used in digitising: 88/42 H ammam, 88/48 G hayata, 92/48 A l ex andria and 92/54 D amanhur. 29. Arabic edition, four maps have been used in digitising: 0° 31’ /29° 40’ A bu El - Matamir, 31° /29° K rair, 31° /29° 40’ A l ex andria and 31° /30° 20’ D amanhur.

. 30. Known at that time as F ouad the F irst air ort 31 he western one s water le el is dire tl infl en ed dr and wet seasons. 32. Fourteen maps have been used in digitising: 90/450 H ammam, 90/465 Mahatet El - G harbaneyat, 91/450 K om Makhoura, 91/465 B urg El - A rab, 92/465 B urg El - A rab, 91/480 Mahatet B ahig, 91/495 A eroport of A mreya, 92/480 K reir, 92/495 A mreya, 93/480 El - D eir, 93/495 Mex x , 93/510 K om Echo, 94/510 R aml eh and 94/525 Munshaat B ul ine.

19

LAKE MAREOTIS: RECONSTRUCTING THE PAST

F ig.18 :

19

, m ap series 1: 50., E

gyptian G eneral Surv ey A uthority. D igitised by I . A w ad. A rchiv es CEA l ex .

F ig. 19:

201

, L andsat satel l ite image, r esol ution 15m

/px l . D igitised by I . A w ad. A rchiv es CEA l ex .

2.07:

19, E gyptian G eneral Surv ey A uthority, scal e: 1: 50. 0 (Fig. 18) 33 A huge development occurred in the southern and eastern parts of the Mareotic region in the early 1990s which enabled the entire lake to be part of the same irrigation system, and large areas of the middle zone of the lake were transformed into fish farms

201, L A N D SA T satel l ite image, r esol ution: 1 5m /pix el (Fig. 19) 34 Satellite images were used to study more recent changes in the lake morphology due to the unavailability of more

These were the last map series produced concerning the Mareotic region, and edited by the Egyptian General Survey Authority.

33. Six maps have been used in digitising: NH35-L5d H ammam, NH35L6c B urj al - ’ A rab, NH35-L6d K ing Mariout, NH35-P3a Sidi K irayr, NH35-P3b A l ex andria and NH36-M1a K afr El - D aw ar. 34. Downloaded from the University of Maryland site (www.umd.edu).

2.08:

20

I. AWAD: MARY UT LAKE THROUGH MAPS T abl e 1: Comparison tabl e betw een different periods of Maryut L ake.

F ig. 20: Ev ol ution diagram of Maryut L ake’ s surface areas

contemporary maps since the last map series was produced in the 1990s.

as the onstr tion of fish farms and artifi ial salt marshes, all of which require stable quantities of water.

A topographic map was extracted from a LANDSAT satellite image of a medium resolution of 15 m/pixel, which is eno gh to digitise the limits of the lake, marshes, fish farms, wide anals and drains, t not eno gh to define the other small features such as buildings, narrow canals, secondary drains or roads.

n the other hand, nat ral infl en es to the sha e of ar ut Lake have a limited impact nowadays (such as rain and ground water). However, any loss in the lake’ s surface area in the future depends on the rate and types of uses of the lake and the implemented management plan. F orecasting the S hape of M aryut L ak e Digitising the topographic maps does provide an opportunity to forecast the future shape of the Maryut Lake region, if left to evolve naturally without any external interference. A forecasting study is undertaken by extracting and digitising different contour levels of the Maryut Lake region. The maps utilised in this digitising exercise were the map series of scale 1:25,000 edited in 1940s, and the same map series edited in 1970s; these maps series were chosen because they are the most accurately contoured maps that covered the whole Mareotic region. The accuracy of the results of such an exercise is after all dependent upon the precision of the maps used, and on the data extracted.

The map showed that the area of Maryut Lake’ s is still about 210 km² (51,000 feddans). Its maximum dimensions attended 66 km in length and 11 km in width. From the table (Table 1) and the diagram (Fig. 20) showing the evolution in surface areas of Maryut Lake from 1801 until 2001, we can notice that the lake has lost 75% of its surface area in just 200 years. The greatest loss occurred between the period from 1801 to 1866, in which the lake lost half of its area d e to dr ing and land fill Maryut Lake became more stable with regard to its rate of surface area lose from the 1970s until now. During this 30 year period it lost 0.1% of its surface area. The main contributor to the reduced rate of water loss was irrigation and the increase in activities around the shores of the lake, such

he 0 m onto r line was the first digitised, whi h refers to the limits of the Maryut Basin, which is the same as sea level (0 m). By comparing this contour map with Maryut 21

LAKE MAREOTIS: RECONSTRUCTING THE PAST

i a is ay ak h a ai in the Mareotic region. D igitised by I . A w ad. A rchiv es CEA l ex .

a s i s

a

Results (Fig. 22) The surface area of Maryut Lake reduced by less than a quarter (24% ) over a period of 200 years (1801-2001). This does not bode well for the future of Maryut Lake, particularly when you consider the intensive construction projects in and around the lake especially at its eastern end, its northern and at its north-western parts having been contin o sl di ided , dried o t and filled to onstr t to rist villages and settlements.

ake s limits in 1801 witho t flood , it an e seen ig 21) that they are approximately identical. That means that the lake’ s limits in 1801 had reached the 0 m level. Using this kind of study, the Maryut Lake’ s limits can be compared to any topographic map with the different digitised contour levels such as the levels of 0 m, -1 m, -2 m and -3 m, in order to study the lake water’ s level in any period. We chose these precise levels because the deepest altitude of the Mareotic region is about -4 m which is limited to a small area in the north-eastern part of the lake and the highest level that the lake’ s water reaches did not exceed 1 m.

he se of ar t ake itself s h as fish farms or salt marshes has increased from 1 km² (0.3% of Maryut Lake’ s area) in 1930 into 39.6 km² (18.6% of Maryut’ s Lake area) 22

I. AWAD: MARY UT LAKE THROUGH MAPS

F ig. 2:

Comparison betw een Maryut L ake 180-

201. D

igitised by I . A w ad. A rchiv es CEA l ex . Charles-Roux, F., 1910, L es origines de l ’ Ex pé dition d’ Egypte. Paris. Carte d’ A l ex andrie, 1801, French campaign. Carte topographique de l ’ Egypte et de pl usieurs parties des pays l imitrophes [ MAP] . El-Falaki, Mahmoud Bey, 1872, Memoire sur l ’ antique A l ex andrie, s es faubourgs, s es env irons. Copenhagen. Jacotin, M., 1809-1828, Mémoire sur la construction de la carte de l’ Egypte. D escription de l ’ Egypte, Etat Moderne, Tome 2, 2è me Partie. Paris. Jondet, G., 1921, A tl as historique de l a v il l e et des ports d’ A l ex andrie. Cairo. Le Pè re, G., 1809-1828, Planche 19, Vallée du Nil et Lac Maréotis. D escription de l ’ Egypte, A ntiquité s, A. Vol. 5. Paris. Livio Sanuto, 1588, A fricae T abul a V I I . Biblioteca Nazionale Centrale di Roma [ MAP] . Lucas, P., 1664-1737 (1724), V oyage du Sieur P aul L ucas, fait en MD CCX I V & par ordre de L ouis X I V , dans l a T urquie, l ’ A sie, Sourie, P al estine, H aute et B asse Egypte. Paris. Pococke, R., 1745, Observations on the Islands of the Archipelago, Asia Minor, Thrace, Greece and some other parts of Europe. A D escription of the East and some other countries. Vol. II. Part II. London. Sauneron, S., & Martin, M., 1982, O euv res I I I - P aral l è l e gé ographique de l ’ ancienne Egypte et de l ’ Egypte Moderne. Cairo. Surv ey of Egypt, 1914, A tl as of Egypt, V ol . I , L ow er Egypt, Cairo. de Vaugondy, R., 1753, Carte de l ’ Egypte ancienne et moderne [ MAP] .

in 2001. That means that the lake’ s use has increased about 40 times in just 70 years for salt production. Conclusion The map can be a very useful tool in a study of temporal evolution of any geographic feature, such as the case of Maryut Lake. When assessing the results of such studies, the following conditions must be considered to increase the reliability of the product: ata m st e a rate to all the region of st d the more accurate the data the better the results). ata ma s and satellite images m st e a essi le This is a particular problem when investigating the Maryut Lake region where some maps of some of the series were not found due to negligence of archiving or due to security reasons.35 ata m st e dated reg larl to make the st d of the region reliable and more accurate. In the case of the present study, the last series of maps produced concerning the whole region was in 1991. This means that for the last 17 years, no other map series concerning the Mareotic region have been produced, despite continued development and huge observable changes having occurred in the Mareotic region since this time. Bibliography D’ Abbeville, N.S., 1655, R oyaume et dé sert de B arca et l ’ Egypte [ MAP] . Bellin, J.N., 1764, Carte des Embouchures du N il , et partie de son cours [ MAP] .

35. Several military zones exist around the Mareotic region causing the exclusion of some maps which include these.

23

LAKE MAREOTIS: RECONSTRUCTING THE PAST

24

L. BLUE: LAKE MAREOTIS RESEARCH PROJECT

Lake Mareotis Research Project Lucy Blue

Alexandria was by any standards one of the great cities of the Mediterranean. Since its foundation in 331 BC and for almost a millennium to follow, it was the political, economic and cultural capital of Egypt, and one of the most signifi ant em oria and om le orts in the Hellenistic and Roman worlds. As such, Greco-Roman Alexandria has been the subject of much historical and archaeological research (Fraser 1972; Haas 1997; Empereur 1998; Goddio, et al. 1998) that has revealed the wealth of the city, much of which was generated by trade through its important and monumental harbour. However, the complexity of the Alexandria harbours system, which included not only harbours on the sea, but also on Lake Mareotis, has ne er een thoro ghl nderstood he signifi an e of Lake Mareotis in the history of the Greco-Roman port-city of Alexandria, is undeniable (Fig. 1). Settlements and industrial units located along the shores of Lake Mareotis, located to the south and west of the city, are known to have produced amongst other things, glass, textiles, pottery and wine (Empereur & Picon 1998), all of which were transported to the city. Y et the dynamics of this productive hinterland have not been fully appreciated in terms of the important contribution the region made to the economy of Alexandria and thus to Egypt as a whole.

The principal objective of the Lake Mareotis Research Project was to undertake a detailed systematic survey of the western extension of the lake in order to record, quantify and assess the archaeological resources of the area in a comprehensive manner. As a result, we hoped to determine a better understanding of the nature and extent of economic and maritime activities in the Mareotic region and to explore the relationship between the component parts of this complex system and how they developed and changed over time, in order to determine a more comprehensive understanding of the economic functions of ancient Alexandria.

In this context, the Lake Mareotis Research Project set out to investigate and determine the role of the lake, the function of the sites around its shores and the part they played in supporting the metropolis of Alexandria in antiquity.

In ancient times, Lake Mareotis was connected to and fed by the Canopic Branch of the River Nile (Strabo 17.1.7). By the 12th century the lake had silted up but prior to its silting it was much larger and deeper (Warne & Stanley 1993: 53-8; Said 2002). The main body of Lake Mareotis currently covers an area of about 90 km2 and is defined to the north a hain of lithified ar onate ridges e tending parallel to the northern shoreline of the lake and the Mediterranean coast (Fig. 2). It is nearly rectangular in shape and extends south of Alexandria for about 12 km, merging along its southern and eastern shores into the Western Deltaic region. It would once have extended considerably further south, quite possibly a navigable distance of about 50 km. It is separated from the lake’ s main body by causeways and shallows and its relative size varies according to the season.

E nvi ronmental and Political Context of L ak e M areotis Lake Mareotis represents one of the most distinctive geomorphological features along the north-west coast of Egypt (Warne & Stanley 1993). It is unique compared to other lakes on the north coast of Egypt, being the only one with no direct connection to the Mediterranean Sea, but originally with direct access to the River Nile and hence to the whole of Egypt. Therefore, it is believed that the location and characteristics of Lake Mareotis gave the city of Alexandria one of its major advantages (Strabo 17.1.7).

Context of the L ak e M areotis Research Project The Lake Mareotis Research Project is a collaboration between the Centre for Maritime Archaeology, at the University of Southampton, and Department of Underwater Antiquities (DUA) of the Supreme Council of Anti ities , le andria he ro e t first started in 2004 when a British Academy Small Grant was awarded to support a pilot survey season in the western arm of the lake, to the west of Alexandria. The survey covered an area of approximately 40 km east-west by 3 km wide, along the shores of the western extension of Lake Mareotis, from idi erir to l- ammam ring this first season o er 60 sites where identified, man e ating to new dis o eries (Blue & Ramses 2005). The potential wealth of the sites in the area was thus realised and subsequently further funding was sought and awarded from both the British Academy and the Leverhulme Trust to conduct a more detailed survey on both shores of Lake Mareotis and on an island that is located in the eastern region of the survey area.

Thus at present, Lake Mareotis is a body of shallow brackish water that occupies about 13% of an extensive subsea-level depression that is sometimes referred to as the Maryut Depression. The remaining 87% of the depression has been drained and is now used primarily for agriculture. The main body of the lake is heavily polluted by untreated sewage and industrial waste. It has also been divided into many sub-basins which are used for industrial and agricul25

LAKE MAREOTIS: RECONSTRUCTING THE PAST

F ig. 1:

L ocation of A l ex andria and L ake Mareotis ( L ake Mareotis R esearch P roj ect) .

26

L. BLUE: LAKE MAREOTIS RESEARCH PROJECT tural purposes. Moreover, the shores of the lake are subject to continuous irrigation, drainage and reclamation, which means that it is constantly changing and unstable (Warne tanle 1993 29-30 rih , et al 1996 282 he a erage depth of the remaining lake is less than 1 m, although it is assumed to have been greater in antiquity to allow boats to sail safely.

and Boussac & El Amouri (this volume) on the Hellenistic port of Taposiris Magna. In addition, rescue work has been undertaken by the Supreme Council of Antiquities (SCA) at a number of smaller sites around the shores of the lake largely in advance of development. The majority of known sites have not been systematically studied however, and the pilot survey conducted in 2004 revealed that only a fraction of the available evidence is currently recognised. In addition, as indicated previously, the area is massively under threat from development and pollution. Large sections have been quarried, particularly of the northern limestone ridge of the study area, while extensive sections of the remaining lake are being exploited for aquaculture and its shores cultivated for agriculture. Much of the land has been sold for development, and along the edges of the lake harmful polluting industries such as salt factories, geochemical and petrochemical processing plants, as well as numerous sewage plants, all present a real threat to this lacustrine environment – the landscape is constantly changing and with it the archaeological record is being destroyed (see Ramses & Omar this volume). For all these reasons it was determined that a comprehensive survey of the area was essential in order to record what remains before the information is lost forever.

Due to its location and characteristics, Lake Mareotis la ed a signifi ant role in the internal and international transport system of Greco-Roman Egypt (Rodziewicz 1998), particularly after it had been connected to the sea by a navigation canal at Alexandria (Strabo 17.1.7: 31). It became an important conduit of communication between the River Nile and the Mediterranean Sea, and river boats would have transported trade goods from the Nile Valley to the harbours of Mareotis and onwards to Alexandria and beyond. As indicated, Mareotis also supported economic activities around its shores such as the cultivation of vines and grain, and the manufacture of ceramics, wine and glass (Empereur & Picon 1998). However, despite the important role of Lake Mareotis, previous research conducted in the region has been largel limited to s e ifi areas and to i s, s h as the work of El-Fakharani (1983) amongst others, on the Byzantine port of Marea (Philoxenite) (see now Haggag, a ra ma ska, od iewi and i hot this olume), Empereur & Picon (1998) on the amphorae and wine production in Mareotis (also see Dzierzbicka this volume),

F ig. 2:

Objective s of the L ak e M areotis Research Project The objectives of the collaborative Lake Mareotis Research project were to:

A n approx imation of the ancient and present l imits of L ake Mareotis ( E. K hal il ) . 27

LAKE MAREOTIS: RECONSTRUCTING THE PAST stemati all s r e the western arm of ake Mareotis (shores 40 x 3 km wide, and Mareotis Island,1 3 7 km in length and to 680 m wide , bearing in mind the complex palaeogeographical history of the lake. e ord, antif and assess the ar haeologi al resource. etermine the e tent, hronolog and, where possible, function, of each site. etermine the nat re and the e tent of the e onomi and maritime activities of the region. etermine the signifi an e of the areoti region in relation to Alexandria, particularly during the Hellenistic, Roman and Byzantine periods when the lake is believed to have been most active. dentif the degree of threat to the ar haeologi al resource in relation to each site before it is lost to development and pollution.

l gate adiometer agnetometer geo h si al survey was conducted to establish the layout of str t res eneath the s rfa e at s e ifi sites 2 erami s r e was ndertaken to determine the chronology, nature and function of each site. edimentologi al s r e to determine hanging lake levels and identify geomorphological changes was conducted at selected sites. imited site- learing to arti late walls and identif more detailed building phases, was undertaken. hotogra hi ar hi e was om iled of ea h site and features within the site. nal sis of the fi e s r e seasons ond ted in the western arm of Lake Mareotis between 2004 and 2008 is almost complete, and brief preliminary observations are presented in this a er he final res lts will e lished in a single monograph detailing the work undertaken and an interretation of the final res lts l e halil forth oming In total, over 70 sites were documented, ranging in date from the Hellenistic period to the 7th century AD (Fig. 3). A clear distinction was noted between the nature of the sites on the northern shores compared to those in the south and on Mareotis Island. The north shore sites were clearl defined, t ite hea il eroded, tell sites he sites to the south appear to represent more dense urban concentrations with associated agricultural and industrial areas, particularly towards the east where the island ridge may have once been an extension of the southern shore and the ridge associated with the site of Marea. Many of the sites, particularly on the southern shore, were partially buried under aeolian and alluvial sediments deposited by the eroding hillside and wind-blown material from the north.

M ethodology As a result of the pilot survey conducted in 2004 a total of o er 60 sites were identified, as re io sl mentioned Some of these sites were already documented in the archaeological record, some having been partially excavated by the Western Nile Delta section of the SCA and foreign missions, or noted but not recorded in detail in previous li ations owe er, at least half of the sites identified are believed to be new discoveries. The sites were prioritised on the asis of their maritime signifi an e, arti larly where maritime installations such as moles, jetties, a s, mooring rings et , had een identified oordinates were taken to fi the osition of ea h site lo ated, each site was photographed and basic data pertaining to its nature were recorded. Since 2004 four further seasons of s r e ha e een ond ted etween a t mn 2006 and summer 2008, with the objective of recording in detail the maritime installations and to place the sites in their environmental and economic context. In order to achieve these objectives, the following methodology was implemented: desk- ased assessment of the region was ndertaken with the aid of high resolution Q uickbird satellite images, to establish the nature and extent of the sites alread identified his information along with all the data acquired as part of the survey project, is archived in a GIS-based database. a h site was s r e ed sing a - ased eal ime Kinematic (RTK) satellite navigation system to produce detailed topographic maps of each site. e ifi ildings and/or feat res were lanned sing a Leica TCR705 Total Station, and the data was downloaded in the field into to ia heo t software.

N orth S hore S ites (Fig. 3) total of ten sites were identified on the northern shore They date from the Hellenistic to Late Roman periods with a concentration of activity in the Roman period. They were primarily tell sites and their size ranged from c. 7000 to 180,000 m2. The ceramic assemblages at the sites included a mi t re of am horae, finewares, ta lewares and ooking wares an of the finewares and am horae were im orts Many of the sites on the northern shore contain wells and cisterns as well as evidence for mortar-lined basins, probably from bath complexes, which would suggest a domestic context. Limited evidence of irrigation exists in the form of water wheels or sakkias, and qadus pots associated with water wheels or sakkias were found at some Early and Late Roman sites. The evidence suggests that these sites were largely domestic in nature and each site represents a fairly contained and distinct unit with a clear association with

2. Preliminary investigations in the 2007 season were not as insightful as we had ho ed, d e to the fa t that the fl gate radiometer 36 we had been using had limitations. Thus, the 2008 geophysical survey was ond ted with a artington nstr ments rad 601-2 d al sensor fl gate gradiometer. The dual sensors survey at twice the rate of a single sensored instr ment and with more refined res lts

1. The Lake Mareotis Research Project refers to this ridge, largely surrounded by water located in the south-eastern region of the survey area some distance from the southern shore of the lake, as Mareotis Island. However, its island status in antiquity has yet to be determined and it is also acknowledged that it is not, and was not in antiquity, the only island in the lake.

28

L. BLUE: LAKE MAREOTIS RESEARCH PROJECT

F ig. 3: Sites l ocated al ong the northern shorel ine, t he southern shorel ine and on Mareotis I sl and ( L ake Mareotis R esearch P roj ect) . sites extend across the lake towards each other. Thorough investigation of the ceramic samples collected from these sites will shed more light on their possible nature and relationship.

the waterfront. A number of the settlements featured jetties extending into the lake on their southern shores. The survey also revealed possible spatial and functional relationships between a number of sites on the northern shore and s e ifi sites on the island and along the so thern shore. For example, it was noticed that at least one of the largest sites on the northern shore of the lake (AlGamal, Site 204) is located immediately opposite a settlement located at the western end of the island (Site 23; see Hopkinson this volume). Q uays and jetties from these

S outh S hore S ites (Fig. 3) total of 26 sites were identified on the so thern shore The sites ranged in date from the Hellenistic to the Late Roman periods, but generally exhibited limited activity in the Early Roman period. The earlier period displays some imports, largely amphorae from the Aegean region, but 29

LAKE MAREOTIS: RECONSTRUCTING THE PAST

is predominantly composed of locally produced material, while the Late Roman material reveals a greater amount of imported pottery from all over the eastern Mediterranean.

F ig. 4: T opographic, c eramic and auger surv ey pl an of Site 4 on M areotis I sl and ( image L ake Mareotis R esearch P roj ect) .

The sites towards the western extremity of the southern shores formed distinct and in some cases (Site 44 and Site 109) quite substantial units, and are spaced well apart. The archaeology at these sites was largely concentrated at the foot of the limestone ridge (Fig. 4). The sites extended north towards the shore of the lake where many exhibited a series of linear features that extend parallel to the shoreline down to the water’ s edge and were open to the E-NE. They appeared to support low-lying areas of marshy ground between the main occupation area and the shoreline ridge, or in many cases a boundary wall to the north. This type of linear feature, known as a ‘ lake wall,’ essentially contained a body of water and could have been used for agricultural purposes, perhaps for feeding animals or harvesting reeds, as is still practiced in the region today. This was noticed at a number of sites and also towards the western end of the island (see below) (Fig. 5).

F ig. 5: A n ex ampl e of a ‘ l ake w al l ’ feature ( Site 4) ( photo L ake Mareotis R esearch P roj ect) .

To the east, the sites were located on a limestone ridge that is effectively an extension of the ridge upon which both Mareotis Island and the site of Marea are situated. The nat re of these sites a ears to refle t similar dates and a tivities to that of the eastern end of the island. A number of sites ha e een identified that ontained wine rod tion fa ilities and/or sakkias designed to facilitate the shifting of water for e am le ite 100, and ites 01 and 106 Generally, however, these sites are less substantial than 30

L. BLUE: LAKE MAREOTIS RESEARCH PROJECT of a few erami s re io sl identified l- akharani l- akharani 1983 176-178 herefore, these ildings might contribute towards resolving the mystery of the so far el si e haraoni har o r of area raser 1972 146 l- akharani 1991 28, 1983 176

those on the island or by comparison with Marea, and they tend to cluster opposite the western end of the island, or along the ridge that extends between Marea and the island. Further investigation is needed to identify the relationship between the eastern end of the island, the south-easterly sites and the western extent of Marea.

M areotis Island (Fig. 3) Mareotis Island is a limestone ridge joined to the southern shore at its western end. It is possibly an adjunct of the southern shore that has extended out into the lake from the Marea Ridge. It is aligned roughly northeast – southwest and is 3 7 km in length and 680 m at its widest oint ome 40 sites were initiall identified rimaril on entrated at the eastern and western ends of the island, with a few sites in its central section that were largely industrial-scale Early Roman kiln sites (see Hopkinson this volume).

Two sites stand out as exceptions to the others noted along the southern shores; the Kibotos (the square-shaped harbour) (Site 09), and the ‘ complex building’ (Site 13) (Figs. 6 7 he i otos ite 09 is essentiall a har o r area constructed of large limestone blocks (0.8 x 1 m), defined a series of moles that en lose an area some 60 m long to and 36 m to wide see halil this volume). A second site, known as the ‘ complex building’ (Site 13), is located a short distance to the southwest of the Kibotos and is composed of a continuous wall constructed of huge blocks (1.3 x 1.1 m), with a smooth outer surface that slopes up and inwards from the foot of the building. It has been proposed that both these buildings date to the Pharaonic period (Haggag 1984: 277-280). No earlier haraoni material has et een identified at an of the sites recorded in the survey area, with the exception

The majority of the material is Hellenistic in date, particularl towards the eastern end of the island, with a fl rr of activity again in the Late Roman period, particularly towards the western end of the island. The ceramic assemblage indicates some imported material especially for the earlier Hellenistic period, but in general it is locally

F ig. 6: T he K ibotos site ( Site 09) ( photo L ake Mareotis R esearch P roj ect) .

F ig. 7: T he ‘ compl ex buil ding’ ( Site 13) ( photo L ake Mareotis R esearch P roj ect) .

31

LAKE MAREOTIS: RECONSTRUCTING THE PAST produced, particularly the production of amphorae. The nature of the structures, with the exception of the centrally located industrial-scale amphorae production kiln sites, is essentially urban in character. The scale of the buildings, particularly those located on the ridge towards the east, is impressive, and many are associated with numerous waterfront structures located along the northern shore on the edge of the coastal plain. At the western end of the island, the settlement e tends from the ridge to the north o er a flat plain. Whereas the ridge supports further urban structures, the plain to the north appears to be more agriculturally orientated, with features including possible ‘ lake walls’ and sakkias similar to those identified along the so th-western shores of the lake (see above).

the lake’ s Mareotic Arm (see Khalil this volume). The frequency and abundance of these installations along the shores of the areoti rm refle t the e tent of maritime activity that was taking place in this region. Besides the jetties and quays which are positioned perpendicular to the shore, another type of maritime installation has also een identified, ositioned arallel to the shore, and des ri ed as a lake wall he so- alled lake walls define the shores and protect them from the effects of silting and sedimentation s man as fi e lake walls were dis o ered along the southern shore of the lake and the northern shore of Mareotis Island, areas which are more subject to silting as a result of the prevailing northwest wind. The lake walls could have been utilised as docking facilities for merchant vessels, but were more probably associated with agricultural activities, perhaps the cultivation of reeds as is still witnessed in the region today.

There is also evidence for quarrying on the island, as well as rock-cut tombs. Towards the eastern end of the island, a ossi le town wall has een identified together with a series of jetties and platforms associated with the waterfront, possible warehouses and a small inlet embayment. All these are discussed in more detail in Hopkinson (this volume).

The Lake Mareotis Research Project has also recorded numerous archaeological sites of industrial and agricultural nature such as amphorae production sites, wineries and the remains of several water wheels (sakkia) which were used for lifting water for irrigation and for use in baths and other domestic purposes. With one exception, all the amphorae and wine production sites are located on the southern shore of the lake or on Mareotis Island. In antiquity, the southern shores of the lake were characterised by a fertile plain which supported and enabled agrilt ral a ti ities to flo rish n addition, the a ndan e of calcareous clay, particularly suitable for amphorae production, resulted in a thriving large-scale amphorae industry. The study of the ceramic assemblages collected from the surveyed sites indicate that the amphorae production centres were active from the Ptolemaic to the Late Roman period.

What is worthy of note is the fact that the majority of sites on the island are concentrated along the ridge and the northern shores, and that several were clearly built in a er s e ifi relationshi with sites on the lake s north shore (see above). The northern concentration of sites has also raised questions about the island’ s status. Does the fact that so few sites have been found along the island’ s southern shores suggest that it may have been linked to the southern shore of the lake in antiquity? Recent geomorphological work suggests that in fact the area to the south of the island has always been waterlogged to some degree, but that the waters were shallow and access by boat was problematic, and as such activities in this area were restricted.

Since no lake-side archaeological sites were recorded on the southern shore of the lake opposite Mareotis Island, questions have been raised about the nature of the island and whether or not it was indeed an island in antiquity. However, sedimentological analysis carried out during the final s r e season re ealed that areotis sland was probably an extension of a ridge that extended some 10 km to the west from the site of Marea. In addition, it was ascertained that the area between the island and the southern shore of the lake was either waterlogged or marshland in antiquity (Flaux forthcoming).

Preliminary Conclusions he ake areotis esear h ro e t has identified and s stematically surveyed over 70 sites dating from the Hellenistic period through to the 7th century, in what still remains of the western arm of the once much larger Lake Mareotis. The sites vary extensively in nature, size and function, but demonstrate the importance this area had in relation to Alexandria and the mechanisms by which Mareotic products were produced and arrived at the city. Local pottery production is indicated by the numerous kiln sites discovered in the region, particularly on the island. There is also considerable evidence for imported pottery from as early as the late 4th century BC and as late as the 7th ent r he nat re and si e of the sites identified aried from rural small holdings to large urban settlements, with associated production sites, water storage facilities and agricultural and industrial complexes.

Previous research conducted in the Mareotic region has een largel limited to s e ifi sites howe er, the ake Mareotis Research Project has systematically investigated all the archaeological remains along the shores of the lake’ s western arm. The project also assessed the results of previous archaeological work undertaken in the Western Deltaic region. Based on this, it is reasonable to suggest that the Mareotic Arm was in fact the most active area in Lake Mareotis in antiquity, and that most of the Mareotic products arriving in Alexandria were in fact coming from the western Mareotic Arm, rather than from settlements associated with the lake’ s main body (see Khalil this vol-

Different types of waterfront installations were recorded, including more than ten anchorage facilities, which form the majority of maritime installation along the shores of 32

L. BLUE: LAKE MAREOTIS RESEARCH PROJECT que d’ A thè nes, organisé par l e CN R S, l e L aboratoire de cé ramol ogie de L yon et l ’ Ecol e franç aise d’ A thè nes, 1 1- 12 dé cembre 198. Athens, Paris. Empereur. J.-Y . (ed.), 2002, A l ex andrina 2 . Cairo. Empereur, J.-Y ., & Picon, M., 1998, Les ateliers d’ amphores du Lac Mariout. In J.-Y . Empereur (ed.), Commerce et artisanat dans l ’ A l ex andrie hel l é nistique et romaine. A ctes du col l oque d’ A thè nes, organisé par l e CN R S, l e L aboratoire de cé ramol ogie de L yon et l ’ Ecol e franç aise d’ A thè nes, 1 1- 12 dé cembre 198: 75-88. Athens, Paris. Flaux, C., forthcoming, Geomorphological survey of the south-west shores of Lake Mareotis. In L. Blue & E. Khalil (eds.), A Mul tidiscipl inary A pproach to A l ex andria’ s Economic P ast: the Mareotis Case Study. Oxford. Fraser, P.M., 1972, P tol emaic A l ex andria, Vol. I. Oxford. rih , , 1996, ome ro osals for oastal anagement of the Nile Delta Coast. O cean & Coastal Management 30.1: 43-59. Goddio, F., Bernard, A., Bernard, E., Darwish, I., Kiss, Z ., & Y oyotte, J., 1998, A l ex andria: T he Submerged R oyal Q uarters. Oxford. Haas, C., 1997, A l ex andria in L ate A ntiquity: T opography a ia i . Baltimore. Haggag, M., 1984, P orts in A ncient Egypt til l the A rab Conquest. Unpub. MA diss., Alexandria University. Hassan, F.A., 1997, The Dynamics of a Riverine Civilization: A Geoarchaeological Perspective on the Nile Valley, Egypt. W orl d A rchaeol ogy 29.1: 51-74. Rodziewicz, M.D., 1998, Fom Alexandria to the West by Land and by Waterways. In J.-Y . Empereur (ed.), Commerce et artisanat dans l ’ A l ex andrie hel l é nistique et romaine. A ctes du col l oque d’ A thè nes, organisé par l e CN R S, l e L aboratoire de cé ramol ogie de L yon et l ’ Ecol e franç aise d’ A thè nes, 1 1- 12 dé cembre 198: 93-103. Athens, Paris. Said, R., 2002, Did Nile Flooding Sink Two Ancient Cities? N ature 415: 37-38 Strabo, G eography, H.L. Jones (transl.), 1949. London. Warne, A.G., & Stanley, D.J., 1993, Late Q uaternary Evolution of the Northwest Nile Delta and Adjacent Coast in the Alexandria Region, Egypt. J ournal of Coastal R esearch 9 1 26-64

ume). Finally, the Lake Mareotis Research Project has outlined the true scale of activities in the Mareotic Arm with respect to industrial, agricultural and urban sites, being much more extensive than previously suggested, thus s orting the signifi ant im ortan e of the area and highlighting the role it played in the economy of Hellenistic and Roman Alexandria. Ack now ledgments The author would like to thank the Supreme Council of Antiquities for permission to work in the Mareotis region, particularly colleagues from the Underwater Department who were instrumental to the success of this project, in particular the support and help of the project co-director Dr Sameh Ramses and project manager Dr Emad Khalil. The ro e t wo ld not ha e een ossi le witho t the finan ial backing of both the British Academy and the Leverhulme Trust particularly for funding Dr Khalil’ s two-year post doctoral fellowship. Finally, thanks is due to all students of Southampton and Alexandria Universities and Egyptian and UK colleagues who worked on the project without whom the survey would not have happened. Bibliography El-Fakharani, F.A., 1983, Recent Excavation at Marea in Egypt. In G. Grimm, H. Heinen & E. Winter (eds.), A egyptiaca T rev ernsia I I , D as R öm isch- B yz antinische Ä gypten. A kten des international en Symposions 26.30. September 1978 in T rier 175-186 Mainz. El-Fakharani, F., 1991, The Kibotos of Alexandria. Studi Miscel l anei 28: 21-28. Blue, L., & Ramses, S., 2005, Lake Mareotis Research Project – Preliminary Report from Pilot Survey, August 2004. N ew sl etter of the A rchaeol ogical Society of A l ex andria 1 5 ( F ebruary) 5-16 Blue, L., & Khalil, E. (eds.), forthcoming, A mul tidiscipl inary approach to A l ex andria’ s economic past: T he Mareotis case study. Oxford. De Cosson, A., 1935, Mareotis: B eing a Short A ccount of the H istory and A ncient Monuments of the N orthW estern D esert of Egypt and L ake Mareotis. London. Empereur, J.-Y . (ed.), 1998, Commerce et artisanat dans l ’ A l ex andrie hel l é nistique et romaine. A ctes du col l o-

33

LAKE MAREOTIS: RECONSTRUCTING THE PAST

34

D. HOPKINSON: MAREOTIS ISLAND SURVEY

The Results of a Preliminary Survey at Mareotis Island Dylan Hopkinson

Introduction

northern and southern limits of the western arm of Lake areotis and form a nat ral flooded de ression known as the Mallahet Maryut Depression (Fig. 1). The island measures 3.7 km long by an average of 400 m (maximum 680 m) wide and is connected to the mainland by an area of low-lying marshy ground to the southwest at the foot of the Gebel Mariout Ridge (Fig. 2). The lands around Lake Mareotis formed an important agricultural production centre in antiquity known to have been richly populated with farming villages, market-towns and villa estates (Strabo G eography 17.1), and would have been dependent on seasonal rainfall and ile floodwaters for their irrigation and fresh drinking water.

L ocation The site known as Mareotis Island lies 65 km southwest of Alexandria within the western arm of Lake Mareotis, closely associated with its southern shore. It lies 4 km to the west of the site of Marea/Philoxenité, and directly north of the modern town of Bahig (Fig. 1). The island was subject to systematic survey in 2007 and 2008 as part of the broader Lake Mareotis Research Project (see Blue this volume). A large number of archaeological features are clearly visible on the ground which date to the Greco-Roman periods largely indicated by the substantial quantities of ceramics present in surface scatters. This paper presents the overall character of the archaeological features observed, and some reliminar inter retation with s e ifi referen e to the littoral zone on the northern coastal plains where ‘ maritime’ and lakeside activities were largely focused.

Env ironmental B ackground In antiquity Lake Mareotis provided a transport connection between Alexandria and its agricultural and industrial hinterland in the Mareotic region; and via the Nile waterway to the Egyptian interior along the Nile Valley. Strabo states that the lake is filled man anals from the ile, both from above and on the sides” (G eography 17.1.7). It was these canals that provided the fresh water which came

The island is formed from the remnants of a small limestone ridge which lies between the two much larger Abusir and e el ario t idges hese two ridges define the i

h

ai

a

is s a

a

h si s

i

i

35

i

ss

LAKE MAREOTIS: RECONSTRUCTING THE PAST F ig. 2: Schematic of the main features of Mareotis I sl and ( H opkinson) .

from the Canopic Branch of the Nile apparently keeping the lake water sweet, and filling the isterns of le andria he dire t link to the ile d ring the seasonal floods is also likel to ha e had a signifi ant im a t on lake le els The exact location and number of these connections is not firml nderstood, t it is likel that the an ient o rses would have mirrored those of the more recent canals (Warne & Stanley 1993: 15-16).

of sites located on the southern shore of the lake, where Roman foundations are observed to have been buried under 3 m of deposits (Khalil 2005: 39). Holocene sedimentation rates between 0.02 and 0.07 cm per year are indicated by Smithsonian boreholes either side of the island (Warne & Stanley 1993: 50). This sedimentation is likely to have obscured archaeological features over time, and to have raised the level of the lake bed while consequently reducing the depth of the lake waters (Flaux 2008: 9-10, III.2).

here is s ffi ient en ironmental e iden e to show that the limate of the region has not hanged signifi antl in the past two thousand years (Ball 1942: 20; Brooks 1949: 333). Evidence for this includes an analysis of a weather diary kept in Alexandria by Ptolemy in the 2nd century, which concludes that “ the number of rainy days per annum was much the same as it is nowadays, though the rainfall was more evenly spread out over the year” (Murray 1951). So we can conclude that the climate was semi arid as it is today. Today rainfall in the Western Coastal Desert region is greatest along the margin the sea and is t ified its great variability from location to location and from year to year, which suggests that in antiquity even this fertile and productive region would have had a precarious existence were it not for the fresh waters of the lake and the reliable ann al ile floodwaters

i ai s h ak si the G reco- R oman P eriod he water regime in the area has een greatl modified since antiquity, most notably by the silting of the connections between the lake and the Nile; when the Canopic ran h of the ile finall silted aro nd the 12th century the lake also began to dry up and shrink in size. Strabo also tells us that the extent of the lake must have been much greater in antiquity than it is today, and extended around 40 km to El Bordan to the southeast and to Kom Al Trouga 70 km to the southwest of Alexandria (De Cosson 1935: 26) (see Fig. 1). Warne & Stanley also support these dimensions which amount to around 700 km2 of land lying below the 0 m sea-level datum (Warne & Stanley 1993: 53).

Light north-westerly winds are generally observed in the region and these transport large quantities of clayey loess sediments from the degrading limestone ridges and silts and sand from the coastal plain. These silts are deposited in colluvial sheets at the base of the ridges and on north facing shores. This phenomenon has been noted previously (Warne & Stanley 1993: 53), and at a number

It is very likely that even without a direct connection to the lake the effe ts of the ile flood wo ld still ha e een noticeable in the water levels due to groundwater seepage (A.G.,Brown, Univ. of Southampton, pers. com.). When the Aswan Dam was constructed in 1964 there was therefore a signifi ant hange to the water le el in the ile 36

D. HOPKINSON: MAREOTIS ISLAND SURVEY Delta aquifers and the lake. As a result the hydrodynamic environment today is very different from antiquity and we have no direct way of understanding the range of water le els that wo ld ha e een e erien ed d ring the floods As Ramses & Omar (in this volume) indicate, modern infrastr t re, fisheries and agri lt ral drainage dit hes ha e f rther modified the lake, s litting it into se eral artifi ial asins and remodelling large stret hes of oastline hese asins ha e water regimes that are artifi iall managed for modern needs, and large amounts of water are m ed into the sea to kee the waters artifi iall low for land reclamation projects.

this sakia to the cisterns of the town. To the south of this are the remains of two very interesting pottery kilns with a large heap of broken pottery thrown there as it was ‘ scrapped’ from them. North of the sakia there is a long jetty running into the Lake from which ferry-boats once plied.” (De Cosson 1935: 130)

The reference here to waterwheels (‘ sakia or ‘ saqiya’ ), cisterns, ceramic mounds from pottery production and lakeside features characterise the island in general terms, but the description bears little direct relation to the position of the remains observed today. Consequently the archaeolog of areotis sland re resents a signifi ant nta ed resource that can enhance our understanding of the region.

Today the lake is fed only by water seepage from the sea, precipitation, and by irrigation ditches that channel agricultural waste-water from the Beheirah Governorate. As a result of all of these changes, it has been calculated that the current lake level is only 13% of the ancient lake surface area (Warne & Stanley 1993). Observations made during the 1970s show that the western arm of the lake was totally dry at least from Marea/Philoxinite and to the west. However, a recent geoarchaeological survey as a component of the Lake Mareotis Research Project indicates that the depositional environment close to the modern shoreline had always been lacustrine except during the recent centuries (Flaux forthcoming).

F ieldw ork In August 2007 under the auspices of the Lake Mareotis Research Project, a six-week survey was undertaken to plot the visible archaeological remains on the island using RTK GPS and Total Station surveys. In order to provide an understanding of the dating of features a controlled surface ceramic collection strategy was adopted. An RTK GPS topographic model was also completed with 5 m transects across the island. In addition to this, a geoarchaeological auger survey was undertaken in 2008 to help contextualise the results and to begin to understand how site formation processes can inform our understanding of the site. Of particular interest here was the relationship between the ancient water levels and the archaeological features, and the impact of the changing lake levels on the development of the sites. A more detailed account of the Lake Mareotis Research Project and its methodologies is presented in this volume (see Blue this volume, see also Khalil and Ramses & Omar this volume)

Deltaic subsidence and tectonic shifts may also have changed the topography of the region as it is known to have done within the Nile Delta region further to the east; however it has not proved possible to use existing geoar haeologi al datasets to nderstand the signifi an e of these events. Consequently as a result of the various geographical factors described, we must accept that the current lake level cannot be used as a reference datum from which to reconstruct ancient lake levels and to assess the im a t of the ann al floods and the s mmer e a oration on the shoreline. However, understanding these relationships is a fundamental factor to our interpretation of the archaeological structures lying near the shore.

The majority of archaeological structures are located on high ground, occupying the central limestone ridge that forms a series of hills that extends as a discontinuous spine along the length of the island. Two main settlements were identified on entrated at the eastern and western e tremes of the island, with further activities evident on the north facing coastal plains below the central ridge. The central section of the island is some 2 km long and is much less densely occupied than the western and eastern extremities. The activities here represent industrial amphorae production and possible agricultural activities.

P rev ious R esearch here has een a signifi ant in rease in o r nderstanding of the archaeology of the Mareotic region in recent times brought about by the work of El-Fakharani (1974, 1983, 1984), Rodziewicz (1983, 1990, 1998a, 1998b, 2002), and Empereur (1986, 1998; Empereur & Picon 1986, 1998, 1992), however previous research on the island itself appears to have been negligible. There are numerous cartographic representations of the region, two of the most informative being the one on Sheet 37 of the A tl as G é ographique in the s i i y from 1809, and on a s i s a i produced in 1866 by Mahmoud Bey El-Falaki; however, the detail pertaining to the island in oth ases is limited and diffi lt to relia l geo-reference. A further cryptic description of the archaeology on the island is given by De Cosson (1935):

Chronology The ceramics recovered from the island indicate that the settlements was well established some time during the Hellenistic period and that occupation potentially continued through to the 7th century AD. The western settlement appears to have been continually occupied as indicated by ceramic forms recovered, whereas the eastern settlement seems to have suffered a decline in activity in the Early Roman period which reached its nadir some time in the 4th to 5th centuries with a late recovery in the 6th century. The results of the ceramic survey are not commented on in detail here, but will be published shortly (Tomber & Thomas forthcoming).

“ Many buildings can be traced, and at the eastern end the circular stone platform and oblong well of an ancient sakia will be found. Stone channels are traceable leading from

37

LAKE MAREOTIS: RECONSTRUCTING THE PAST Archaeology of the E astern S ettlement The eastern settlement (Fig. 3) is concentrated on the high ground of the limestone ridge which measures approximately 650 m by 350 m; the majority of this area is covered with building remains of various construction styles and materials. The layout and construction styles of the visible building remains indicate some degree of social organisation; such as possible administrative and commercial areas, some are grouped around what appear to be large, open areas at a number of locations on top of the ridge. A long boundary wall is located winding along the top of the ridge in the north-western limit of the settlement; this wall measures 240 x 1.20 m wide, and may have served a defensive function protecting the settlement from the west. To the west of the settlement there is a small low-lying depression that could well have formed a natural inlet prior to silting; around its northern and southern margins the remains of buildings that may have been possible wharf str t res ha e een identified his inlet a ears to have served as an inland harbour which would have provided sheltered mooring. Auger cores taken during the 2008 season onfirm that the sediments in this de ression were indeed deposited in lacustrine conditions in antiquity (Flaux forthcoming).

are diffi lt to inter ret in detail e a se there are se eral phases of activity. The relative dating of these features has not yet been possible. These structures appear to fall into several categories of linear features, walls and lakeside buildings. L inear F eatures ( G roup A ) he first ategor of str t res is a series of fi e linear features located on the low-lying northern coastal plain that extends between the present northern shoreline of the island and the foot of the limestone ridge (Fig. 3, Group A). These features range between 65 to 120 m in length, and appear as low, roughly built earthen ridges up to 4 m wide which are spaced roughly between 30 to 100 m apart. At the waters edge, additional features can be seen that extend along the current shoreline. As yet, these water front feat res ha e an ndefined hronologi al relationshi to the linear features (Group A) and represent primary and secondary activities over a period of time. The landscape between the base of the limestone ridge and the lakeside features appears to be largely devoid of additional structures and the linear features are relatively isolated with only a few abutting or adjacent features. It has already been established that little is known about the absolute water-levels of the lake in antiquity, and unfortunately the 2008 geoarchaeological survey did not target the area around the linear features therefore our interpretations

To the north of the eastern settlement, along the northern coastal plain of the island, between the lake and the foot of the ridge, are a series of structures and buildings which i h i ( H opkinson) .

a

a

s

i

i

his a

a

a hi s

38

y

h

as

s

h is a

D. HOPKINSON: MAREOTIS ISLAND SURVEY of the exact nature and function of these features remains speculative, however two possibilities present themselves. Firstly, they may represent crude earthen jetties projecting from the ridge into the lake during periods of higher water-levels. The other possibility is that the coastal plain was dry in antiquity and that the earthen ridges may have divided the coastal land into units, potentially for cultivation as is suggested by the interpretation of similar features identified in the western settlement e know from an ient sources that the Nile had a dramatic effect on the Delta region; Strabo states “ at the rising of the Nile the whole country is under water and becomes a lake, except the settlements; and these are situated on natural hills or on artifi ial mo nds, and ontain ities of onsidera le si e and villages, which, when viewed from afar, resemble islands” (G eography 17.1.4). It is therefore possible that the annual variation in the level of the lake level was large enough for both scenarios to be valid.

‘ quay’ wall to demarcate the edge of the lake either as protection from inundation or to form a working platform, perhaps for the mooring and loading of boats. In some instances the construction of these sections of wall seem to be related to the linear earthen ridges of Group A, and it would appear that in some cases they were built to connect or at least respect, the linear features lakeside termini. The elevations recorded from various sections of these walls indicate that the surviving top course of the walls are not all at the same level but in fact exhibit a difference in height of over 1.83 m. This may indicate that either the top course of some of the sections is missing, or that these walls were built at different times and at different mean lake-levels rather than as a single planned ‘ quay’ building effort. as a i i s sso iated with the first gro of linear feat res ro A) are a large number of walls indicating the presence of a number of buildings that had been constructed along the shoreline in this area (see Fig. 3). Since the relative dating of structures has yet to be determined it is not possible to discuss the phasing of the structures. However, what is apparent is that if the linear features were jetties and if they were built at the same period as the coastal buildings, then these buildings would have been susceptible to serio s flooding t is ital that the relationshi s and hasing of these features is investigated as it will have important implications with respect to interpretation of the landscape and the features located within it.

L inear W al l s ( G roup B ) A second group of linear walls were observed to the west of Group A along the north-western shores of the lake in the eastern settlement area. This group is located to the west of the wall that winds along the top of the western side of the limestone ridge, and down onto the coastline (see Fig. 3). These linear features were observed as low rubble walls 0.50 m thick, their northern lakeside ends are close to the current waterline and appeared to extend inland for distances of between 25 to 60 m. None of these walls reached the foot of the limestone ridge and in most cases extended less than half way across the plain. Three of these walls were spaced approximately 35 m apart with a further group 30 m to the east that was composed of three roughly parallel walls in very close proximity. The coastal ends of the first three walls had asso iated low r le linear structures or platforms which extended for a short distance along the coastline.

Amongst these coastal buildings are a group of structures that deserve particular attention. Three isolated groups of ildings of similar onstr tion were identified at three different areas of the plain (see Fig. 3; Group D). Two of these are located on the north coast and a third on the eastern extreme of the island with a south-easterly coastal aspect. The buildings appear to be arranged in symmetrical blocks facing each other and around 13 m apart, each measuring approximately 10 m wide and between 30 and 55 m in length, and were aligned perpendicular to the coast. The building pairs consist of single rows of cellular rooms measuring roughly 10 m by 5 m. Their location and alignment relative to the coast suggests that their function was somehow related to the lake and that they could have been warehouses, shops or boathouses.

The alignment of the main walls of Group B is interesting since they were constructed on the same orientation as the linear features discussed above (Group A). However, the Group B linear features are not perpendicular to the shoreline since this section of the coast turns to the south and is west facing. As a result these features are on an oblique alignment relative to the current shoreline. The stone construction, relative close spacing, alignment to the coast, and their location to the west of the long winding ‘ boundary’ wall, suggest a distinct function as compared to the Group A linear features. One possibility is that these features may have been concerned with land reclamation or stabilisation, serving perhaps as a series of ‘ groins’ .

Summary Overall there appears to be a great deal of established activity in the eastern settlement. Buildings on the ridge settlement hint at developed social organisation and those on the northern coastline show a strong association with the lake. Features interpreted as an inland harbour, lakeside walls, and possible jetties, all suggest an established interaction with the lake. There are a far greater number of buildings on the northern coastline as compared to the south coast in this part of the island, which may suggest that the southern coast was possibly dry or marshy and therefore inaccessible to boats in antiquity.

sh a ak a s Large sections of dressed ashlar blocks are observed forming stretches of wall along the northern edge of the plain (see Fig 3). These walls closely respect the natural topography of the current shoreline and appear to have originally been laid as stretchers extending parallel to the shore. This might suggest that the walls were built as a sort of 39

LAKE MAREOTIS: RECONSTRUCTING THE PAST Archaeology of the W estern S ettlement The western settlement occupies a similar location to the eastern settlement, situated on the high ground of the limestone ridge which measures 820 by 200 m (Fig. 4). The area exhibits the remains of various buildings, with a mixed function suggested by the construction materials and surfa e finds erall the reser ation of these str t res is not good; a large area in the extreme west has been heavily damaged by earthmoving machines, and in other instances only patchy areas of construction and dislocated wall lines survive. As a result, it is not possible to identify any zoning of functionality or social organisation, although a few signifi ant gro s of str t res were o ser ed s in the eastern settlement the northern coastal plain appears to be the focus of activities. Once again it is the lakeside features that are the focus of interest in this paper although some of the structures on the ridge will be discussed.

A shl ar Shorel ine W al l A l ignment ( F eature E) The area along the waterline is densely overgrown with water reeds and has silted up making the water level very shallow and the ground marshy. Amongst the reeds can be seen the upper surface of sections of a long straight wall (see Fig 4). All the shoreline features in the western settlement appear to respect this northern boundary. The wall is ilt from limestone ashlars laid as stret hers with a fl sh north-fa ing as e t, s ggesting that it was laid in a s e ifi relation to the lake. The various sections of this feature appear to suggest a continuous wall with overall dimensions 0.40 m by as much as 245 m in length. No excavations of this feature were undertaken so it is not possible to indicate the number of courses or the height of the wall. The location and construction of this wall suggest that it has a signifi ant relationshi to the shoreline if water le els similar to those seen today are assumed. It appears to be a lakeside wall protecting the land behind or creating a working platform from which to conduct activities related to the lake. Some of the other features close to this wall appear to butt against this wall suggesting that it may also form part of a larger pattern of structures.

The features of the coastal plain are sparsely distributed and appear to fall into three clear categories: a section of ashlar wall aligned along the shoreline; long, thin linear walls forming low ridges perpendicular to the coastline; and lo alised masonr feat res t ified r elinear walls and associated ashlar lined pits, there is an overall lack of secondary activities such as the coastal buildings on the eastern coastal plain, which makes interpretation somewhat easier.

L inear W al l s ( G roup F ) The second group of features located on the coastal plain of the western settlement, is a series of roughly paral-

F ig. 4: T he l ittoral features mentioned in this paper and topographic surv ey of the w estern settl ement of the isl and ( H opkinson) .

40

D. HOPKINSON: MAREOTIS ISLAND SURVEY hi ler identifies 14 different t es of saqiya from Roman, Islamic, and modern contexts (Schiøl er 1973: 13), including wheels with qudus pots attached directly to the rim, and wheels which powered a ‘ pot-garland’ or bucket chain. In the case of the features associated with the western settlement it seems that the curved walls represent part of the animals’ circular track, while the ashlar lined pit is likely to be associated with either the wheel or the gear mechanism.

lel linear features aligned perpendicularly to the current coastline; they are constructed from faced ashlar blocks that in some cases have degraded to stone rubble and in some areas only survive as low earthen ridges. Seven potential features were observed in this group (see Fig. 4), all of which have some connection with the current coastline and appear to respect the alignment of the possible lakeside wall. These linear features extend back from the coast, south towards the main ridge; their lengths vary between 25 and 100 m but on average only reach halfway across the plain and appear to show no direct relationship with the ridge itself. Two of these walls are connected by a linking wall at their southern ends.

Many sherds of qudus pots were observed along the northern flank of the limestone ridge ad a ent to the lain t were not recorded anywhere else on the island; the location directly adjacent to the water is also appropriate for this interpretation of water management, and places the northern coastal plain in an agricultural setting.

These walls have much in common with similar features at the eastern end of the island, they have the same orientation relative to the coastline and occupy the ground between the coast and the ridge, with their coastal terminus closely associated with sections of ashlar wall aligned along the coast. This suggests that the walls could have potentially functioned as jetties when lake waters were high or as boundary markers potentially prescribing plots of land along the coast that butt against the ashlar wall.

In the light of this interpretation it is possible to reinterpret the long linear walls between the ridge and the quay wall; these may potentially be seen as boundaries demarking areas of farmed land irrigated by saqiyas. Oleson concludes that rudimentary saqiyas from modern contexts could usefully irrigate similar portions of land to those possibly represented by the area of land contained within the linear walls (Oleson 1984: 369).

i a a sa sh a i is a s The land between the linear walls had a sparse distribution of ashlar constructed features that form the third group of structures. They are represented by sections of curved wall in two locations where additional associated wall lines were observed. These curved walls are incomplete but appear to represent rounded features with projected diameters of about 10 m (see Fig. 4).

L inear Mol e Ex tending into the L ake tow ards a Smal l I sl and At the north-eastern extent of the western settlement in the marshy ground of the shoreline plain is a site that is unique on the island see ig 2 his is signifi ant e a se it is the only ancient feature on Mareotis Island that currently extends into the lake, and as such it represents a useful indicator of lake levels in antiquity. The feature is a 250 m long linear mole construction that extends out into the lake to join with a small island that measures roughly 30 by 15 m. This small island supports a rectangular building measuring 20 x 10 m that was built from substantial masonry blocks. Three jetties extended from the north of this building into the lake measuring between 7 and 15 m in length. The feature extends towards a promontory settlement on the opposite side of the lake where there is a similarly built reciprocal linear construction that extends for a short distance into the lake on the same alignment towards the small island. These features imply a functional relationship with the lake perhaps for controlling the movement of transport along the lake as well as between the two settlements.

The interior structure of one of these rounded features is preserved and takes the form of ashlar built components lining a rectangular slot which measured 4.80 by 0.90 m. This feature, although unexcavated and partially observed, ears a signifi ant similarit to an ient water-lifting devices known as saqiya which have been observed archaeologically from broadly contemporary Greco-Roman sites in the region such as Abu Mina located some 15 km south of the island (Schiøl er 1973: 129-136; Oleson 2000: 263; see Fig. 1), and Tuna Al-Gabal near Hermopolis in the Nile Valley (Schiøl er 1973: 141-148). There are further saqiya features in the Mareotic region which have not been published in detail; examples in the immediate vicinity of the island are described as having a “ circular stone platform and oblong well” (De Cosson 1935: 130).

T ow er- l ike F eature ( F eature H ) final feat re in the western settlement is worth of mention, this is a small artifi ial mo nd of olla sed ilding material some 4 m in height which lies just northeast of the main ridge on which the settlement is located. After leaning the flat to of the mo nd, masonr elements were observed that formed a structure similar to the ashlar lined pits mentioned above. Rather than having a single pit slot this feature had a pit that was divided into two separate apertures with overall dimensions very close to the saqiya discussed above.

Waterwheels were in common use in Egypt from Hellenistic through to medieval times, and are simple devices using a gear mechanism to convert the horizontal movement of an animal walking round a circular track into vertical movement for lifting water. The water was held in pots called ‘ q s which are connected to the wheel. Saqiya are used predominantly in agricultural contexts (Oleson 1984: 126) but have also been recorded in the bath-house buildings, as at Abu Mina (see Fig. 1). Waterwheels are also known to o r in man onfig rations 41

LAKE MAREOTIS: RECONSTRUCTING THE PAST i i i a and transv erse schematic sections through h a Mareotis I sl and ( H opkinson, af ter a a

top of the tower and a further 1.5 m to the vat on top of ridge, an overall height in the order of some 5.5 m, the additional height of about 1.5 m being accounted for by the assumed superstructure which held the qudus pot-garland above the surviving tower.

It is possible that this feature is the top of a tower used to lift water from a well known to exist close to the bottom of the mound using a pot-garland passing through each aperture. An elevated ‘ bridge’ constructed from earth and masonry blocks connects the tower to the main ridge at a point where a plaster lined vat is located as part of a larger building complex that could potentially represent a public bathhouse or cistern and water management feature (Fig. 5). An example of such a tower from the late 1st century AD is known from the port of Cosa in Italy (Oleson 2000: 258). The so called ‘ spring house’ at Cosa initially lifted water to a height of 3 m but it burnt down and was replaced with a structure that lifted water over a height of 10 m to an elevated conduit. If this hypothesis is correct the structure on the mound would have lifted water to about 4 m to the

Summary The western settlement is generally poorly preserved, however there is an apparent lack of secondary activities on the northern plain which makes it easier to interpret the features that survive. The presence of water lifting devices and possible plots of agricultural land suggest that the majority of the coastal plain was above the mean waterlevel in antiquity. As such an alternative interpretation of the linear walls as possible jetties is not so well supported, 42

D. HOPKINSON: MAREOTIS ISLAND SURVEY although they could have served this function during the flood season he a sen e of f rther ildings on the lain may suggest that this community did not have a strong association with the lake for their subsistence; however the presence of the promontory towards the small island may s ggest that at some oint the settlement had a signifi ant role in administering activities on the lake.

plications for the interpretation of the other ‘ coastal’ walls on the island. To the south of the central ridge of the island, there are two adjacent ‘ plots’ of land which are enclosed by low walls of faced stone and rubble (see Fig. 2). These features resemble ‘ ka or vineyard plots; features that are usually found inland in a 15 km wide band south of the Gebel Mariout Ridge, as indicated on Surv ey of Egypt maps from 1927. K arum are features that are poorly researched but the author believes they were used to raise the depth of topsoil on agricultural plots with the advantage of increasing the space for healthy root growth and for storing rainwater in the soil reservoir (Hopkinson 2007). There seems then to have been some small scale industry and subsistence agriculture in the middle of the island that is not directly related to either of the main settlements.

Archaeology of the Central Island The land between the eastern and western settlements on the island is roughly 2 km long and appears to have been sparsely utilised in antiquity for industrial and agricultural activities. There are four sites of large scale pottery manufacture (see Fig. 2; Kiln “ A” & Kiln “ B” ), each with extensive associated storage rooms, and large mounds of cerami sherds wasted d ring the firing ro ess t seems that these otteries, like man identified near the oast on the southern mainland, were exclusively producing amphorae forms known to have been used for making and packaging the large quantities of wine for which the region is renowned (Empereur & Piccon 1986). Two of the kilns originate in the Hellenistic period (see Fig. 2; Kiln “ A” ), one close to both the eastern and western settlements. These sites continued in use through to the mid-1st century AD, when amphorae production switched to two adjacent sites (see Fig. 2; Kiln “ B” ). Production of amphorae at the new sites continued from the late 1st to the 5th centuries AD.

W ater L eve ls and Connection to the M ainland One important consideration for the interpretation of the features on the island is the relationship between the island and the mainland to the south, and the local hydrological environment throughout its occupation. The island is currently largely surrounded by water but it should not be assumed that this was the situation in antiquity because modern irrigation and d ke formation ha e modified the movement of water. A number of the features that have een identified on the island ma s ggest a greater onne tivity to the mainland to the south. All of the potential maritime focused features such as the possible jetties, quays and lake walls are located on the north facing shores of the island, and the absence of these features on the southern shores suggests the absence of water in this area that could be usefully exploited. Moreover the features that are observed along the current southern shore appear to be associated with dry land activities. The ‘ ka plots are agricultural features thought to be concerned with collecting and maximizing the use of precipitation and as such suggest that there is no alternative readily available water resource in close proximity. The second feature that appears to re l de the e isten e of a signifi ant so therl od of water in antiquity is the large kiln structure observed in the west of the island (see Fig. 4; ‘ kiln site’ ). These kilns are known to ha e signifi ant s terranean fire ham ers to 4 m dee where the fire was lit, with a erforated firing plate above to allow the hot air to circulate (El-Fakharani 1983 owe er, the firing late of this kiln is rrentl only a short height above the lake waters to the south allowing little room for the firing ham er and indi ating that it could only have been used if the land to the south between the island and the mainland had been dry. The inference that there was connectivity between the island and the southern shores of the lake, is supported by what we know about the organisation of wineries of the region, consisting of three essential elements: the kiln, the winery, and the estate villa (Empereur & Picon 1992). At least three winery structures are known in the immediate vicinity of the island on the southern mainland, and no structures s ita le for the rod tion of wine ha e een identified on the island itself. It is possible to suggest therefore that

Further evidence of light industry on the island is indicated by two small quarrying sites on the limestone ridge (see Fig. 2), although their capacity is very small compared to the quantity of stone used in the settlements and it is more likely that the majority of the building stone was transported across the lake from the large quarry sites known to exist in antiquity on the Abusir Ridge to the north (Oliver & De Cosson 1938: 169-170). In the middle of the island, associated with the kiln sites, there are a group of structures that occupy the top of the ridge that could have been workshops or accommodation for the workmen involved in these industries. A number of features from these structures extend down towards the northern coastline, although on the plain below there is only one visible structure. This is a 75 m long section of ashlars laid as stretchers aligned with the current coastline (see Fig. 2). The wall lies 70 m inland from the current oastline, whi h is s r rising sin e the fl sh fa e of the wall at first s ggests a oastal f n tion similar to the sections of lake wall or quay found associated with the eastern and western settlements. However, if this was its function then the land directly north of the wall equates to a large area of siltation and aeolian soil deposition. Small scale excavations either side of the wall in 2008, however, indicated that there was only one course of stones surviving and that the soils surrounding the feature appeared to display similar sediment regimes and were not deposited underwater. It therefore seems that this section of wall had no direct association with the lake, and its purpose is unlear his finding is also signifi ant sin e the onstr tion of a fl sh fa e a arentl res e ting the oastline has im43

LAKE MAREOTIS: RECONSTRUCTING THE PAST the kilns on the island were producing amphorae for the wineries on the southern mainland. However, the recent geomorphological enquiry undertaken as part of the Lake areotis esear h ro e t has onfirmed that the od of water that currently extends from the southern shores of the island was in fact waterlogged in antiquity (Flaux forthcoming). The island ridge was an extension of the Marea Ridge and thus enclosed a body of shallow water to the south. How continually waterlogged this area was in antiquity has yet to be determined but the geomorphological survey has indicated that there was continuous lake sedimentation between the island and the southern shores since antiquity. Thus, it is assumed that access from this area into the main body of the lake was restricted by the ridge, thus deterring settlement on the southern shores of the island, as witnessed by the lack of archaeological sites in this region. It is clear however that the situation is very complicated and that our understanding of the archaeology of the island is incomplete.

El-Fakharani, F.A., 1983, Recent Excavation at Marea in Egypt. In G. Grimm, H. Heinen & E. Winter (eds.), y ia a sia as is h y a tinische Ä gypten. A kten des international en Symsi s i i , 175-186. Mainz. El-Fakharani, F., 1984, The Kibotos of Alexandria. Studi Miscel l anei 28: 23-28. El-Falaki, Mahmoud Bey, 1872, i s a iq a i s s a s s s i s. Copenhagen. m ere r, - , 1986, n atelier de dressel 2-4 en g t au IIIe siécle’ de notre ére’ . In J.-Y . Empereur (ed.), a isa a a s a i h is iq ai s q h s a is a a a i a i y a ais h s 599-608. Athens, Paris. Empereur, J.-Y ., 1998, A l ex andria R ediscov ered. New Y ork. Empereur, J.-Y ., & Picon, M., 1986, A La Recherche des Fours d’ Amphores. In J-Y . Empereur & Y . Garlan (eds.), R echerches Sur L es A mphores G recques. A ctes q i ai a a is a ai a a h h s i i q i si s a ais h s h s 103-126. Athens, Paris. Empereur, J.-Y ., & Picon, M., 1998, Les atelier d’ amphores du Lac Mariout. In J.-Y . Empereur (ed.), a isa a a s a i h is iq ai s q h s a is a a a i a i y a ais h s 75-88. Athens, Paris. Empereur, J.-Y ., & Picon, M., 1992, La Reconnaissance des Productions des Ateliers Cérmiques: L’ Exemple de La Maréotide. ai s ahi s a a iq y i 3: 145-152. Flaux, C., forthcoming, Geomorphological survey of the south-west shores of Lake Mareotis. In L. Blue & E. Khalil (eds.), A Mul tidiscipl inary A pproach to A l ex a ia s i as h a is as y Oxford. Hopkinson, D.J., 2007, A n A ssessment of G reco- R oman L akeside A gricul ture and T rade in the Mareotic R egion of Egypt. Unpub. MA diss., University of Southampton. Khalil, E.K.H., 2005, Egypt and the R oman maritime trade: a focus on A l ex andria. Unpub. PhD thesis, University of Southampton. McCann, A.M., Bourgeois, J., Gazda, E.K., Oleson, J.P., & Will, E.L., 1987, T he R oman P ort and F ishery of sa a i a Princeton. Murray, G.W., 1951, The Egyptian Climate: An Historical Outline. T he G eographical J ournal 117.4: 422-434. Oleson, J.P., 1984, G reek and R oman Mechanical W ateri i i s h is y a h y London. Oleson, J.P., 2000, Water-Lifting. In Ö . Wikander (ed.), a k a i a h y, 217-302. Leiden

Conclusion The archaeology of Mareotis Island consists of two settlements which appear to be concerned with agricultural, industrial and water management activities as well as lastrine fo sed feat res whi h ma indi ate that fishing activities or lake transportation services were conducted from the island. The island represents a largely unexplored archaeological asset that has the potential to give insights into daily life in a non elite community; when looked at in conjunction with other known sites in the area it may provide an opportunity to begin to address settlement distributions and social organisation and make comparison between the lake region and those communities on the elta, and erha s ar haeologi al findings from the a um or Nile Valley, where the most studied examples of agricultural settlements are known. Such comparisons with sites from the Egyptian interior may prove to be less than representative of ordinary Egyptian practices where the majority of the population was located on the Nile Delta (Bagnall 2007: 227). There is a great potential for future research on Mareotis Island and the broader Mareotis region and these ro isional findings onl egin to s rat h the surface of what the island and Mareotic region in general has to offer. Bibliography Bagnall, R.S., 2001, Archaeological Work on Hellenistic and Roman Egypt, 1995-2000. A merican J ournal of A rchaeol ogy 105. 2: 227-243. Ball, J., 1942, y i h assi a a h s. Cairo. Brooks, C.E.P., 1949, i a h h h a s New Y ork. De Cosson, A., 1935, Mareotis: B eing a Short A ccount of the H istory and A ncient Monuments of the N orths s y a ak a is. London. El-Fakharani, F., 1974, The “ Lighthouse” of Abusir in Egypt. a a i si assi a hi y 78: 257272. 44

D. HOPKINSON: MAREOTIS ISLAND SURVEY a a i a i y a ais h s 27-36. Athens, Paris. Rodziewicz, M.D., 2002, Mareotic Harbours. In C. Déco bert (ed.), A l ex andrie Mé dié v al e 2 . Etudes A l ex andrines 8: 1-22. Schiøl er, T., 1973, R oman and I sl amic W ater- l ifting W heel s. Denmark. Strabo, G eography, H.L. Jones (transl.), 1949. London. Surv ey of Egypt (27/135) 1:100,000 Map; Sheet 88/48 “ El Ghayata” . Survey of Egypt/Ministry of Finances, Cairo. Tomber, R., & Thomas, R.I., forthcoming, Pottery from the Lake Mareotis Research Project. In L. Blue & E. Khalil (eds.), A Mul tidiscipl inary A pproach to A l ex a ia s i as h a is as y Oxford. Warne, A.G., & Stanley, D.J., 1993, Late Q uaternary Evolution of the Northwest Nile Delta and Adjacent Coast in the Alexandria Region, Egypt. a as a R esearch 9.1: 26-64.

Oliver, F.W., & De Cosson, A., 1938, Note on the Taenia Ridge – with especial reference to quarries, sites and an ancient road between Alexandria and Abu Sir. i a i ya h i andrie 32: 163-175. Rodziewicz, M.D., 1983, Alexandria and District of Mareotis. a a i a 2: 199-216. Rodziewicz, M.D., 1990, Taenia and Mareotis: Archaeological Research West of Alexandria. A nnual of the Egyptian Society of G reek and R oman Studies 1: 62-78. Rodziewicz, M.D., 1998a, From Alexandria to the West by Land and by Waterways. In J.-Y . Empereur (ed.), a isa a a s a i h is i q ai s q h s a is a a a i a i y a ais h s 93-103. Athens, Paris. od iewi , , 1998 , lassifi ation of ineries from Mareotis. In J.-Y . Empereur (ed.), a isa a a s a i h is iq ai s q h s a is a

45

LAKE MAREOTIS: RECONSTRUCTING THE PAST

46

M. HAGGAG: EXCAVATIONS AT MAREA/PHILOXENITÉ

he it of area/ hilo enit efle tions on the Alexandria University Excavations, 1977-1981 Mona Haggag

In 1977 Alexandria University patronized an expedition under the direction of the late Prof. Fawzi El-Fakharani to search for the ancient city of Marea, mentioned by different classical authors as the capital of the Mareotic Nome during the late Pharaonic period (Thucydides I.68; Herodotus II.30; III.12; 15; IV.161; Diodorus I.68.).

adek on l ded that all the remains identified were not earlier than the 4th-5th centuries AD (Sadek 1978: 67-80). All trial trenches proved to have remains of stone buildings dating to the Byzantine period. Pot sherds were entirely of Byzantine manufacture (Shahin 1983). The same is true for the architectural material, construction methods and te hni es hese res lts raised a ma or estion a o t the location of the pre-Byzantine town mentioned in classical literature (Thucydides I.68; Herodotus II.30; III.12; 15; IV. 161; Diodorus I.68). . T he Pre- Byz antine Remains (Map 1) By exploring the area some 5 km west of the Byzantine remains, earlier remains of Marea proved to exist beyond the main perimeter of the Byzantine site. A group of harbour installations were found on an island to the west.1

Mahmoud Bey El-Falaki located the site of Marea some 45 km west of Alexandria on the southern shores of the western extremity of Lake Mareotis near El-Hawaria (ElFalaki 1872: 85, 86, 91, 96, 97, 101; see other Haggag ha ter in this ol me l- alaki ased his identifi ation of the site according to its location on the map of Ptolemy the eogra her ,5 16 17 his lo ation is onfirmed by Ball (1942: 25-30), Breccia (1922: 337), De Cosson (1935: 131-135), Rowe (1954: 128-145) and Fraser (1972: 143-146) due to four visible jetties which still project into the lake today. The jetties are associated with a broad ridge of hard limestone extending from east to west across the town parallel to the southern coast of the lake and bordered, from both north and south, by calcarenite land.

T he ‘ Pr’ - S haped H arbour (Fig. 1, Map 1) The main harbour of the island is a three sided structure projecting into the lake basin about 30.5 m wide and 57 m long. It is built of undressed blocks of limestone of Cyclopean size. A number of the upper blocks in this structure have drilled cylindrical holes which were either bollard holes or mooring rings (Fig. 2). According to the traditions of ancient Egyptian harbour architecture, ports of rectangular plan in the shape of the sign “ pr” were designed for harbours facing the open seas. Earlier examples existed in the seaport of Alexandria (Jondet 1916: 13-20; El-Fakharani 1991: 25), the most conspicuous of which is the one inside the Western Harbour mentioned by Strabo as the Kibotos (G eography, Xvii, I, 10), another one is associated with the Island of Antirrhodes which lies inside the Eastern Harbour of Alexandria discovered by Goddio (Goddio, et al. 1998: 21-29, pl. I). This feature consists of two rows

T he S urve y As part of the 1977 campaign and in collaboration with the University of Guelph, Canada, Mohamed Sadek conducted a proto-magnetometer survey at the site, concentrating on the shoreline area from which the four jetties projected. Data was collected in lines parallel to the lakeshore from west to east in a grid of 800 m long and 100 m wide at intervals of 4 m. Soundings were taken in the area west of the first a , and some trial tren hes were in estigated in three locations on the waterfront near the jetties. The trenches went deep into the bedrock which was found at intervals of 4.6-6.00 m below the present ground surface. F ig. 1: K ibotos harbour structure on the south shore of L ake Mareotis ( Site 9 – L ake Mareotis R esearch P roj ect) ( photo M. H aggag) .

1. It is worthy of notice here that Gauthier, referring to some hieroglyphic text, stated that the ancient city of Marea was located on an island in the lake (Gauthier 1925-1931: vol. III, 53-54).

47

LAKE MAREOTIS: RECONSTRUCTING THE PAST Map 1: Marea in L ake Mareotis ( V . A tef) .

F ig. 2: K ibotos harbour structure, m ooring ring ( photo M. H aggag) .

F ig. 3: R ock- cut K ibotos harbour structure at I kingi Mariut ( photo M. H aggag) .

48

M. HAGGAG: EXCAVATIONS AT MAREA/PHILOXENITÉ of stakes or piles approximately parallel, running southwest to north-east, along the axis of the main branch of the island, that forms a rectangular shaped-harbour against the island’ s shoreline. A third one, hewn in the rock at Ikingi Mariut, about 20 km west of Alexandria (Fig. 3), was identified l- akharani2 (2002: 203-208).

to a dromos. Another similar dromos on the opposite side leads to a funerary chamber. The ceiling is carved to imitate wooden beams in the manner common to Pharaonic cemeteries. False doors were also carved in the walls with a shaft in front of each one leading to the burial chamber below. All tombs have been looted. The only remains were a few local pointed bottom jars which are likely to be Late Pharaonic in date (El-Fakharani 1983: 176-178).

T he F ort (Map 1) To the south-west of the pr-shaped port remains of a substantial structure built on a monolithic platform about 21 m wide and 24 m long was uncovered. The structure has an L-shaped plan. The walls seem to have consisted of rough core masonry with outer casing walls. The blocks of the o ter asing are aref ll t and fitted together in o li e oints witho t sing an mortar e e t for a foliage layer of non reddish mortar used in some of the courses to facilitate the sliding of such gigantic sized blocks. These external walls are slanting inward at an angle of about 40 degrees, and are based on larger foundation blocks with s ared edges he ilding is a roa hed from the so th means of a ram of dressed and aref ll fitted stone blocks (Haggag 1984: 277-280). The use of rough masonry for the core while the huge blocks of the facing are very aref ll olished, the o li e oints of the lo ks, the use of such a thin layer of mortar, the inward slanting of the outer walls as well as the general plan of the building, are clear indications to a Late Pharaonic date. The building is interpreted as the fort of Psammaticus III (525 BC), from which his son Inaros sailed with his army, aided by some Athenian troops (Thucydides I.104), against the Persians in 465 BC, in an unsuccessful attempt to liberate Egypt.

T he Byz antine H arbours (Fig. 4, Maps 1 & 2) The focal point of the town to the east of the Pharaonic/ Ptolemaic remains is the Byzantine harbour area which extends 2 km along the lake shore with four well preserved jetties projecting into the lake at an average height of nearly 2 m above the lakebed. The area comprises three harbours marked by the jetties and a natural promontory namely the west port, the middle and the east port. Along the shore there is a stone built seawall occasionally interrupted by either steps or stone ramps leading directly down to the water. The wall extends some 500 m with an average width of 1 m. t the o ter e tremit of the western a of the middle port, there are a number of circular foundations about 4 m in diameter. These features are assumed to have been the base of a light beacon or lighthouse that would have aided navigation at night. Some mooring features can be detected on the upper course of this jetty in the shape of cylindrical bollard holes (Petruso & Gabel 1983: 63). T he E ast Port (Fig. 5, Map 2) Although the east port was completely buried under the sands of the lake which by the time of excavation had dried out, digging revealed that it differs from the other ports of Marea in both shape and function. It is located between the promontory and a north-easterly island which has its own landing la e formed a short artifi ial d ke to the north-east and a natural promontory to the northwest. This northern island is connected to another smaller one to its so th-west third a onne ts this island to the promontory forming a wide entrance to the eastern ort, at least in its first stage of onstr tion n the se ond stage, the shape, size and function of the port seem to have een modified lmost 100 m of the a that e tended from the eastern small island to the promontory were remo ed and an e tension to this a onstr ted to the south. It was located parallel and adjacent to the eastern shore of the promontory and was about 200 m in length. Thirty metres before reaching a third island to the south, the a ends forming an o ening etween its so thern extremity and this small island. This opening provides an exit or entrance to the east port by way of the southern body of the lake, which extended considerably inland at this oint this modifi ation, the er long a s of the east port became totally isolated from both the mainland and the promontory. Accordingly, the link between the anchored ships with their cargoes and the town was severed. This fact is very important as far as the function of the port is concerned.

T he Ptolemaic Q uay (Map 1) To the east of the island, a stone built jetty about 104 m long extends into the lake. Near its southern end, the jetty inclines to the east at an angle of about 175 degrees to extend a further 20 m in length. It is constructed of local regularly sized rectangular blocks, arranged in alternate courses of headers and stretchers (Haggag 1984: 280-283). he fine workmanshi , medi m si e nat re of the lo ks as well as the ashlar masonry construction style with its vertical joints, are indications of a Ptolemaic date for the a T he Cemetery (Map 1) To the southwest of the island on the limestone ridge, the team came upon a group of burials that have nothing in common with Byzantine tomb types. The cemetery consists of some anthro oid it tom s s are shaft tom is carved into the limestone ridge to a depth of 6 m. Loculi with gabled ceiling are cut either side of the shaft. A nearby chamber tomb is composed of a sloping passage hewn into the rock that ends with six steps leading to an open court. An opening cut in one side of the court leads

2. For detailed discussion of this pr-shaped ports of Pharaonic date as well as e am les, see aggag 1984 263 ff l- akharani 1991 25, fig 1,2.

49

LAKE MAREOTIS: RECONSTRUCTING THE PAST Map 2: Marea’ s publ ic buil dings on the l ake shore ( V . A tef) .

F ig. 4: B yz antine harbour of Marea ( photo M. H aggag) .

de th of 0 25 m elow the a ement, a s are drainage canal, coated with a thick layer of plaster and covered by a row of huge rectangular blocks, extends across the decumanus. This drains into a similar canal that extends under an unpaved side street and from south to north crossing the decumanus at a right angle. Thus, the drainage waters pour directly into the lake (El-Fakharani 1983: 178-179).

El-Fakharani’ s interpretation of the function of the east ort 1983 181 is that it worked as a one wa traffi harbour for transit navigation: so that seafaring ships carrying goods from different parts of the Mediterranean could enter the east port via its northern opening, their cargoes were to e nloaded on the long a s in order to e reloaded on board river boats to transport goods to the Nile and vice versa.

T he S hops (Map 2) The arcade of the decumanus lines a group of 12 constructions (Fig. 7), with different ground plans. The buildings are located side by side with doors that opened out onto the sheltered arcade. They seem to have been built to function as sho s with residential arters for their owners Remains of white plaster coating are still visible on some of the walls. An upper storey for each of the shops is confirmed the e isten e of stairs at the rear of some of the shops and also in some of the back rooms (El-Fakharani 1983: 178-179; Soleiman 2004: 145-164).

T he Public Buildings (Map 2) Many, if not most, of the public buildings of Byzantine Marea were mainly erected along the shoreline. Parallel to the lake shore in front of the harbour installations runs the main arcaded street of the city. This decumanus (Fig. 6) extends eastward until it reaches the natural promontory which forms the eastern part of the middle port. It is 10 m wide marked on its southern side by an arcade of which some of the bases of columns can still be seen in situ. The street is paved by big rectangular limestone blocks. At a 50

M. HAGGAG: EXCAVATIONS AT MAREA/PHILOXENITÉ T he Basilica (Fig. 8, Map 2) To the west of the decumanus, at the intersection of the western side street, a huge complex of structures came to light during excavations (Sadek 1992: 549-554). The structures consist of two conjoined apsidal basilicas furnished with what seems to be public bathing facilities for sanitary purposes, similar to those discovered and interpreted by Grossman at St Menas (Grossmann 1986: 12-13; Haggag 1984: 284-289). T he Bak ery At the promontory which forms the eastern extremity of the middle port, a huge building with an area of about 700 m 2 was uncovered. A large granite mill inside one of the building’ s multiple rooms indicates the relation of this building with grain grinding processes. Inside one of the rooms, a marble tile with a carved Greek cross was uncovered. In a compartment annexed to one of the rooms in this building, a clay jar containing a hoard of coins was also found. Courses of Byzantine burnt bricks are inserted in the stone masonry of the walls. The building is surrounded by an external wall supported with buttresses typical of Byzantine methods of construction (Moussa 2002: 479480). F ig. 5: ( abov e) P l an of the eastern harbour of Marea, after El - F akharani 197 ( thanks to El - F akharani) .

F ig. 6: Decumanus betw een the eastern and middl e ports of Marea ( photo M. H aggag) .

F ig. 7: P l an of shops off the main decumanus ( thanks to El - F akharani) . 51

LAKE MAREOTIS: RECONSTRUCTING THE PAST F ig. 8: B asil ica, M area ( photo M. H aggag) .

F ig. 9: Enigma B ul iding, Marea ( photo M. H aggag) .

T he E nigma Building (Fig. 9, Map 2) Next to the western edge of the promontory at its meeting point with the lakeshore along the middle port, an enigmatic building has been uncovered. The building stands completely in the water. It is composed of podia and walls leaving canals 0.75 m broad between them. Some of the canals run in parallel lines with the coast, while others run at right angles with it and the other canals. The entire building is sliding towards the water leaving the outer parts of the podia under water. The canals were studded with small artefa ts he finds, whi h were mostl inta t, included nearly one thousand pieces of small bronze coins, about 200 jugs of red or yellowish clay, St Menas ampulae, other ampulae with crosses, small statuettes of horsemen, fishing hooks and roken glass ottles his n s al lan made scholars differ in their interpretations of the building’ s function. Some suggestions included its possible function as a dry-dock owing to its distinct slope into the lake water adek 1978 70 , t the h ge amo nt of finds inside the canals stands against this hypothesis. Another e lanation is that it ser ed as a fisher ikkos ianos pers. comm.). While the discoverer tended to interpret it as a building for votive offerings in honour of both St Menas

and t ar s, d e to the large antit of oins, am lae and other artefacts that were found inside (El-Fakharani 1977: 16; Soleiman 2004: 121-138). Boston F ield S chool In 1979, the Boston University’ s Summer T erm and Study A broad P rogram s orted an ar haeologi al field s hool at the site of Marea. Their work lasted for three consecutive seasons, one month each year.3 In addition to the fruitful student training, the mission made new and important discoveries, the most important of which was the Diolkos of Marea. Along the shore between the middle and western jetties, the team of Boston University came upon remnants of a dockyard (Petruso & Gabel 1982: 12). It consists of two sloping walls, each is over 7 m wide and 20 m long,

3. The team was directed jointly by Profs. El-Fakharani & Gable in collaboration with Petruso and the architect, Boyd. The author of this arti le had the hono r of arti i ating in this field s hool first as an undergraduate trainée and then as an assistant archaeologist.

52

M. HAGGAG: EXCAVATIONS AT MAREA/PHILOXENITÉ F ig. 10: W inery ex cav ated by El - F akharani 1983 ( photo M. H aggag) .

pletion of the excavation of the building, it proved to be a huge double peristyle that occupied an area of more than 1,500 m2. Both El-Fakharani (1983: 184-186) and Rodziewicz (1988a: 175-178, 1988b: 267-276) asserted that this dis o er re resents the first and largest erist le built for residential purposes to be discovered in Egypt hitherto.

emerging from massive stone foundations going down to the bedrock, to a depth of about 2 m. The runners slip into a central aisle that lies between the jetties. This aisle has some sort of drainage facilities in the form of holes cut into the floor in order to drain water and kee the floor dry. The entire building slides down into the lake basin at a gradient of approximately 1:16 which allows for the use of manpower and log rollers to hall ships out of the water (Petruso & Gabel 1983: 76).

In the middle of the town, some Byzantine dwellings were uncovered. Each is composed of two rooms. Byzantine local pottery ware were scattered inside. Muslim burials were dug inside the room. Green glazed pottery sherds of the Islamic period were found as well as one piece of a Fatimid coin.

T he City’ s Residential Area Far from the harbour area to the south of the limestone ridge and south of the modern highway leading to Borg ElArab, in the area now called “ Hawariya” . During a second season conducted by the Alexandria University expedition team in 1977, various buildings were uncovered in this region which seem to relate to residential and daily life activities. T he Peristyle Complex (Map 1) Adjacent to the modern highway (Map 1), some 2.5 km to the south of the harbour, a peristyle court which leads to a variety of rooms, was uncovered. One of the rooms has a stair ase leading to an er floor of the ilding. It seems that the building went through two stages in its onstr tion the first stage is indi ated a la er of white plaster coating on the faç ade. The second stage is marked by the addition of a massive thick walled faç ade with two windows which have tunnels underneath leading to watercourses that are connected to cisterns. A modern ditch for irrigation water that was dug through the rooms halted a complete investigation of the remains. In 1980, Rodziewicz4 replaced that ditch with a pipe in order to continue uncovering the rest of the building. On com-

W ineries (Map 1) About 200 m to the southwest of the peristyle building, in the middle of the town, a big elaborate winery (Fig. 10) was uncovered (El-Fakharani 1983: 182-184). Another idential in la o t t smaller than the first, was also dis o ered 2 km to the southwest (Arafa 1985: 78-80). Both are of the type known in other places in the Mareotic region.5 They represent the archaeological testimony for the excellence of the Mareotic wine praised by the classical authors (see Dzierzbicka this volume). Each of the two buildings comprises two rooms separated by a low screen wall for s ee ing gra es ne is smaller than the other oth are entirely coated by four layers of red plaster to prevent any possible seepage of the juice. In the center of the smaller room is a raised round base covered with plaster possibly to s ort a mo a le s ee er or a hand ress means of a lion headed spout, the juice from both rooms pours into a large s are asin whi h is d g in the gro nd, with another smaller and dee er asin d g in its floor for olle t-

4 od iewi was then a ons ltant for the g tian nti (EAO) and was in charge of the restoration of the area.

5. A third smaller one that lies to the south of the big winery has been reported by Rodziewicz, see Rodziewicz 1998a: 27-36.

it

er i e

53

LAKE MAREOTIS: RECONSTRUCTING THE PAST ing residues. The inner surface of the big basin’ s walls is s rat hed to allow the laster oating to hold firm he inner sides of the basin are surrounded by a shelf approached from all sides a flight of ste s ther ste s lead to the asin s floor t is tho ght that this shelf is designed to s ort wooden eams that held a ie e of loth for filtering the juice poured from the lion shaped spout. In the middle of the north side above the spout, there is a small podium with two funnels pierced each by a hole opening at the basin. El-Fakharani’ s hypothesis is that these funnels acted as measures for adding certain amounts of some aromatic fla o rs that rod ed the famo s taste of area s wine6 (1983: 182-184).

town all 1942 114, 117, fig 17, l , ore than two and a half centuries after Ptolemy, we hear nothing about Marea. From the 5th century onwards, Byzantine Marea appears. During this time, the city has to play another vital role in Egypt’ s history. The town in which the Martyr Abu Mina was buried lies about 15 km to the south of Marea. The sanctuary of St Menas was famous for its miraculous healing capacities. The importance of the area increased gradually until it reached its climax at about the end of the 5th and beginning of the 6th centuries (Kaufmann 1910: 103). Flux of pilgrims and patients from all over the Roman provinces headed to the place seeking cures for their ailments rossman 1986 12-13 lasks am lae filled with the holy waters of Abu Mina were specially made for pilgrims to take with them on their way back home as a blessing.8

T he Byz antine T omb (Map 1) To the north of the railway line, and at the southern boundaries of the inhabited area of Byzantine Marea, lies a rock cut chamber tomb (El-Fakharani 1977: 19-22). The entrance to the tomb is an arched doorway with basalt jams and lintel. It opens onto a passage with a staircase leading to a s are mo rning ham er he eiling of the assage is barrel vaulted with a helicon vault covering the part which turns to the east towards the arched entrance of the mourning chamber. The latter is surrounded by a stone bench and covered with a cross vault with pendentives. Three burial chambers with cross vaulted ceilings are cut into the walls of the tomb, forming a trefoil plan for the H ypogeum.7

The most suitable and more convenient way for those who are coming from both the Mediterranean and the Nile is to reach the lake harbour of Alexandria and navigate the lake to the nearest harbour before taking the land route to St Menas. In this respect, Marea is the nearest point. At that time, Marea’ s earlier harbours were isolated on an island and pilgrims had to be ferried to the southern coast of the lake in order to take the caravan route to the sanctuary. To solve this problem, the old town of Marea was shifted 2.5 km to the east of the earlier harbour area, where the line of the ridge inclines far from the lake shore leaving a larger area including a promontory, a bay and several islands in front of the shoreline, that combined offer a naturally sheltered har o r his area seemed more s ita le to i kl establish the new harbour facilities and provide accommodation for pilgrims (Rodziewicz 1998b: 101-103). Other factors may have contributed to the shifting of the site, the most important of which is the changing lake levels, a fact indicated by the different levels of rising water during the flood seasons o er man ent ries

T he N ame of the City More than two centuries before the establishment of Alexandria, Marea, according to Herodotus (II.30, IV-161; III.12, 15), Thucydides (I.104) and Diodorus (I.68) functioned as a strategic staging post on Egypt’ s northern frontiers. After the establishment of Alexandria, the former military role of Marea appears to have come to an end and the city takes on a more civic and commercial role (Rodziewicz 1998b: 101-103). Excavation works yielded neither wea ons nor fortifi ations of neither tolemai or Roman date. At the same time, classical authors, such as Athenaeus (I.33), Vergil (G eorgics, II.91-92), Horace (O des, I.37) and Columella (D e R ustica, III.24), give us glimpses of the reputation of Mareotic wines. Marea became an agricultural centre as well as an intermediate station for goods moving between Alexandria and the Nile Valley (Rodziewicz 1983: 199-208). After the time of Strabo (XVII.I.14) who spoke of the city’ s wealth in papyrus, bean and vine plantation, the city seems to have shrunk to a small village as stated by Athenaeus (I.33) and as could be inferred from Ptolemy the Geographer who mentioned it as “ Palae-Marea” by which name he indicates the existence of the earlier Pharaonic

his leads s to the estion that od iewi 2003 27-39 first raised as this new site of antine area the it of Philoxenité? 9 Philoxenité was built by the Praetorian Praefect Philoxenus upon orders of the Emperor Anastasius (AD 491-518) in order to serve the needs of pilgrims on their way to the shrine of St Menas. Rodziewicz counted on the Coptic Encomium in raise of t enas, first lished by Drescher (1946: 35-72; 126-149), to reach the conclusion of interpreting the later site of Byzantine Marea as the city of Philoxenité. According to the text of the Encomium, Drescher placed Philoxenité on the shores of the main basin of the lake, somewhere close to Kom Truga 35 km to the east of the Holy Shrine. Rodziewicz believes

6. For more detailed interpretations, see Rodziewicz 1998a: 27-36. 7. Various rooms on top of the H ypogeum and its vicinity were uncovered a team of the reme o n il of nti ities later in 1987/88. They represent a complex of funerary chapels for the tomb (Soleiman 2004: 283-286).

8. Just as Muslims nowadays do with the waters of the holy well of Z amzam at Mecca. 9. Before tackling this problem, it is worth noting that Rodziewicz in addition to his responsibility for the restoration works carried on at the site of Marea, made additionally impressive discoveries at the site.

54

M. HAGGAG: EXCAVATIONS AT MAREA/PHILOXENITÉ Bibliography

that the site of the Byzantine harbours at Marea accurately fit the lo ation of the landing la e of ilgrims seeking a commodation, food and beverage, before taking the short caravan route to the shrine.10 In this respect, it is important to point out that the Encomium, which is the sole source we know of to date about the city of Philoxenité, includes the following passage:

A ncient Sources Athenaeus, D eipnosophists, C. Burton (transl.), 1927. London. Augusti Meinekii, 1849, Stephani B yantii, Ehnicorum quae supersunt, I. Berlin. Columella, O n A gricul ture, H. Boyed (transl.), 1931. London. Diodorus Siculus, L ibrary of H istory, C.B. Welles (transl.), 1963. London. Herodotus, H istories, H. Carry (transl.), 1912. London. Horace, T he O des, C.H. Bennet (transl.), 1925. London. Ptolemy, G eographica, E.L. Stevenson (transl.), 1932. New Y ork. Strabo, G eography, H.L. Jones (transl.), 1949. London. Thucydides, H istory of the P el oponnesian W ar, C.F. Smith (transl.), 1969. London. Virgil, G eorgics, J. Jackson (transl.), 1908. Oxford.

hen it was finished he ga e it the name hilo enité… He ordered to build water stations, where he placed water jars, every ten miles along the route between the hospices and the shrine in order to serve the needs of these multitudes” (Drescher 1949: 15-16).

Y et, the distance between the Byzantine harbours of Marea to the sanctuary do not exceed 15 km, that is less than 10 miles, a distance that can not take more than a few hours if riding on camels and half a day if walking on foot. Then, why did the Praefect order the building of water stations every ten miles along this route? 11 Rodziewicz’ hypothesis is mainly based upon believing that there is nothing evident at the site to be dated earlier than the Byzantine era. This assumption has been proved to be incorrect through what was discovered on the island and the ridge by the Alexandria University expedition. More recently our Polish and French colleagues have also uncovered some pre-Byzantine remains (Szymanska & Babraj 2001: 37-42, 2002 47-58 see a ra ma ska this ol me i hot 2004). Moreover, if the newly established Byzantine centre had the name of Philoxenité, why did a Christian author, who lived and wrote after the time of Anastasius and his Prefect, Stephanus of Byzantium (Augusti Meinekii 1849: 432), retain the old name of Marea in his writings? 12

Secondary Sources Arafa, D., 1985, W ine P roduction in Egypt from A l ex ander to the A rab Conquest: A s Compared to the G raecoR oman W orl d. Unpub. MA diss., Alexandria University. Ball, J., 1942, Egypt in Cl assical G eographers. Cairo. Botti, G., 1902, Studi sul III Nomo dell’ egitto inferiore E piu specialmente sulla regione Mareotica. B ul l etin de l a Socié te A rché ol ogique d’ A l ex andrie 4: 41-84. Breccia, E., 1922, A l ex andrea ad A egyptum. Bergamo. De Cosson, A., 1935, Mareotis: B eing a Short A ccount of the H istory and A ncient Monuments of the N orthW estern D esert of Egypt and of L ake Mareotis. London. Drescher, J., 1946, A pa Mena: A Sel ection of Coptic T ex ts R el ating to Saint Menas. Cairo. Drescher, J., 1949, Topographical Notes for Alexandria and Distric. B ul l etin de l a Socié te A rché ol ogique d’ A l ex andrie 38: 13-20. El-Fakharani, F.A., 1977, A P rel iminary R eport on the F irst Season of Ex cav ation at Marea. Excavation Reort resented to the nti ities e artment, g t Cairo. El-Fakharani, F.A., 1983, Recent Excavation at Marea in Egypt. In G. Grimm, H. Heinen & E. Winter (eds.), A egyptiaca T rev ernsia I I , D as R öm isch- B yz antinische Ä gypten. A kten des international en Symposions 26.30. Se ptember 1 978 i n T rier, 175-186. Mainz. El-Fakharani, F.A., 1991, The Kibotus of Alexandria. Studi Miscel l anei 28: 21-28. El-Fakharani, F.A., 2002, The Pharaonic Port on the Mediterranean: Its Shape, Development and Importance. In Z .A. Hawass & L.P. Brock (eds.), Egyptol ogy at h a h y s y i s of the 8t h I nternational Congress of Egyptol ogists, Vol. 2: 203-208. Cairo. El-Falaki, Mahmoud Bey, 1872, Memoire sur l ’ antique A l ex andrie, s es faubourgs, s es env irons. Copenhagen. Empereur, J-Y ., 1998, A l ex andria R ediscov ered. New Y ork.

Finally, one would say that it is not unusual for a certain town in anti it to ha e more than one name o er the different periods of its history. Rhacotis/Alexandria, Thonis/ Heracleion, are very distinctive examples in this respect. It seems reasonable that the old town of Marea shifted eastward to build new facilities suitable for the new role the city had to play.13 Whether the new site took a new name or not is a estion that annot e answered et with an certainty. In this respect, further readings of the original Encomium are re ired as well as an other material e idence that may appear through investigations currently being undertaken by different missions at the site.

10. Haas and Empereur share Rodziewicz’ s opinion, see Haas 1997: 349; Empereur 1998: 229-239. 11. Whether this is due to miscalculations of the wording of the text, as Rodziewicz mentions, it is something that cannot be judged unless further readings of the original Encomium are undertaken, see Rodziewicz 2003: 27-39. 12. It is not evident that the various citations of Marea in the ancient literature refer to exactly the same spot as in the case of Procopius: a o s a r 1913 ol 2, 171 e find some onf sion in the use of “ Marea” and “ Mareotis” used to describe both the town and the lake or sometimes the whole lake district (Petruso & Gabel 1980: 1-27). 13. In 1902 Botti stated that there were two towns named Marea, the ancient and the new (Botti 1902: 73-75).

55

LAKE MAREOTIS: RECONSTRUCTING THE PAST Fraser, P.M., 1972, P tol emaic A l ex andria, Vol. I. Oxford. Gauthier, H., 1925-1931, D ictionnaire des nomes gé ographiques contenus dans l es tex tes H ié rogl yphiques, Vol. III. Le Caire. Goddio, F., Bernard, A., Bernard, E., Darwish, I, Kiss, Z ., & Y oyotte, J., 1998, A l ex andria: T he Submerged R oyal Q uarters. Oxford. Grossmann, P., 1986, A bu Mina: A G uide to the A ncient P il grimage Center. Cairo. Haas, C., 1997, A l ex andria in L ate A ntiquity: T opography a ia i . Baltimore. Haggag, M., 1984, P orts in A ncient Egypt til l the A rab Conquest. Unpub. MA diss., Alexandria University. Haggag, M., 2002, Two Religious Buildings at Byzantine Marea. In Z .A. Hawass & L.P. Brock (eds.), Egypya h a h y s y ceedings of the 8t h I nternational Congress of Egyptol ogists, Vol. 2: 284-289. Cairo. Jondet, M.G., 1916, L es P orts Submergé s de l ’ ancienne I l e de P haros. Le Caire. Kaufmann, C.M., 1910, D ie Menasstadt und der national heil igtum der al tchristl ichen Ä gypter, I. Leipzig. o ssa, , 2002, el es as e ts de la ie otidiè nne représentés á Marea Byzantine. In Z .A. Hawass & L.P. Brock (eds.), Egyptol ogy at the D aw n h y s y i s h 8t h I nternational Congress of Egyptol ogists, Vol. 2: 478-486. Cairo. Petruso, K.M., & Gabel, C., 1980, Marea: An Environmental and Cultural Study at Lake Maryut, Lower Egypt: A Research Prospectus. B oston U niv ersity A frican Studies Center W orking P apers 25: 1-27. Petruso, K.M., & Gabel, C., 1982, Marea: A Byzantine Port in Northern Egypt. B oston U niv ersity A frican Studies Centre, W orking P apers 62: 1-23. Petruso, K.M., & Gabel, C., 1983, Marea: A Byzantine Port on Egypt’ s Northern Frontiers. A rchaeol ogy (Sept./Oct.): 62-77. Pichot, V., 2004, L a fouil l e de l ’ I l e de Marea: L e site, P ros pections et campagnes de fouil l e: http://www.cealex. org/sitecealex/activities/terrestre/marea/genef.htm, CEA l ex (accessed March 2008). Rodziewicz, M.D., 1983, Alexandria and District of Mareotis. G raeco- A rabica 2: 199-208.

Rodziewicz, M.D., 1988a, Remarks to the Peristyle House in Alexandria and Mareotis. P raktika, I nternational Congress of Cl assical A rchaeol ogy in A thens, September 4t h - 10t h, 1983 : 175-178. Athens. Rodziewicz, M.D., 1988b, Remarks on the Domestic and Monastic Architecture in Alexandria and Surroundings. In E.C.M. vanden Brink (ed.), T he A rchaeol ogy of the N il e D el ta: 267-276. Amsterdam. od iewi , , 1998a, lassifi ation of ineries from Mareotis. In J.-Y . Empereur (ed.), Commerce et artisanat dans l ’ A l ex andrie hel l é nistique et romaine. A ctes du col l oque d’ A thè nes, organisé par l e CN R S, l e L aboratoire de cé ramol ogie de L yon et l ’ Ecol e franç aise d’ A thè nes,1 1- 12 dé cembre 198: 27-36. Athens, Paris. Rodziewicz, M.D., 1998b, From Alexandria to the West by Land and by Waterways. In J.-Y . Empereur (ed.), Commé rce et artisant dans l ’ A l ex andrie hé l l é nistique et romaine. A ctes du col l oque d’ A thè nes, organisé par l e CN R S, l e L aboratoire de cé ramol ogie de L yon et l ’ Ecol e franç aise d’ A thè nes,1 1- 12 dé cembre 198 : 95-102. Athens, Paris. Rodziewicz, M.D., 2003, Philoxenité - Pilgrimage Harbour of Abu Mina. B ul l etin de l a Socié te A rché ol ogique d’ A l ex andrie 47: 27-47 Rowe, A., 1954, A Contribution to the Archaeology of the Western Desert: II. B ul l etin of the J ohn R yl ands L ibrary 36: 128-145. Sadek, M., 1978, The Ancient Port of Marea. Cahiers des É tudes A nciennes VIII: 67-80. Sadek, M., 1992, The Baths at the Ancient Harbour of Marea. T he Sesto Congresso I nternaz ional e di Egittol ogia, Vol. I: 549-554. Turin. Shahin, B., 1983, L ocal P ottery in B yz antine Egypt: A y h s a h iy a a. Unpub. MA diss., Alexandria University. Soleiman, N.M.S., 2004, Marea: A n A rchaeol ogical Study and the Manner of its T ourist I nv estment. Unpub. PhD diss., Alexandria University. Szymanska, H., & Babraj, K., 2001, Marea: First Interim Report, 2000. P ol ish A rchaeol ogy in the Mediterranean XII: 37-42. Szymanska, H., & Babraj, K., 2002, Marea: Second Interim Report, 2001. P ol ish A rchaeol ogy in the Mediterranean XIII: 47-58.

56

V. PICHOT: MAREA PENINSULA

area enins la ation and orksho ti ities on the hores of ake ario t in the ork of the enter d t des le andrines le , 3134 al rie i hot G eneral Introduction to the S ite of M area he site, identified as area on the late of le andria in the A tl as gé ographique d’ Egypte1 (D escription de l ’ Egypt 1809 l 10 and then ahmo d e l- alaki in 1866 l- alaki 1872 96 , is one of the rare e am les of a har o r town on the shores of ake ario t that has essentiall remained nto hed sin e anti it and a essi le for ar haeologi al st dies it ated some 40 km to the so th-west of le andria, it stret hes more than 25 ha along the so thern shore of ake ario t at a oint where

the width of the lake from north to so th is less than one kilometre ig 1 in e 1977 the site of area has een the s e t of arhaeologi al e a ations and those dis o eries made rior to 2003 all s ggested a somewhat late o ation of the site he e a ated remains, dating from the 5th to 7th ent ries , indi ate a har o r town of onsidera le a a it with large storage fa ilities as well as li ildings of high alit ig 2 l- akharani 1983

F ig. 1: Map of Egypt: l ocation of L ake Mariout and the site of Marea ( CEA l ex A rchiv es) .

1

drogra hi ma of ower g t drawn in 1801-1802, D escription de l ’ Egypte, E tat moderne, ol , aris, 1809, late 10

57

LAKE MAREOTIS: RECONSTRUCTING THE PAST

i a a ( CEA l ex A rchiv es) .

etr so od iewi

h si i

i

as

a a

a el 1983 adek 1992 e osson 1935 131 1983, 1998 manska a ra 2008

will e more re ise on e we ha e the res lts of the eramologi al and n mismati st dies resentl nder wa

in e 2003, the entre d t des le andrines le has een working on the enins la sit ated some 100 m to the northeast of the an ient town of area to ogra hi al s r e s, ros e ting on foot, geo h si al e aminations and ar haeologi al e a ations ha e all eg n to re eal the general organisation of the enins la and the e isten e of a large worksho arter that e tended o er m h of its s rfa e

n im osing ilding onstr ted of large lo ks and h dra li mortar stood on this a sewa a o t 100 m efore its n tion with the enins la t is s all inter reted as a lightho se or landmark ett stret hing 26 m into the lake is sit ated to the so th-east of this igs 3 7 he an ient a sewa is far from straight riented so th/north for less than 300 m, it takes a t rn to the east to follow, firstl , a so th-west/north-east dire tion then west/east to the oint of the lightho se or landmark t then ret rns to the so th-west/north-east and rea hes a worksho arter he rather arti lar sha e of the a sewa and the ett to the so th-east of the landmark reate a mooring asin that is relati el well rote ted from the re ailing winds

G eneral Organisation of the Peninsula ig 3 riginall an island, a ess to the enins la was ossi le either from the lake onto a ier to the north that allowed for oats to moor, or land a ross a a sewa that was some 5 m wide that onne ted with the mainland igs 4a 4

n the north of the enins la there was a ilding o er 100 m long also ilt of large lo ks and h dra li mortar ig 3 ts northern limit ends in a s stem of terra es and stairwa s t into the ro k and leading to a ier ts general la o t aro nd a large re tang lar o rt ard might s ggest a li ilding for ommer ial r oses on erned with trade and storage, tho gh it might also e a ri h illa he resen e of istern s nder this ilding is attested a des ending assage sit ated in the entre of the west wing tr t res in ri k and h dra li mortar isi le on the s rfa e of the east wing o ld e onne ted with the isterns, or o ld e e iden e of ath fa ilities rfa e s r e s and drawings ha e not et made it ossi le to nderstand the e a t role and organisation of this ilding or the f n tion of its asso iated h dra li s stem

n 2003 the first two e a ation se tors were esta lished on this a sewa e tor 1 igs 3 5 was o ened on a er adl reser ed art of the tra k his e a ation allowed s to re eal the resen e of fo ndations of a str t re ro a l asso iated with ontrolling mo ement on and off the a sewa e tor 2 igs 3 6 was esta lished at the oint where the tra k disa ears toward the so thern art of the on ession he a sewa effe ti el sto s at this oint and ends in a sort of latform made of n mero s am hora fragments and mortar he ar haeologi al material fo nd here was relati el homogeno s and dates the onstr tion of the a sewa to etween the 1st and 3rd ent ries he hronolog of its onstr tion 58

V. PICHOT: MAREA PENINSULA

F ig. 3:

T he peninsul a of Marea ( CEA l ex A rchiv es) . 59

LAKE MAREOTIS: RECONSTRUCTING THE PAST

F ig. 4: A ) ( abov e) pier to the north of Marea peninsul a, v iew from the north- w est. B ) ( right) ancient causew ay connecting the peninsul a to the mainl and, v iew from the w est. P hoto V . P ichot ( CEA l ex A rchiv es) . F ig. 5: ( l eft) A ncient causew ay under ex cav ation of Sector 1, v iew from the east. P hoto V . P ichot ( CEA l ex A rchiv es) .

F ig. 6: A ncient causew ay under ex cav ation of Sector 2, v iew from the north. P hoto V . P ichot ( CEA l ex A rchiv es) .

F ig. 7: L andmark and j etty to the south of the peninsul a, v iew from the south- w est. P hoto V . Merl e ( CEA l ex A rchiv es) .

60

V. PICHOT: MAREA PENINSULA Occupation and W ork shop Activi ties le is rrentl fo sing on the entral art of the enins la where a arter de elo ed and e ol ed d ring the ellenisti and oman eriods his was rin i all a worksho arter the resen e of whi h is attested from the 3rd ent r

a later eriod, o ation re-ado ts the north-north-west/ so th-so th-east orientation resent d ring the ellenisti eriod on the ele ated art of the enins la r e of the enins la led to the dis o er of n mero s ar haeologi al estiges onne ted to light ind str an ones with on entrations of slag and f rna e walls, oint o t the resen e of what were worksho s, the greater art of whi h a ear to ha e een onne ted to metall rgi a ti ities he n mero s lots showing lear anomalies dete ted the magneti s r e ertainl orres ond to areas of intense metall rgi a ti it ther anomalies ma ossi l re resent the resen e of otteries and more ertainl halk f rna es from the later era 5th-6th ent ries , as has een shown the e a ations

eo h si al magneti s r e s ndertaken on the entral art of the enins la ha e re ealed the e isten e of three ma or orientations, s se entl onfirmed e a ation, that orres ond more or less to the la o t of the area at different eriods ig 8 n the ellenisti eriod 3rd-1st ent ries there are two orientations one so th-west/ north-east for the low-l ing art of the enins la and the other north-north-west/so th-so th-east for the ele ated art t the end of the ellenisti eriod and the eginning of the oman era, the ma or orientation sed within the low-l ing areas follows a north-north-east/so th-so thwest a is, while the north-north-west/so th-so th-east orientation a ears to ontin e in the ele ated areas t

Metal l urgic activ ities he e a ations ndertaken in e tor 4 see ig 3 ha e re ealed e tensi e f rna e a ti it e tensi e not so m h in terms of rod tion, whi h is still diffi lt to estimate,

F ig. 8: R esul ts of the geophysical surv ey undertaken on the Marea peninsul a. ( T . H erbich & CEA l ex A rchiv es) . 61

LAKE MAREOTIS: RECONSTRUCTING THE PAST t rather in the n m er of hearths dis o ered n fa t, in this se tor where e lorations oth geo h si al and is al s ggested an em t terrain, e e t for a dit h sit ated to the western edge, some 40 hearths ha e een nearthed most of whi h are l ing a mere 0 15 m eneath the s rfa e and all are on entrated in the west and north-western art of the se tor f different si es the si e de ending to a large art on the e tent of destr tion the are either ir lar in lan or o asionall re tang lar hile ertain e am les still hold fragments of o e ts and/or slag, others are ra ti all destro ed he are not all ontem orar , t the ma ha e f n tioned in gro s of two, three or fo r, se eral gro s eing in o eration at the same time he est onser ed hearths were leared when e a ating r ish dit h 4024 sit ated on the western edge of the se tor ts fill om osed fragments of am hora, erami s, ones, some fragments of faien e o e ts, har oal, some fragments of iron and seashell he ma orit of the material nearthed was hara teristi of the 1st ent ries and , howe er, ertain elements an e dated to an earlier ellenisti eriod, e g a worked lam of the 3rd ent r and am hora and erami fragments of the 3rd and 2nd ent ries

F ig. 9: Sector 4 - H earths in F S4024, unde P hoto V . P ichot ( CEA l ex A rchiv es) .

r ex cav ation.

ti eriod o ation itself otentiall related to worksho a ti ities, an e roadl dated to the first ent ries of the oman era e tor 3, resentl nder e a ation see ig 3 , was o ened in 2003 on the western art of the island and was e tended in 2007 into a one of noti ea le magneti anomalies towards the east ork on the western art was temoraril s s ended to allow, firstl , for the om letion of the st d of e tor 4 and then the e a ation of the eastern art of e tor 3 n 2003 two ensem les of ildings se arated an alle were nearthed in the western art of the se tor ig 10 he ensem le oriented north-north-east/ so th-so th-west was dist r ed in its north-eastern art large onstr tions orientated on a north-north-west/ so th-so th-east a is a h ilding is om osed of small worksho rooms o ening onto the alle , while at the a k there was a large room sed for storage or ha itation ne of the a ti ities identified in this worksho arter is ol metall rgi al and e his is attested the remains of hearths, flooring and o ation la ers hara teristi to this t e of a ti it , t also the dis o er of ref se material onne ted to metall rgi al a ti it slags, dri s, hammers ales, nfinished o e ts, et

his dit h 4024 was t se eral gro s of hearths ig 9 mong these str t res, two feat res are ite well reser ed 4034 and 4035 ltho gh made of hearths with different lans one is om osed of re tang lar hearths, the other of ir lar the are onstr ted in a similar fashion a hearth of 0 30 m diameter to whi h is atta hed a m h smaller t dee er hearth, the e a t f n tion of whi h remains to e determined hese feat res were onstr ted of la in whi h fragments of erami s and am horae were la ed to a t as s ort he str t res are well onstr ted and the interior walls are erfe tl smooth heir fill is of har oal, aked la , small fragments of iron and small f rna e slag he de ris dis o ered s ggest that metal working a ti it was taking la e, asso iated with the fa ri ation and/or re airing of small iron o e ts with the se, in ertain ases, of o er allo s and lead sso iated with the hearths, tra es of ost-holes re eal the resen e of light str t res that o ld ha e een more or less ermanent take holes near the hearths, howe er, indi ate the ro a le se of fireg ards, wind reaks and/or walls marking o t ertain s e ifi ones, e g for storage

o r works a es are resentl eing e a ated he o tline of walls marking o t storage and ir lation ones, ha e een re ognised in three of these rooms n a e 10 a series of re-laid eaten earth floors ha e een re ealed eneath the last o ation le el ording to the initial st dies of the ar haeologi al material, it seems that le els of worksho a ti it an e dated to the 1st ent r to the 1st ent r limited tren h has een o ened in the low-l ing one to the so th of e tor 3 he water ta le was i kl rea hed at a little more than 1 m elow the s rfa e hree s essi e floors and asso iated metall rgi a ti it la ers lots of har oal, n mero s small metal dri s dating to the 2nd to 1st ent ries were re ealed, as well as a wall whose orientation re ealed the e isten e of an ensem le laid o t on a north-west/so th-west a is

he remains of ertain f rna es onsisted of no more than the er ottom tra es of stakes and osts, a few stri s of eaten earth floors and a art of the first fo ndation o rse of a wall 4040 oriented west/east set on the edro k he estiges of this se tor are er worn down, hen e the diffi lt in en isaging the s atio-tem oral organisation of these worksho s itho t going into onsiderations of ontem oranit of f n tion and the length of o ation, and ha ing et to om lete the st d of the ar haeologi al material, the im ortant metal working a ti it in this one, im lanted on the remains of ellenis62

V. PICHOT: MAREA PENINSULA

F ig. 10:

O v eral l pl an of the w estern z one of Sector 3 ( CEA l ex A rchiv es) . 63

LAKE MAREOTIS: RECONSTRUCTING THE PAST

F ig. 1 1:

O v eral l pl an of the eastern z one of Sector 3 ( CEA l ex A rchiv es) .

A rea D ev oted to Cul t A ctiv ity his o ation is ontem orar with that of a onstr tion 300 artiall nearthed in 2007 in the eastern art of e tor 3 igs 11 12 a ed wa some 1 15 m wide sit ated to the north, leads to the entr of the ilding he ilding2 is re tang lar in sha e a ro imatel 10 35 m 11 75 m, oriented north-north-west/so thso th-east along its longit dinal a is he eri heral walls of an a erage thi kness of 1 4 m and with on a e e ternal s rfa es, en lose a s a e di ided into se eral ells se arated walls that are 0 8 and 0 9 m thi k

he ilding onsists of fi e different s a es rom the o tside one enters into a large re tang lar room of 3 5 m 7 55 m, di ided into two a a ement of sla s that ts thro gh the middle his a ement is a ro imatel 0 85 m wide, th s narrower than the o ening of the doorwa 1 m a ro , and is a ed with irreg lar limestone sla s t onne ts the entran e with the onl door within the ilding that is ier ed thro gh the interior trans ersal ross-wall ts slight irreg larit and the fa t that is deli eratel off the a is of the room, leads one to onsider the nat re of the flooring that on e la on either side here are few l es left to satisf this er hree fragments of sla s atta hed to the a ement at three different oints along its line might indi ate that the missing flooring was made of the same materiel as the a ement itself, at least on the edges he walls of this room were o ered with a laster ainted to re resent ala aster he remains of another ainted oating made to re resent red, white and la k mar le fa ing might ha e elonged to a later stage or more ro a l to the de oration of a room on the er floor n front of the interior door, a re tang lar a it 0 33 m 0 18 m had een d g into one of the sla s of the a ement t is entred on the median a is of the doorwa and in the ottom one an see in the middle, a small de ression, o al in lan and oni al in se tion, whi h is filled with a l g of n aked la his re tang lar a it

n the north-eastern art of the ilding, the do le-fa ed masonr is om osed of medi m si ed t lo ks of limestone with an inner filling his onstr tion method differs learl from the opus incertum of the other walls of the onstr tion his o ld re resent an a t of restoration d ring a ossi le se ond hase of the ilding s se he er art of the walls were ilt of m d- ri k that were fo nd fallen and disintegrated in the interior of the ilding

2 he reliminar st d of this str t re was ndertaken air , ar hite t-ar haeologist, , 3134, entre d t des le andrines

64

V. PICHOT: MAREA PENINSULA F ig. 12: ( l eft) Eastern z one of Sector 3 unde r ex cav ation. P hoto V . P ichot ( CEA l ex A rchiv es) . F ig. 13: ( bel ow ) Sector 3 - O ne of the tw o l imestone sphinx es discov ered in the tow er buil ding ST 30. P hoto V . P ichot ( CEA l ex A rchiv es) . F ig. 14: ( bottom) Sector 3 B ronz e candel abra after restoration in the CEA l ex l aboratory. P hoto A . H ussein ( CEA l ex A rchiv es) . ro a l held a s ort, a ase earing either an element of de oration, e g stat ar , or a f n tional o e t he interior doorwa was framed two ilasters, of whi h onl the ases remain, that ore two little, limestone s hin es ig 13 that were fo nd dismantled in the ne tdoor room his interior doorwa o ens onto a room some 2 40/2 35 m 4 45 m t the end of this s a e, to the so th, two small assem lies of lo ks, the sides of whi h ha e een o ered leanl t stone fa ing, stood against the walls in the orners hese two assem lies form the ase of two stair ases that led to the er floors he two lateral s a es sit ated to ea h side of this distri tion s a e are om letel losed and ina essi le at this le el he eastern s a e was filled with la and other r le, while within the filling of the other s a e there is a lo k of masonr in opus incertum that most ro a l s orted the weight of a now-disa eared s erstr t re hen onsidering the different as e ts of this onstr tion, its interior la o t, as well as the ar haeologi al material dis o ered here, a h othesis ma e that the f n tion of this ilding o ld relate to the re resentations of the tower-ho se on iloti mosai s, or that it was religio s his ilding was art of a m h igger ensem le, the e tension of whi h to the north-west was artiall n o ered e a ation in 2008 and was still relati el inta t eneath the onstr tion le els of ilding 301 wo s a es learl similar in si e and limited to the east wall 3114, were dis o ered heir western limit has not et een determined he most so therl s a e is diffi lt to inter ret t was ra ti all destro ed in its entiret the im lantation of wall 3082 that lies in art on wall 3121 that se arates the two s a es he e a ation of the northern s a e, delimited to the north 3116, has re ealed the remains of flooring of om a t la in whi h were set se eral erami storage ars ltho gh we do not et know for s re the d ration of its se, the e iden e of the ar haeologi al material, in l ding a ellenisti lam with se en wi k-holes, a faien e endant and a ron e andela ra ig 14 , wo ld s ggest that this ensem le was 65

LAKE MAREOTIS: RECONSTRUCTING THE PAST settlement and organisation of a si ea le town on the lakeshore ntil the resent, o r knowledge of the o ation of the site of area was limited to the antine eriod 5th to 7th ent ries he work of the le on the enins la has ro en a settlement e isted here well efore the 5th ent r he e a ations ndertaken in e tors 3 and 4 ha e re ealed a large o ation dating to the end of the ellenisti and the eginning of the oman era, as well as n mero s signs of a ellenisti o ation rior to the 2nd ent r he o ort nit at area to e a ate and st d a worksho arter of this si e, in an en ironment so well defined the fa t that it stands on a enins la, is reall ite e e tional n addition to gathering information on the rod tion itself and on the rod tion lines, it is also ossi le to st d the s atio-tem oral organisation of the site in the er heart of the enins la and to resit ate it within its immediate en ironment ha itation, religio s area, et Bibliography e osson, , 1935, Mareotis: B eing a Short A ccount of the H istory and A ncient Monuments of the N orth- w estern D esert of Egypt and of L ake Mareotis ondon D escription de l ’ Egypte, Etat moderne, ol 1, 1809 aris l- akharani, , 1983, e ent a ation at area in g t n rimm, einen inter eds , A egyptiaca T rev ernsia I I , D as R öm isch- B yz antinische Ä gypten. A kten des international en Symposions 26.30. Se ptember 1 978 i n T rier, 175-186 ain l- alaki, ahmo d e , 1872, Memoire sur l ’ antique A l ex andrie, s es faubourgs, s es env irons o enhagen etr so, , a el, , 1983, area antine ort on g t s orthwestern rontier A rchaeol ogy 36 5 62-63, 76-77 od iewi , , 1983, le andria and distri t of areotis G raeco- A rabica 2 199-216 od iewi , , 1998, rom le andria to the est and and aterwa s n m ere r ed , Commerce et artisanat dans l ’ A l ex andrie hel l é nistique et romaine. A ctes du col l oque d’ A thè nes, organisé par l e CN R S, l e L aboratoire de cé ramol ogie de L yon et l ’ Ecol e franç aise d’ A thè nes, 1 1 - 12 dé cembre 198: 93-103 thens, aris adek, , 1992, he aths at the an ient har o r of area Sesto Congresso internaz ional e di Egittol ogia, ol 549-554 tra o, G eography, eineke transl , 1877 i siae ma ska, , a ra , eds , 2008, B yz antine Marea. Ex cav ations in 20203 and 206 , Marea, ol 1 rak w

F ig. 15: Sector 3 - H oe chal k burner. P hoto V . P ichot ( CEA l ex A rchiv es) . in se from the end of the 3rd ent r t was artiall destro ed the installation of one or se eral o ations to the north d ring the oman eriod that were in art onne ted to worksho a ti it ,3 and thereafter halk rners that were a ti e in a later eriod ig 15 , most ro a l in the 5th and 6th ent ries , and that largel destro ed all onstr tion in this one Conclusion he areotis area, toda semi-desert, was on e a fertile region with wides read agri lt ral a ti it n the raeo- oman era, ake ario t was a erita le inland sea with intense traffi s a one of mo ement and e hange, it was onne ted to the ile anals and with the sea at se eral oints thro gh the o rse of its histor assage also allowed traffi a ess to the anal of le andria he tolemai a ital had a lakeside ort with a s and wareho ses that tra o 17 1 7 onsidered ri her than the maritime har o rs of the it area is one of the est e am les of arioti lakeside towns whose de elo ment was tied to e hange etween le andria and its hinterland e orts of agri lt ral rod tion and lo al ind str glass, metal, et , im orts of raw materials and ilgrimages d ring late anti it ts state of reser ation allows s to st d the ro lems of

3 he mission of 2008 has ro ght to light n mero s str t res onne ted to worksho a ti ities of the oman era to the west and north of the tower ilding he are resentl nder st d and will e the o e t of a s lementar mission in 2009

66

M. RODZ IEWICZ : MAREA AND PHILOXENITE’ S ARCHAEOLOGICAL EVIDENCE

On Interpretations of Archaeological Evidence Concerning Marea and Philoxenite Mieczyslaw D. Rodziewicz

The largest and best preserved ancient lake harbour southwest of Alexandria was located on the map of the Mariout region by Mahmoud El-Falaki in 1872 with a description and interpretation of the site as the centre of the ancient areoti distri t eing identified as the site of area lFalaki 1872: 96). For a long time this interpretation was considered as correct and unshakable. It was expected that with ontin o s ad an es and de elo ment of field archaeology in the region, this interpretation would be substantiated by more archaeological evidence. However, in the first half of the ast ent r the site was not in estigated by the most active archaeologists in Alexandria such as Breccia and Adriani. Instead, two other sites in the Mareotic region captured the attention of these great arhaeologists of the last ent r he first s h la e was Abu Mina discovered in the desert south of Alexandria and excavated from 1905 by German archaeologist C.M. a fmann 1908 . The second important site was the large city of Taposiris Magna, west of Alexandria, that spread around the Ptolemaic temple of Osiris, with the neigho ring site of linthine, where re ia 1922 353 and driani 1940 on entrated their field resear h in the first three decades of the last century. There was also a short German campaign near Amreyia, situated in the desert south of Alexandria, where very informative architectural remains of earl hristian date ha e een n o ered ilmann, et al 1930

ten re ia 1922 335 and orster 1922 , the ort of Marea is hardly mentioned compared with such important places west and south of Alexandria as Taposiris Magna and Abu Mina. The best equipped excavations in the Southern Desert were made by expeditions organized by the German Archaeologi al nstit te in airo in the 1930s he were arried out near the modern market place and settlement of Amre ia ilmann, et al 1930 e ifi r ral t es of earl Christian structures characteristic of the region, although rather modest and of an introductory nature, were recorded and published very quickly. It is a great loss that the excavators did not come back to the region after the second world war to continue the research. In the second half of the past century, excavations at Abu Mina, undertaken by the German Archaeological Institute at Cairo, were resumed under the direction of architects, Mü ller-Wiener and later Grossmann. They extended the area of research around Abu Mina, the largest pilgrimage centre of early Christian Egypt, and surveyed the neighbouring territory of the ake ario t har o r ller- iener 1967 and der 1967 , de oted their time to in estigating not onl the ancient settlements in the region, but also the sources of local building material, including ancient quarries located along the southern shore of the lake, particularly one lo ated near the modern illage of ahig ller- iener 1967 104-117 der 1967 118-131 heir interests also extended to an analysis of the building materials visible on the surface of the land along the southern shores of Lake Mariout and the surrounding area, with the clear intention of identifying a link between the chain of settlements in the desert and the pilgrimage centre at Abu Mina. The arti le of ller- iener 1967 , entitled iedl ngsformen in der Mareotis” , was based on rational observation and the description of visible remnants of ancient structures on the s rfa e igs 1 2 n the har o r of so- alled area, rior to an e a ations, he identified on the s rface, a double bath, structured frames of the quays, with accompanying buildings, and the most important construction, the mon mental three a sidal h r h ller- iener 1967: 106).1 ll his identifi ations were orre t, and the have not altered as a result of later excavations. More im-

The most complete description of the whole area was offered by Anthony De Cosson, the former director of the railwa west of le andria, in the 1930s is ook on Mareotis was based on then available literature on the subject and his extensive knowledge of the land from Ale andria westwards far e ond arsa atr h an ient arethoni m e osson 1935 131 n his des ri tion of Marea he follows the description and name of the extensive lake port remains situated south-west of Alexandria given by Mahmoud El-Falaki. In both, El-Falaki’ s and De Cosson´ s descriptions, the site was presented as 1.5 km long, but not very broad, and limited to the chain of visible ruins of fallen walls along the lake shore. Nevertheless, in their opinions the place was the most important centre of the historical Mareotic region, as mentioned in ancient literary sources. Other modern publications concerning ancient Mareotis have dealt very little with the largest lake har o r in the region et, in the first half of the last entury the site was recorded on several maps of the Surv ey of Egypt series, but was rarely presented in published photographs. The best known are pictures showing the western jetty of the harbour surrounded not by water, but instead by marshes. In the most popular books of that time writ-

1 ller- iener 1967 106, n 16 n area sind o wohl das a fgehende a rerwerk weitgehend fortges hle t sein s heint no h ahlrei he a ten erkrnn ar grosses o el ad, afen efestig ngen mit Kaianlagen und drei langen Molen, Kirche mit drei-apsidialem r ndriss sw sw eine n here nters h ng des rtes w re mal angesi hts seiner ede t ng in der r h eit sserst w ns henswert

67

LAKE MAREOTIS: R

PAST F ig. 1: Central part of the pl an of Mareotis draw n by Mül l erW iener in 196, w ith the pil grimage road from the l ake harbour to A bu Mina and settl ements, am ong them settl ement “ M” in H uw aryia v il l age ( after Mül l erW iener 1967: F ig.1) .

F ig. 2: A bu Mina w ith surrounding settl ements/karms, and pil grimage road tow ards the l ake harbour ( after Mül l er- W iener 1967: F ig. 4) .

68

M. RODZ IEWICZ : MAREA AND PHILOXENITE’ S ARCHAEOLOGICAL EVIDENCE ortantl , he identified a se arate settlement on the lan re rod ed on his ig 1 see ig 1 , at a distan e of 1.5 km south from the ruins of the lake harbour known as area ller- iener 1967 110 e resented its dimensions 600 800 m and des ri ed riefl the remains of a brick cistern located at the edge of the modern road, the same that later was identified as a water reser oir ser ing the do le erist le ilding, lished its first e a ator l- akharani, as the antine o se l- akharani 1983 175 h s, ller- iener learl se arated all these inland situated structures from the harbour to the north, and er learl stated that settlement was not a part of a larger town plan comparable with ancient Alexandria, as was to be suggested a decade later, and supported by his successor heading the excavations in Abu ina rossmann 2003 13

saved most of the monuments excavated by the Alexandria ni ersit edition, headed rof l- akharani The winery in Huwaryia village, excavated in 1977, reei ed limited re onstr tion mainl for rote tion t also didactic reasons), and was covered with a roof. The excavation of the so called Byzantine House, which most probably served as a hospice for pilgrims heading to Abu Mina, was completed. It was partly reconstructed to prevent further damage and also for didactic requirements. The remains of buildings at the lake harbour were treated similarl od iewi 2002, 2003 ring these works many previously completely unrecognised structures were identified, th s e tending the ossi ilities of defining more features of the local topography, its culture, economy and technology. If they would have been known to Prof. ElFakharani at the time of his activities in the area, he would ertainl ha e hanged some of his first h otheti al identifi ations, and his general o inion of the site

n 1978 the first ar haeologi al e a ations of the soalled site of area were finan ed le andria ni ersity under the direction of Professor of Archaeology, Dr Fawzi El-Fakharani, a very energetic organizer and gifted speaker. In a very short time his discoveries at the site became very popular both nationally and internationally. rof l- akharani lished in 1983 an e tensi e re ort on the survey and excavations of a vast area of Mareotis, a o t 6 km long and 4 km road ther li ations of the site of area adek 1978 67 and etr so and a el 1983 62 followed, altho gh anal sis of a aila le material and structures excavated there were not completed. At that time, there arose the notion of a very large and prosperous city of Marea around the existing remains of the lake harbour, which some supposed to be equal in size to ancient Alexandria. The Mareotic settlements 1.5 km south of the port, and the Ptolemaic rock-hewn tombs situated a o t 5 km west at the site alled assim ia , reviously investigated by Mül ler-Wiener, were included by l- akharani in the it of area l- akharani 1983 176, 186). The lack of any continuation and evolution of chronologically parallel architectural structures between these distant places was not taken into consideration by the enth siasts of this notion of the great it of area

The most required change in the general knowledge of the site on erns the e tension of the r an area of area to the west, i e to the tom s of the tolemai eriod at modern assimi ia site, ca 5 km from the city), which is still considered by some archaeologists as a part of the western necropolis of the capital of the whole Mareotis region l- akharani 1983 176 agnall ath one 2004 75).2 Of three hypogea, still well preserved in the 1970s, only traces of the largest, located on the southern slope of the ro k ridge, are still re ognisa le he identifi ation l- akharani of assimi ia site as the western nero olis of area is not ar haeologi all s stantiated, because similar hypogea existed on the rocky ridge near war ia illage 1 5 km from area , and all along the same ridge were located other types of tombs such as chambers, shafts or pit tombs, which were dispersed over the area, also considered by some archaeologists as belonging to the it of area rthermore, the e tensi e ancient ruins with a subterranean tomb, now located near the modern Huwaryia railway station, which supposedly delimited the so thern r an area of an ient area , are according to the results of the survey of 1990s EAO Delta est ns e torate n lished , the remnants of an isolated structure with a church inside, which may indicate one of the numerous Christian monasteries that existed here in the early Christian period.3 There were no detectable traces of urban features such as street connections between this structure and the group of low quality houses with an associated wine press, that were excavated and published by l- akharani 1983 183-184 his site was earlier identified ller- iener as a se arate settlement ller-

In 1972 Fraser, the author of a major work on Ptolemaic le andria, wrote, ltho gh area had een an im ortant town in Pharaonic times, these remains are evidently late structures, they are not Ptolemaic and may be Byzantine or earl ra raser 1972 146 ome ears later an essential question arose about the location of the pilgrimage lake harbour, known from the Coptic Encomium on St. Menas as Philoxenite, a place that is also mentioned in other hristian so r es res her 1946 147-148, 1949 15-16 od iewi 1983, 2003 The next phase of research connected with the lake harbour, went together with the protective work undertaken by the Egyptian Antiquities Organization in 1982, which required a factual analysis of all preserved archaeological material essential for planning and intensifying the protection of the cultural heritage in this area. I was involved in this process, particularly in the protective works, which

2 l- akaharani 1983 176 he dis o er of rials in the west and south marked the limits of the town westwards and southwards since ancient cemeteries were customarily located outside of inhabited area agnall ath one 2004 75 n the emeter to the west of the town …” . 3 om are these str t res to those e a ated the wiss at ellia, asser 1983, 1986 with f rther referen es in A ctes du Col l oque de G enev e, 13 au 15 aout 1984 , Geneva 1986.

69

LAKE MAREOTIS: R

PAST ing a now horizontal stratigraphy, unlike most of the Egyptian ancient settlements which grew upwards and form hara teristi hills alled koms ontrar to the first and still alid identifi ation of this ar haeologi al site ller- iener in the 1960s in arti lar the water istern belonging to the big Byzantine House that he published), as a separate construction in a rural environment, the site is still directly or indirectly considered by some archaeologists as a centre of the urban area of an extended town alled area agnall ath one 2004 74-76 5 Thus, despite the growing archaeological evidence that strongl o oses the identifi ation of the lake ort as Pharaonic Marea, we still read about the remains of an an ient town tadtanlage lo ated as far so th as the modern illage of war ia rossmann 2003 15 s a consequence of such interpretations of the ancient village, general conclusions were formulated concerning the lo ation of wineries in r an en ironments agnall ath one 2004 74-76 rossmann 2003 15, n 27 rthermore, another conclusion drawn, was that the so-called rof l- akharani antine o se at war ia is the most convincing archaeological document attesting the existence of the urban centre of ancient Marea, where in reality at this location there existed a large building structured in a rural manner, surrounded by cultivated land. Judging from its huge dimensions and the church built inside, it may have served as a very comfortable and safe pilgrimage hospice from the early 6th to late 7th century AD, i.e. exactly at the time of prosperous pilgrimage acti ities to ina od iewi 2003 27-47 o older or younger constructions in the area of the building have been registered. The proper understanding of the function of this large building is crucial for the interpretation of the character of the area, which was described by Mül leriener 1967 who ndertook a detailed s r e of the area in the 1960s), as a rural settlement, not the centre of any larger town. He mentioned visible surface remains of a brick built water cistern that was located right on the edge of the modern road, that are still isi le toda see ig 3 n late anti it , this istern was sit ated in a garden or in an open space surrounding the large double-peristyle ilding the antine o se of l- akharani llerWiener thought that the cistern may eventually be conne ted with the ath ller- iener 1967 owe er, the e tensi e neigh o ring ilding the antine o se , whose excavation was completed in the EAO in the early 1980s, a eared to e the largest residential edifi e known hitherto in the whole Mareotis region. It was equipped with a church in the central, eastern wing, between the two extensive colonnaded peristyles. The church had multi-

F ig. 3: B yz antine H ouse/H ospitium at the pil grimage road from the l ake harbour to A bu Mina ( draw n by M. R odz iew icz ) . iener 1967 110 see ig 1 n the so thern art of this settlement an underground water cistern, a group of low quality rooms and, further to the west, a medium-sized winer , were artl n o ered in 1992/3 he were associated with pottery of the late 1st century BC to the middle of the 2nd ent r a thor s identifi ation This winery appeared to be much older than the winery dis o ered rof l- akharani 1983 af 36 ther architectural remains from this settlement are younger. At the most northern part of the ruins typical Muslim burials were located, which, according to Prof. El-Fakharani, were associated with glazed pottery sherds and a probable atimid oin l- akharani 1983 176-177 4 rther to the north, an e tensi e ilding identified El-Fakharani as the Byzantine House was separately built o tside the settlement l- akharani 1983 184-186 t occupied a surface of over 1,500 m2 and was surrounded on all sides lti ated land od iewi 1983, 1988a 267-277, 1990, 2002 1-22 ig 3 he ho se was not covered by structures belonging to other buildings, which means that it was situated in a typically rural environment, at the far edge of the older, but still inhabited village, hav-

5 agnall ath one 2004 74-76 76 r haeologi al e iden e for wine production has also come to light. South of the limestone ridge are the remains of two wine-producing establishments. The larger and more interesting is aligned with the middle of the town” . However, such installations as wine factories and pottery kilns are not known in the very centres of ancient towns, where life was regulated by city law. ite the ontrar , these are er t i al r ral esta lishments

4 l- akharani 1983 176-177 ew ie es of slami otter sherds with green glaze and some white lines, and a piece of Fatimid coin were discovered just above the dead” .

70

M. RODZ IEWICZ : MAREA AND PHILOXENITE’ S ARCHAEOLOGICAL EVIDENCE

F ig. 4: B yz antine pil grimage harbour/P hil ox enite ( N os. 5- 17) and w estern/coastal rural remains ( N os. 1- 4) . O n the eastern side, s tructured causew ay ( N o. 18) ( draw n by M. R odz iew icz ) . K EY 1– J etty of pre- B yz antine date 2– R ural estate w ith l arge w ater w heel 2a – Smal l kil n 3– i i s rural character 4– Structures w ith mooring pl ace

si i i h pil grimage harbour: 5– P ier 6– W ater w heel ( saqia) 7– P ubl ic doubl e bath 8– I nsul a w ith shops 9– P ubl ic buil ding 10 – P ubl ic l av atory 1 1– K ibotos?

coloured, geometrical opus sectil e mosai s on the floor and a ri hl ornamented northern a tister od iewi 1983, 1988a, 1988 he a tister floor was made of mortar and painted red in a geometrical pattern composed of triangles. After extensive cleaning of the area, it appeared that beside the large church in the centre and numerous rooms aro nd an e tensi e o rt ard and ro a l on the er floors , a olle ti e nine seat la ator e isted there with two smaller ones aside. They were all built in the so thern art of the edifi e, the stair ase leading to the er floors, and lose to the athing e i ment lo ated in the south-western corner of the house. The bathing facilities were not s ffi ient to ser e the needs of ilgrims e entually stationed there, but a full programme of cold and hot baths was available just about 1 km to the west, in the large public bath that was surveyed, excavated and restored by EAO Delta West Inspectorate in the early 1990s.6 The bath, datable to the Byzantine period, was not attached to

12 – 13 – 14 – 15 – 16 –

17 – 18 – 19 –

P ier T ransept B asil ica ( after G rossmann 20 2: F ig. 9) Eastern harbour W ater w heel ( saqia) P ubl ic bath R ock- hew n tomb/hypogeum Causew ay l eading to the isl and P il grimage track to A bu Mina

any architectural remains. It was built in an open area and surrounded by cultivated land. The area of the settlement, which extended from the uplands to the lake, bears traces of various human activities, such as limestone quarrying, various burials on the rocky ridge, clay extraction, pottery firing in the n mero s kilns registered there, and lti ation down to the shores of the lake der 1967 118-131 The individual buildings, whose traces have been excavated and surveyed there, were built in a rural manner. Strictly urban construction methods of building are identifia le onl aro nd the lake har o r, erroneo sl identified as haraoni area l- alaki 1872 96 ost of the buildings at the port were constructed in one phase, with similar onstr tion rin i les ig 4 he onstr tion methods are comparable to buildings at Taposiris Magna, Alexandria and also the pilgrimage centre of Abu Mina rossmann, et al 1982 rossmann, et al 1984 123-151 They are not similar to the structures of neighbouring rural settlements and the karms/karum” located further south. West of the Byzantine passenger harbour Philoxenite area , on a ro gh nat ral ro k oastline, are the re-

6. The results of which were presented by Ahmed Abd El-Fatah at the aln orient onferen e, held in le andria, e em er 2006 in ress

71

LAKE MAREOTIS: R

PAST

served traces of a small size mooring facility and a quite well preserved short but wide and relatively high, jetty that differs remarkably from the two long Byzantine quays of the ilgrimage ort ig 4 1 he ett is onstr ted from very large sandstone blocks without hydraulic mortar n front of it a l ster of r ins ig 4 2 is still isi le, whi h s rro nd a large o rt ard, water wheel saqia), cistern and channels. They probably mark structures of a coastal rural estate similar to the one at Borg el-Arab. Early Roman building material has also been recorded. In addition, there are traces of another water wheel, a small kiln ig 4 2a , and widel dis ersed tra es of f rna e activities. These remains depict the kind of rustic-coastal h man a ti ities with small si e mooring la es ig 4 4 , created prior to the construction of a large passenger harbour at the promontory. It is worth noting that during the very last era of the passenger harbour of Philoxenite, with large buildings already abandoned and decaying, old local activities returned, motivated by the rural economy, with an oil press and a pottery kiln installed on top of the ruined constructions.

t res at the lake ort rossmann 2003 13-20 7 Arguments against this proposition vary from self-contradictory to those based on out of date, unproven archaeologically theories, particularly that of the existence in that place of a large town, area , the haraoni a ital of the distri t, with a generalised opinion that the old town experienced an unprecedented economic boom in a period of well-documented and steady economic decline of Egyptian cites.8 With a new round of excavations at the lake port starting in 2000, we e e ted a flow of new information and o jective interpretations of freshly unearthed structures and artefacts.9 However, this was not to be and the lack of a rational argument in the presentation of newly excavated material from the harbour during the conference on Medieval Alexandria in 2002, motivated me to publish a paper entitled hilo enite- ilgrimage ar o r of ina od iewi 2003 hree ears later, fo nd a statement in the am hlet lished the olish ission 2006 saying that there are several reasons to assume that Philoxenite was built within the city of Marea, thus supporting the interpretation of the site as that of the location of Philo enite, whi h had first s ggested in 1983 od iewi 1983 202, 2002 owe er, the elief that area lies nder the Byzantine buildings of Philoxenite, despite the lack of any older urban structures there that should date back at least to the time of Psametic I, is still maintained. In the

It is the author’ s belief that the lake port in Mareotis should be interpreted as a disembarkation point for Christian pilgrims travelling to Abu Mina, and not as a town with a long tradition matching the history of the ancient a ital of a haraoni nome od iewi 1983, 1988a, 1988b, 2002). This opinion was formed on the back of extensive personal study of the whole lake harbour area from the late 1970s to the mid-1990s. The text of the Coptic Encomium and other Christian sources serve to support this iew res her 1946 147 in e, in the earl 1980s, the water of Lake Mariout was at a very low level, all jetties and other structures at the harbour were dry, and research, as well as all necessary examination of the construction hases, was m h easier than nowada s or identifi ation of the chronology of the site I used my knowledge of the pottery acquired on other excavations in Egypt and elsewhere, t rimaril le andria od iewi 1976 The most informative and numerous sherds were those spotted in the mortar of the lowest parts of the jetties. They belonged to Late Roman Amphora 1, which clearly fi es the hronologi al hori on of the ig har o r in the Byzantine period and parallels the chronology of Abu Mina. The types of the structures, the extension of the r anised s rfa e, three large water wells ig 4 6, 4 15 and a er small lo al emeter hewn in a ro k hill ig 4 17 , and a er large h r h ig 4 13 and e tensi e harbour, lead me to conclude that the site did not develop gradually, but that is was constructed in a short time and a ording to a er s e ifi rogramme rthermore, there were no older or younger pottery sherds collected at the site.

7 rossmann 2003 13-20, onsiders m o inion na e ta le, t his arguments are not backed by any factual documentation concerning the chronology of the artefacts in the lake harbour and pilgrimage centre in ina whi h are mostl not et lished h s, his o inion on differences in chronology of the bath in the lake port and of the pilgrimage centre at Abu Mina can be taken only as his personal supposition, not supported by the necessary documentation and study of the pottery collected there. 8 ording to agnall 1993 108 , des ite ontin o s e onomi decay in late antiquity, not all cities in Egypt suffered regress, and Marea in the immediate hinterland of Alexandria enjoyed a period of ros erit e sa s that the flo rishing de elo ment at area, on the lake which served Alexandria´ s back door by water, points to vigorous commerce by this route in the late period” . So, the old Pharaonic city of Marea, of which we still do not have any archaeological evidence and whose lo ation is still nknown, in agnall s li ation flo rished in the 3rd century AD. 9. The Polish Mission, active there since 2000, is headed by Dr. Hanna Szymanska from the Archaeological Museum in Krakow. With great sadness, we observe subjective methods of interpretation of the archaeological evidence obtained there. At the conference devoted to Medieval Alexandria organized by CEAlex in 2002, Mrs. E. Wipszycka from the olish ission s oke a o t the identifi ation of hilo enite in the light of new discoveries. Her weak documentation mobilised me to publish the 2003 the arti le hilo enite- ilgrimage ar or of ina n 2006, a special exhibition was prepared by the Polish team about the e a ations in area t a om anied the onferen e aln orient at which Mrs. Szymanska presented a paper on a new bath from the har o r his e hi ition entitled area - olish a ations in g t 2000-2004 ond ted ar haeologists from the r haeologi al seum in Kracow under the auspices of the Polish Centre of Archaeology of arsaw ni ersit , was s orted a am hlet with a short te t signed in whi h we an find the assage area or holo enite Identifying the accurate name of the town is a challenge that the Polish Archaeological Mission has undertaken. We have several reasons to assume that it was within Marea that the city of Philoxenite was built.”

h s, in 1983 at a onferen e in thens rae o- ra i a I interpreted the harbour as a specialised port for pilgrims travelling to Abu Mina, as described in the Coptic Encomium, whi h also mentions its name as hilo enite res her 1946 15 he des ri tion of the site in the o ti Encomium matches quite well the preserved Byzantine struc72

M. RODZ IEWICZ : MAREA AND PHILOXENITE’ S ARCHAEOLOGICAL EVIDENCE Bibliography Abd el-Aziz, M., 1998, Recent Excavations around Abou ina n m ere r ed , Commerce et artisanat dans l ’ A l ex andrie hel l é nistique et romaine. A ctes du col l oque d’ A thè nes, organisé par l e CN R S, l e L aboratoire de cé ramol ogie de L yon et l ’ Ecol e franç aise d’ A thè nes,1 1- 12 dé cembre 198: 65-73 thens, Paris. Abd el-Fatah, A., 1998, Recent Discoveries in Alexandria and the hora n m ere r ed , Commerce et artisanat dans l ’ A l ex andrie hel l é nistique et romaine. A ctes du col l oque d’ A thè nes, organisé par l e CN R S, l e L aboratoire de cé ramol ogie de L yon et l ’ Ecol e franç aise d’ A thè nes,1 1- 12 dé cembre 198: 37-53 thens, Paris. driani, , 1940, A nnuaire du Musé e gré co- romain 1935193. Alexandria. agnall, , 1993, Egypt in L ate Antiquity. Princeton. agnall, , ath one, , 2004, Egypt from A l ex ander to the Copts. London. Breccia, E., 1922, A l ex andrea ad A egyptum: G uide de l a V il l e A ncienne et Moderne et du Musé e G ré co- R omain. Alexandria, Bergamo. Décobert, C., 2002, Maréotide médievale. Des Bédouins et des hr tiens n o ert ed , A l ex andrie Medié v al e 2, É tudA l ex 2 , 127-167. Le Caire. e osson, , 1935, Mareotis: B eing a Short A ccount of the H istory and A ncient Monuments of the N orthW estern D esert of Egypt and of L ake Mareotis. London. res her, , 1946, A pa Menas. A Sel ection of Coptic T ex ts R el ating to St Menas. Cairo. res her, , 1949, o ogra hi al otes for le andria and District. B ul l etin de l a Socié té A rché ol ogique d’ A l ex andrie 38 13-20 ilmann, , angsdorff, , tier, , 1930, eri ht Ü ber die Voruntersuchungen auf den Kurûm El-Tiuwal bei Amrijr. Mitteil ungen des D eutschen A rchaol ogi schen I nstituts A bteil ung K airo 1: 106-129. El-Ashmawi, F., 1998, Pottery Kiln and Wine-Factory at rg el- ra n m ere r ed , Commerce et artisanat dans l ’ A l ex andrie hel l é nistique et romaine. A ctes du col l oque d’ A thè nes, organisé par l e CN R S, l e L aboratoire de cé ramol ogie de L yon et l ’ Ecol e franç aise d’ A thè nes,1 1- 12 dé cembre 198: 55-64 thens, Paris. l- akharani, , 1983, e ent a ation at area in g t n rimm, einen inter eds , A egyptiaca T rev ernsia I I , D as R öm isch- B yz antinische Ä gypten. A kten des international en Symposions 26.30. Se ptember 1 978 i n T rier, 175-186. Mainz. El-Falaki, Mahmoud Bey, 1872, Memoire sur l ’ antique A l ex andrie, s es faubourgs, s es env irons. Copenhagen. m ere r, - , i on, , 1998, es ateliers d amhores d a ario t n m ere r ed , Commerce et artisanat dans l ’ A l ex andrie hel l é nistique et romaine. A ctes du col l oque d’ A thè nes, organisé par l e CN R S, l e L aboratoire de cé ramol ogie de L yon et l ’ Ecol e franç aise d’ A thè nes, 1 1- 1 2 dé cembre 198: 75-91. Athens, Paris.

same pamphlet, the author presents important archaeologial e iden e, re io sl lished rossmann 1993 The pamphlet informs us that: he most interesting str t re in the town is a asili a situated on an eminence overlooking the lake shore near the longest of the harbour piers… The greatest surprise however was concealed under the apse. About 1.80 m below the s r i ing to s of the a se walls, the floor of the firing ham er of a great kiln for the rod tion of amphorae was discovered. The part of the church where liturgical ceremonies were held had been founded on these manufacturing remains. The kiln, which is one of only a few known from the Delta so far, turned out to be 8 m in diameter and the firing ham er floor was a 50 m thi k nside it, fired am horae of the 2nd-3rd century were discovered next to clay stacking rings that had separated the essels d ring firing

Because no further evidence was presented, I feel obliged to on l de that, firstl , this kiln alone ro es that the hill is to a great e tent artifi ial, e a se it was formed by industrial refuse produced by the ruins of a vast potter worksho t that se ondl , these r ins sho ld e understood as part of an extensive rural establishment of Early Roman type, comparable to several groups of similar remains hitherto n o ered aro nd the lake od iewi 2002 12, ig 3 d-el atah 1998 l- shmawi 1998 d l- i egem 1998 ne site dis o ered in the 1980s by the Marsa Matruh Inspectorate of the Egyptian Antiquities Organization, was located on the southern shore of this same Lake Mariout, near modern Borg el-Ara illage l- shmawi 1998 10 Thus, the kiln discovered nder the h r h in area ig 4 13 elongs to a hain of Early Roman rural enterprises located all along the southern shore of the Mareotis Lake and has nothing to do with the great city of Marea. Therefore, we should retain the view that the Byzantine port of Philoxenite, created for pilgrims travelling to Abu Mina, which was built not upon the r an r ins of area an ient a ital of areotis as des ri ed erodot s 149 t on the r ins of a nameless, extensive Early Roman rural estate, similar to many others in the region, a number of which have been located along the shores of Lake Mariout.11

10. El-Ashmawi 1998. I had the privilege to work there in 1987/88 as an adviser in the protective works of this large Early Roman rural estate, situated on Lake Mariout. 11 reliminar re orts on e a ations in area are s ar e see P ol ish A rchaeol ogy in the Mediterranean XII-XVII, 2001-2007, Warsaw ni ersit ress he e hi ition ill strating res lts of the olish e avations in Marea presented in the Bibliotheca Alexandrina, December 2006, was supplied with texts describing photos. A pamphlet was also presented. Both ignored all previous extensive research and restoration contributions to the site done by individuals and institutions. Reconstr tion done at the site in the 1980s shown there on ig anels that was carried out by the Egyptian Antiquities Organization was not properly described and the suggestion was that the works had been done re entl the olish ission ee manska a ra 2007, with references to previous reports.

73

LAKE MAREOTIS: R

PAST

Forster, E.M., 1922, A l ex andria: A H istory and a G uide 5th edn.). New Y ork. Fraser, P. M., 1972, P tol emaic A l ex andria, Vol. I. Oxford. rossmann, , os i k, , e er n, , e er n, , 1984, orl eri ht Mitteil ungen des D eutschen A rchaol ogischen I nstituts A bteil ung K airo 40 123-151 rossmann, , arit , , omer, , 1982, orl eri ht Mitteil ungen des D eutschen A rchaol ogischen I nstituts A bteil ung K airo 38 131-154 Grossmann, P., 1986, A bu Mina. A G uide to the A ncient P il grimage Center. Cairo. rossmann, , 1993, ie ers hiff asilika on a wariya und die Bauten dieses Typus in Ä gypten als e r sentanten der erlorenen r h hristli hen Architektur Alexandreias. B ul l etin de l a Socié té A rché ol ogique d’ A l ex andrie 45 107-121 rossmann, , rnold, , os i k, , 1997, e ort on the Excavations at Abu Mina in Spring 1996. B ul l etin a l a Societe d´ A rcheol ogie Copte 36 83-98 Grossmann, P., 2002, Christl iche A rchitektur in Ä gypten. Leiden. rossmann, , 2003, o hmals area nd hiloxenite. B ul l etin a l a Societe d´ A rcheol ogie Copte 42 13-20 Herodotus, T he H istories de lin o rt transl , 1971 Middlesex. asser, , 1983, Surv ey A rcheol ogique des K el l ia ( B asseEgypte) . R apport de l a Campagne 198 1 , Vol. I-II. Louvain. Kasser, R., 1986, L e site monastique de K el l ia. Sources historiques et ex pl orations archeol oqigues. Geneva. Kaufmann, C.M., 1906, D ie A usgrabung der MenasH eil igtüm er in der Mareotisw üs te. Cairo. Kaufmann, C.M., 1908, L a D é couv erte des Sanctuares de Mé nas dans l e D é sert de Maré otis. Cairo. a fmann, , 1924, D ie H eil ige Stadt der W üs te. U nsere Entdeckungen, G rabungen und F unde in der al tchristrl ichen Menasstadt. Kempten. ller- iener, , 1967, iedl ngsformen in der areotis. A rchäol ogischer A nz eiger 82 2 103-117 etr so, , a el, , 1983, area antine ort on Egypt´ s Northwestern Frontier. A rchaeol ogy, Sept/ t 62-63, 76-77 Röde r, J., 1967, Antiken Steinbrüc he der Mareotis. A rchäol ogischer A nz eiger 82 2 118-131

Rodziewicz, M.D., 1976, L a cé ramique romaine tardiv e d´ A l ex andria, A l ex andrie I. Varsovie. od iewi , , 1983, le andria and istri t of areotis. G reco- A rabica 2: 199-216. Rodziewicz, M.D., 1988a, Remarks on the Domestic and Monastic Architecture in Alexandria and Surroundings n an den rink ed , T he A rchaeol ogy of the N il e D el ta, P roceedings of the Seminar hel d in Cairo, 19- 2 O ctober 1986 : 267-276. Amsterdam. Rodziewicz, M.D., 1988b, Remarks on Peristyle House in Alexandria and Mareotis. P raktika 1983 : 175-178. Rodziewicz, M.D., 1990, Taenia and Mareotis, Archaeological Research West of Alexandria. A cta of the F irst I nternational Col l oquium of the Egyptian Society of G reek and R oman Studies I: 62-81. Cairo. Rodziewicz, M.D.,1995, Eco-Archaeology of Ancient Alexandria and Mareotis. In A.A. Hussein, M. Miele iad eds , P roceedings of the Seminar on G eosciences and A rchaeol ogy in Mediterranean Countries, Cairo, N ov ember 28- 30/ 193 127-139 airo od iewi , , 1998a, lassifi ation of ineries from areotis n m ere r ed , Commerce et artisanat dans l ’ A l ex andrie hel l é nistique et romaine. A ctes du col l oque d’ A thè nes, organisé par l e CN R S, l e L aboratoire de cé ramol ogie de L yon et l ’ Ecol e franç aise d’ A thè nes,1 1- 12 dé cembre 198: 27-36 thens, Paris. Rodziewicz, M.D., 1998b, From Alexandria to the West by and and aterwa s n m ere r ed , Commerce et artisanat dans l ’ A l ex andrie hel l é nistique et romaine. A ctes du col l oque d’ A thè nes, organisé par l e CN R S, l e L aboratoire de cé ramol ogie de L yon et l ’ Ecol e franç aise d’ A thè nes, 1 1- 1 2 dé cembre 198: 93-103 thens, aris Rodziewicz, M.D., 2002, Mareotic Harbours. In C. Décoert ed , A l ex andrie Mé dié v al e 2, É tudA l ex 8: 1-22. od iewi , , 2003, hilo enit ilgrimage ar o r of Abu Mina. B ul l etin de l a Socié té A rché ol ogique d’ A l ex andrie 47 27-47 Sadek, M., 1978, The ancient port of Marea. Cahiers des É tudes A nciennes VIII: 67. ma ska, , a ra , , 2007, area i th eason of Excavations. P ol ish A rchaeol ogy in the Mediterranean XVII. Reports 2005: 55-66.

74

K. BABRAJ & H. SZ Y MANSKA: MAREA OR PHILOXENITE?

Marea or Philoxenite? Polish Excavations in the areoti egion 2000-2007 r s tof a ra anna ma ska

ar o r installations of all kinds an e identified all along the south-western shores of Lake Mareotis. The most imosing r ins of antine date lo ated st 45 km so thwest of le andria, e tend along the oast for 1 5 km and are rrentl nder e loration a olish mission ig 1

gion remained the most im ortant agri lt ral rod er in northwestern g t aas 2001 47 he lake was an important communication route in Ptolemaic and Roman times. Goods were transported down the channels from inland, reloaded in Marea and sent to Alexandria, from where the were shi ed to other arts of the editerranean he information ontained in the eri l s of se dola esanges 1978 404-405 , dated res ma l to the earl , indi ates that ake areotis ma ha e 4th ent r een na iga le in haraoni times and that the water was ota le tra o , 1, 7 re orts that the lake ort in area handled more goods than e er ame to le andria the sea ro te o otte har et 1997 83 o th of le andria, there was a har o r, no longer isi le, alled P ortus Mareoticus it handled the goods that were shi ed ia the lake rom there the were trans orted down a anal named i otos thro gh le andria to the editerranean ne of the anals linking areotis with the so- alled ano i ran h of the ile od iewi 1983, 1998 101 he fertile areotis region was ele rated in anti it for its ine ards, oli e gro es, fr it or hards and a r s lantations t rod ed food for the a ital it and ri h landowners had their estates here. In the mid-7th ent r it was still a la e for g tians to ome to a ease h nger H istory of the P atriarchs 1 14, 501 n ient writers s oke warml of the alit of the lo al wine e orted to ome irgil, G eorg. , 91 tra o, , 1, 14 Horace, O des , 37, 14 thenae s, D eipn. , 33 en toda one of the most renowned g tian gra e ine growers and wine producers is based on the southern shores of the lake lmost 30 f rna e sites for firing am horae from the earl ellenisti to late anti it ha e een identified on the so thern lake shore m ere r i on 1998 75-91 El-Ashmawi 1990 55-64 n 2003, the olish e edition nearthed eneath the asili a at area, a otter kiln for firing am horae of 2nd-3rd ent ries date see elow his e iden e, when onsidered together with the n mero s remains of glass worksho s in the region, gi es an idea of the crafts that were an important part of the economic life of the region har k 2008 129-143

n 2000, e a ations at the site were ndertaken a olish team working nder the a s i es of the olish entre of editerranean r haeolog of arsaw ni ersit and the r haeologi al se m in rak w he li ense granted the reme o n il of nti ities of g t, o ers an area of 19 ha h sometri grid was t in la e and a s rfa e s r e of ar hite t ral remains arried o t 16 o e ts were fo nd ma ska a ra 2008 ig 1 n the o rse of the following eight field am aigns, the team has f ll leared a antine ath of the 5th to 7th ent r onne ted with a well that s lied it with water, and a f nerar ha el he resent fo s is on the e loration of a h ge hristian asili a standing on the romontor a o e the ort H istory of the S ite he site was identified as an ient area αρεία ahm d e l- alaki, o rt astronomer to g t s i ero smail asha l- alaki 1872 96 ees 1930 16761678 m linea 1973 241-243 alderini 1980 233 oma 1980 647 , t toda this identifi ation is eing reevaluated as indicated in the discussion below. The extensi e town r ins sit ated on ake ar t an ient areotis , 45 km from le andria, near the illage of a waria, ha e fas inated resear hers for a long time e osson re ared one of the first ma s of the oast resenting the lo ation of mainl ar hite t ral remains e osson 1935 131 he histor of this im osing har o r town with fo r grand etties for shi s to do k and ort asins to rote t against wind and wa es, an e tra ed in the an ient so r es erodot s , 30, 2 re orted that in the times of sammeti h of the 26th nast , area was a strategi osition garrisoned troo s defending the order with i a General Amasis defeated the armies of Pharaoh Apries at area and took the throne in 570 iodor s i l s , 68 nder ersian r le, the town was allegedl the a ital of an inde endent i an- g tian kingdom e tending from the ano i ran h of the ile to renai a ts r ler the name of naros, hallenged the ersians and lost his life after the fall of em his in 454 h dides , 104 inni ki 2006 135-142

The channels that fed the lake silted up in the 8th and 9th centuries due to lack of maintenance, and the fact that fresh water from the ile, mainl the ano i ran h, was t off he lake egan to dr od iewi 2002 9 leading to the de line of towns arti larl along its so thern shores n 1801, the lake ed was filled with sea water after the nglish o ened the sl i es in order to t off the a oleoni arm from fresh water he resent s rfa e of the lake is 90 km2, and its de th a erages at 1 50 m l e amses 2005 7

Marea never lost its importance as a commercial harbour des ite the fo nding of le andria, and the areotis re75

LAKE MAREOTIS: R

F ig. 1. R K EY 1. 1a. 2. 3. 4. 5.

PAST

emains of structures al ong the southern shore of L ake Mareotis

T hol os baths Stone pier 41 m l ong Stone pier 1 1 m l ong Stone pier 125 m l ong H arbour facil ities i i s presumabl y dry dock uncov ered by P etruso

6. 7. 8. 9. 10.

s h i i structure ya i a hs ish a ai s W el l operated w ith saqiyah ( P ol ish a ai s D oubl e baths uncov ered by F . E l - F akharani D ecumanus al ong the shops cl eared by F . E l - F akharani

a

y

i 1 1. 12.

16.

13. 14.

15.

k

y ki a

as

yk

R emains of saqiya G ranary uncov ered by F . E l - F akharani asi i a ish a ai s F unerary chapel ( P ol ish a ai s T ombs carv ed in the rock R emains of the ancient road l eading to the isl and

time of antine domination the region had lost its leading role he town fell d e to la k of water as dis ssed above, but the decline was also due to political and social fa tors areotis was o err n semi-nomadi edo in tri es who l ndered whate er stood in their wa oert 2002 t the eginning of the 8th ent r , in the r le of the Abbasids which coincides with the period when all life seems to have disappeared from the site, a new wave of ra nomads a eared in the region hese eo le were oliti all a ti e and the in reasingl took ontrol of loal h r h instit tions o ert 2002 137 io sl , the atmos here of the la e was no longer fa o ra le to

areotis was hristianised e ent all t the relations with the atriar hate in le andria assed thro gh different hases imm 1988 1593-1603 he first e iden e of hristianit is linked to wa es of re ression, the effe t of whi h res lted in adherents of the new religion eing anished to the region n 538, the ower of stinian s edi t, the entire areotis region, re io sl art of the Roman province of A egyptus P rima, was incorporated into i a Edict 13 1, 9, 17-22 n the eriod immediatel rior to the on est of g t the ra s in 641, areotis was a antine ro in e he infre ent referen e to areotis in o ti writings is roof that the 76

K. BABRAJ & H. SZ Y MANSKA: MAREA OR PHILOXENITE? hristians, arti larl as a sto o er for ilgrims he wine from Mareotis, however, still retained its reputation thro gh to the 7th ent r at least this is what we are led to elie e from a stor a o t the a stinen e of monks ited ohn os h s 1946

diffi lt in iew of ontin ed ind striali ation and n mero s to rist illa om le es eing ilt along the shore emains of har o r fa ilities an e seen along oth the southern shore and the northern shores of the lake. The iggest of these is area he a at the waterfront in the eastern end of the site preserves stone blocks that were used in its construction. Beside it there are four almost wholl reser ed etties where the shi s do ked hese etties formed the eastern, entral and western har o rs he were ilt of se eral o rses of stone lo ks of different si e three or fo r o rses ha e still een reser ed and can be studied thanks to the overall drop of the level of water in the lake he dimensions of the etties are o nting from the west to east 41 m, 6 50 m, lo ks a eraging 0 50 m 1 35 m 111 m, 5m he longest ett is lo ated the romontor with the r ins of the asili a, 125 m, 7 m, lo ks a eraging 0 65 m 1 70 m 0 35 m res ma l onl the art on land has s r i ed , 4 m fo nd ne t to the island that was onne ted in anti it with the mainland an artifi ial a sewa

owe er, identifi ation of earl har o r remains at area was estioned mainl on the strength of field s r e res lts, whi h re ealed nothing earlier than 6th to earl 8th ent r otter he ar hite t ral remains also ro ed to e essentiall antine in nat re raser 1972 146 n these gro nds od iewi 1983 202-204 ro osed to identif the r ins re orded on the s rfa e with a town established as a transfer point on the southern shores of the lake for ilgrims on their wa to the san t ar of inas some 20 km awa he o ti Encomium of St. Menas written the atriar h oannes 775-789 , mentions n mero s on enien es like hos i es the lake and rest-ho ses for those wishing to rest res her 1946 147148 , a market la e, orti oes and e en a fa ilit that we wo ld refer toda as a left-l ggage offi e his im ressi e rest sto was said to ha e een fo nded hilo enos, refe t of the m eror nastasi s 491-518 , hen e the town s name of hilo enite i s ka 2002 has identified this offi ial as the ons l hilo enos oteri hos, who held offi e in 525 and who is known also from ons lar di t hs artindale 1980 879-880

M area or Philoxe nite? he si e of the har o r installations, whi h are rrentl eing resear hed a reek- g tian team of nderwater ar haeologists headed ikos ianos from the elleni o iet for the t d and reser ation of arine lt ral eritage and the e artment of nderwater r haeolog in g t when considered in connection with the considera le si e of the asili a on the lakefront and the inn mero s t on l si e e iden e of arl oman otsherds, indi ates that a large r an agglomeration flo rished on the s ot alread in reantine times either an the e tensi e sewage network e onsidered as elonging to a town of little onse en e hilo enos ma ha e fo nded his it ad a ent to e isting har o r installations that had formerl een art of area, efore it was a andoned as an r an entre rior to the silting of the fresh-water channels that supplied the lake, but there is little doubt that settlement e isted here rior to its fo ndation lread in thenae s 33d wrote of Marea as the the 2nd ent r so r e of a wine of fine alit he alled area a illage thanasios 85 ol 400 - had the same im ression of the town 200 ears later rossmann 2003 16 thinks that the ellenisti ar hite t re of area was of m d- ri k whi h is easil disintegrated t sho ld e noted that at the margins of the settlement m d- ri k is a ilding material sed in the lower la ers of the asili a, hronologi all ontem oraneo s with the otter kiln of the 2nd-3rd ent ries

L ocation of the S ite he geogra hi al oordinates of the site are 30o 59 32 85 and 29o 38 58 34 to 30o59 48 84 and o 29 39 28 32 he e a ation area meas res 19 7 ha o the west it is limited a do le-lane road linking the waterfront with hak sh settlement the lake is a nat ral o ndar on the north and east, and to the so th there is the desert which extends as far as Shakush settlement. The o erall height differen e does not e eed 8 m an where on the site he site grid onsists of 42 geodeti oints, mostl di tated the nat ral to ogra h he oordinates were ass med lo all , while the heights are listed the g tian state geodeti ser i e he to ogra hi al lan of the site so far onl o ers the area e a ated in 2000-2003 see ig 1 rom a geologi al oint of iew, area lies in a road stret h of land etween allahet ar t, the long westward arm of ake areotis, and a arallel ridge of oolithi limestone he soils on either side of the ridge are mostl al arenite, er fa ora le for lti ation with ro er watering arne tanle 1993 l- akharani 1983 175 od iewi 1995 ielska- owgia o oronko 2008 17-18 similar ridge lies on the o osite side of the lake, se arating ake areotis from the sea his is the aenia αι ία idge whi h e tends along the oast from le andria all the wa to ina and i a oda these ridges are mostl destro ed as a res lt of hea e loitation of limestone arries n the 4th and 5th cent r , hristian monasteries a eared in the aenia the were called after the mile-stones that served to measure the distan e from le andria heir lo alisation toda is er

There is however, opposition to the idea that this was the site of Philoxenite. The distance from Philoxenite to Abu ina is one area to estion ording to res her s 1946 translation of the Encomium, the distance amounted to a o t 30-35 km he te t also mentions rest sto s 77

LAKE MAREOTIS: R

PAST

where water was a aila le, whi h the refe t had arranged e er 10 miles or so 14 80 km etween the two towns owe er, the distan e in realit is some 20 km, not eno gh for e en two water sto s et the ke arg ment against the identifi ation with hilo enite is rossmann s resear h on the fre en of ilgrimages to ina ilgrimages to the san t ar did not eak ntil the se ond half of the 5th ent r , and afterwards s rges in the n m er of ilgrimages a om anied the feast da s of t enas rossmann 2003 18 he r ins rrentl eing in estigated on the site are onsidered earlier the do le aths e lored Elakharani 1983 see oth aggag s a ers in this ol me originate from the end of the 4th ent r and the otter kiln under the apse of the basilica operated in the 2nd-3rd ent ries he si e of this latter fa ilit is nmistaka le roof of a flo rishing entre e isting on the s ot ossi l area rior to the antine fo ndations, erha s rod ing the wine that areotis was famo s for and e editing it to other towns of the re o- oman world in amphorae produced at the site. One of the sherds from the fill of the otter kiln ore the stam of a otential owner of the ine ard ιο υσίου ma ska a ra 2005 130, n 41 an artifa ts dis o ered in the neigh orhood of area onfirm the resen e of otter kilns for firing am horae in oman times m ere r i on 1998

lo k- oining te hni e that went o t of se in tal in the 1st ent r , e identl oint to an earlier origin dam 2008 57 oreo er, the a sed f nerar ha el, dis o ered the Polish mission, with its three chambers intended for an affl ent famil , o ld not ha e een ere ted and for ilgrims transiting thro gh the town n the other hand, there an e no do t that ilgrimages to the tem le of the hol mart r m st ha e assed thro gh this area, if an thing ea se of the lose ro imit to the lake ore e iden e is ro ided a ate oman v il l a rustica, uncovered in a wari a illage, transformed into a dormitor in the middle of the 6th ent r and f rnished with a small h r h inside it od iewi 1988 271-273, ig 2 large r an centre with developed harbour facilities must have been erfe tl s ited as a sto o er for wa worn and s rel often ailing ilgrims o ert 2002 129 identifies an ient area with the lo alit of kin ar t to the east of the site, where e a ations are rrentl eing ond ted et the onl arg ments in fa o r of this theor is the strategi osition of this lo alit , affording ontrol of the road from renai a and g t to le andria and eing a essi le also from the opposite side of the lake thanks to a considerable narrowing of its width at this oint

he great etties are of ke im ortan e for the dating of the ort he str t ral ariet of ett onstr tion different si e lo ks s ggests se eral ilding stages, t the hara teristi im rints left swallow-tail metal lam s, a

i

78

a a

ya i

a hs

a

y

aaa

K. BABRAJ & H. SZ Y MANSKA: MAREA OR PHILOXENITE? Baths ma ska a ra 2008a 27-83 he freestanding ilding follows an orthogonal lan, enom assing the aths itself ig 2 and the well, whi h is furnished with a sāqiyah installation drawing water for the aths he a t al ilding of the aths, o ered an area of 642 m2, in l ding two o rt ards and was en ir led a wall of stone lo ks he aths itself was ilt entirel of red ri k on a ri k fo ndation ig 3 ho s and a latrine lined the north wall on the o tside olonnaded o rt ards were located to the east and west.

he ndergro nd art of the ilding in l ded the ser i e area ellars, two heating f rna es and one for heating water in a oiler, finall , there were fo r h o a st ellars two in the men s art and two in the women s art he interior de oration of the ilding onsisted of mar le a ements, ol mns with orinthian a itals, and a few oatings of ainted wall laster feat ring a olorf l egetal frie e he floors and sele ted walls of ools were fa ed with mar le sla s his was e identl a l rio s om le and definitel not the onl one in town to dge the solid water hannel s stem to whi h it was onne ted

he ilding was a do le ath, onsisting of two ne en arts se arated a ri k wall the so thern one, for men, igger than the northern one for women a h of these arts was heated a se arate f rna e s l ing warm air to fo r h o a st s stems he entran e for men was in the western fa ade it led from the o rt ard, following widel a e ted ir lation rin i les, that is, dire tl from the apodyterium into the tepidarium, and then into the two cal daria he women s art was not as e tensi e, a essi le from a rather s a io s o rt ard to the east side It consisted of three rooms: apodyterium and two cal daria. o rteen small ools were sed for athing, eight of whi h were lo ated inside the two arts of the aths, the o ter wall of the caldaria. Two of them were accessible from the o rt ard, leading to the men s art o r ools were lo ated o tside the aths, s mmetri all in the two o rt ards, a tting the walls he ools were differentiated in sha e, from semi ir lar thro gh re tang lar to ro nd, de ending on the re ilding hase al lations made on the asis of si fragments of ri k ar hes fo nd in the de ris ermit a reconstruction of the width of baths halls, which were res ma l o ered with arrel a lts he h otheti al height of the ilding, deri ed from the ro ortions of the s r i ing walls, was 3 50 m

owhere in the ilding was a lear stratigra h of the construction and occupation of the baths visible and the fa t that the ilding had olla sed v ide brick arches from the a lts la immediatel on the floors re l des reise dating n anal sis of the otter finds has ro ided a timeframe for the operation of the complex between the first half of the 6th ent r and the first 30 ears of the 8th ent r , that is, efore the a earan e of slami arl ead la ed wares, of whi h not ns r risingl , not one sherd has een fo nd at the site a herek 2008 ra oins alread iss ed after the monetar reform, that is, after 686, and remaining in ir lation for 750 ears, onfirm the ontin ed o eration of the aths after the damages in rred d ring the ersian o ation and the ra in asion alar k 2008 he most intensi e se of the aths o rred in the first half of the 7th ent r , a fa t onfirmed oth the otter finds and the re alen e of oins iss ed hosroes i ho ka 2008 Sāqiyah ma ska a ra 2008 85-99 well o erated a sāqiyah stood about 5 m north of the aths he well itself was 5 m dee , re tang lar in lan and ilt of stone lo ks igs 3 4 he water in the

F ig. 3. M area. B aths and w estern courtyard. V iew from the south- w est ( photo s k

79

LAKE MAREOTIS: R

PAST F ig. 4. M area. W el l operated w ith s i ah V iew from the south h y aa s

it for olle ting water in ase of dro ght n to of this installation was a treading ir le for the animals t rning the sāqiyah wheel er sim le h dra li s stem ased on onne ted essels was em lo ed ater was olle ted in the reservoir of the sāqiyah and carried to the baths via a s stem of terra otta i es

well ame from a still f n tioning ndergro nd s ring in its northwestern corner. The mineral content of this water is much more abundant compared to the lake water. South of the well and le el with its rim, there was the floor of a reser oir made of fired ri ks in water roof mortar he so th side of this asin, whi h was filled with water from the well, was reinforced with a solid platform built of stone blocks. A capacious cistern closed off the complex of the sāqiyah to the north t was res ma l intended as a fa il-

nl one ond it rea hing asin in the aths so th of the sāqiyah see ig 3 has een reser ed the west wall F ig. 5. M area. F unerary chapel . V iew from the w est h s k

80

K. BABRAJ & H. SZ Y MANSKA: MAREA OR PHILOXENITE? F ig. 6. M area. P l an of the basil ica ( draw n by aaa

of the aths see lan in ma ska a ra 2008 ig 1 he i e, a fragment of whi h is isi le nder the last block of the sāqiyah reservoir with two blocks of the limestone edding, ran to the north wall of the aths he i e t rned into oom 1, whi h it rossed ntil it rea hed the so th wall, where it was raised 1 50 m in order to em t into asin here is e er reason to elie e that there were more sāqiyahs o erating in the town om le of area gro nd s r e of the site has identified the remains of at least one other fa ilit of the kind in the area to the west of the aths ma ska a ra 2008 88

Basilica ma ska a ra 2006 107-117 he most interesting ilding at the site, howe er, is a asili a sit ated on the hill near the longest har o r ett t was dis o ered ller- iener 1967 106, n 16 , former dire tor of e a ations at the near san t ar of ina t it was rossmann who first determined the lan and dimensions of the ilding in 1986 rossmann 1993 107-121 r risingl , none of the an ient written so r es mention this h r h, whi h was one of largest ildings of its t e et known in g t see rossmann 2002, assim

F unerary chapel ma ska a ra 2008 177-185 One hundred meters to the south there were the ruins of a ilding whi h were identified as a f nerar ha el t was f rnished with an east-oriented a se and three masonr gra e ham ers ontaining 23 rials asso iated with er oor gra e goods ig 5 he ontained oth genders of different age, all learl of famil hara ter t f n tioned for a little o er a h ndred ears and onse entl m st ha e elonged to the town s inha itants a a am hora nder the floor of the a se erified a 6th ent r date for the construction of this complex.

t was a s at-sha ed ilding, di ided ol mns into three aisles, f rnished with a wide transe t terminating in ro nded arms and a relati el tin a se ma ska a ra 2004 53-56, 61-64, 2005, 63-67, 2005a, 43-54 he lit rgi al rooms were not on the o tside, as was the r le at other asili as, t in or orated into the od of the ilding on either side of the a se ig 6 see rossmann 2002, assim he a tister with ro nd a tismal font a arentl remained from an earlier ilding, ossil a small ha el, whi h re eded the great asili a at the site. 81

LAKE MAREOTIS: R

PAST F ig. 7. M area. A pse of the basil ica. V iew from the w est. ‘ A ’ and ‘ b’ l ocate the tw o burial s h y aa s

kind of mone e hange ilding intended res ma l for ilgrims isiting the asili a, was dis o ered e ond the so thwestern orner of the ilding are e am les of ron e weights sed to erif oin weight were fo nd inside this room ma ska a ra 2005a 54

rated ol mns of ario s si es, all of them in ro onesian mar le and ro a l im orted ia le andria he interior decoration of the basilica also included pavement mosaics of whi h small mar le es fo nd in the ilding are the onl s r i ing e iden e

Two burial chambers with multiple burials were discovered nder the floor of the a se ig 7, a nthro ologi al e amination identified o er 100 indi id als men, women, hildren and e en n orn hildren he a ear to ha e een ried there d ring the in asion of hosroes in 619 when ersian troo s tor hed le andria and ra aged the region

Amphora K iln ma ska a ra 2004a 53-56, 6164, 2004 26-28, 2005 63-67 om lete s r rise awaited ar haeologists digging nder the h r h a se t a de th of 1 80 m elow the reser ed to s of the walls, a grate elonging to a large am hora kiln was s se entl dis o ered he art of the h r h intended for lit rgi al ra ti es sed this earlier kiln as a fo ndation ig 8 ore of the grate a eared on e the floor of the rial ham ers nder the a se had een cleaned.

The rich interior decoration of the basilica included several fragmentar ol mn shafts and orinthian a itals de o-

F ig. 8. M area. B asil ica. A mphora kil n. V iew from the w est ( photo by y

82

K. BABRAJ & H. SZ Y MANSKA: MAREA OR PHILOXENITE? Bibliography

The kiln had a diameter of 8 m, and the thickness of the grate was 0 50 m ne of onl a few dis o ered in the ile elta, this kiln still ontained on dis o er the last batch of amphorae. These vessels were dated to the 2nd and 3rd ent ries he wall of the kiln still stood 0 93 m high and was ilt of m d- ri ks he str t re of the kiln is not unlike others from the period with pillars under the grate ier ed with holes to s ort eas ir lation of hot air l- shmawi 1998 55-64

A ncient Sources Athenaeus, T he D eipnosophists, h li k transl , 1927 am ridge, Athanasios, A pol ogia contra A rianos, it transl , 1934 erlin, ei ig iodor s i l s , L ibrary of H istory, ldfather transl , 1989 am ridge, s B ook I I , lo d transl , 1976 eiden H istory of the P atriarchs of the Coptic Church of A l ex a ia atrologia rientalis 1 14, etts ed , 1904 aris Horace, sa s dd transl , 2004 amridge, John Moschus, P ratum Spiritual e = a nes, 1947, The Pratum Spirituale. O riental ia Cristiana P eriodoca 13 404-414 Justinian, i s is i i is s i ia s ol , r ger ed , 1877 erlin Strabo, G eography. In Strabon. L e v oyage en Egypte. U n a ai o otte har et transl , 1997 aris h dides, H istory of the P el oponnesian W ar, h orster mith transl , 1928 am ridge, irgil, Ecl ogues. G eorgics, shton air lo gh transl , 1999 am ridge,

Conclusions he site of area with its ate nti e agglomeration remains a le, generating onsidera le ontro ers as to its hronolog and e en its er name he identifi ation of the town r ins l ing on the so thern shore of ake areotis, st 45 km so th-west of le andria, is amongst the riorities of a team from the olish entre of editerranean r haeolog of arsaw ni ersit and the r haeologi al se m in rak w, whi h ha e ond ted e aations here sin e 2000, ased on a on ession granted the reme o n il of nti ities of g t that o ers an area of 19 7 ha The Polish expedition has concentrated on the extant ruins, in estigating the role and im ortan e of this antine town as a religio s enter ilt aro nd a h ge asili a ring nine field seasons 2000-2008 , three inde endent ar hite t ral om le es ha e een in estigated the aths with sāqiyah-well, a f nerar ha el, and the hristian asili a ll the str t res were dated to the 6th end of 5th thro gh earl 8th ent ries site s r e in l ded an inentor of all the r ins dis erni le on gro nd le el within the limits of the on ession see ig 1

Secondary Sources Adam, J.P., 2008, L a construction romaine. Paris. m linea , , 1893, re rint 1973 , G é ographie de l ’ É gypte à l ’ é poque copte sna r k l- shmawi, , 1998, otter iln and ine- a tor at rg el ra n m ere r ed , Commerce et artisanat dans l ’ A l ex andrie hel l é nistique et romaine. s q h s a is a L aboratoire de cé ramol ogie de L yon et l ’ Ecol e franais h s 55-64 thens, Paris. a ra , , ma ska, , 2004, ie ierte ra ngssaison in area, eg ten ra ka elle nd asilike K emet 13 3 61-64 a ra , , ma ska, , 2005, nfte ra ngssaison in area, g ten asilika K emet 14 3 63-67 a ra , , ma ska, , 2008, āqiyah n ma ska a ra eds , B yz antine Marea. Ex cav ations in 20203 and 206 , a a ol 1 85-99 rak w a ra , , ma ska, , 2008, F unerary chapel . In ma ska a ra eds , B yz antine Marea. Ex cav ations in 20 - 203 and 206 , a a ol 1 177-185 rak w Blue, L., & Ramses, S., 2005, Lake Mareotis Research ro e t reliminar e ort from ilot r e g st 2004 N ew sl etter of the A rchaeol ogical Society of A l ex a ia a y 5-16 alderini, , 1980, i i a i i i a i a i i a ol , fas 3 Milano.

his large ort o erated in oman and later antine times and perhaps also in the Ptolemaic era. However, the estion of the a t al identit of the site remains still n ertain he rrent e a ated r ins of antine date e tend along the oast for 1 5 km he great etties are of ke im ortan e for the dating of the ort he si e of the har o r installations, the small t im ortant antit of arl oman otsherds, as well as h ge sewage network of s stantial ild, indi ates that a large r an agglomeration flo rished on the s ot from reantine times hilo enos ma ha e fo nded his it rior to its a andonment as an r an enter d e to the silting of the sweet-water ile hannels s l ing the lake owe er, e iden e s ggests that a site e isted rior to the esta lishment of hilo enite, eing lo ated ne t to e isting har o r installations that ma ha e een art of area he re ent dis o er of the otter kiln nder the a se of the asili a that operated in the 2nd-3rd ent ries , is nmistaka le roof of a flo rishing enter e isting on the s ot ossi l area rior to the antine fo ndations erha s this site was were the wine that Mareotis was famous for was eing rod ed, and from here the rod t was eing e ported to other towns of the Greco-Roman world in amphorae produced at the site. 83

LAKE MAREOTIS: R

PAST

e osson, , 1935, a is i a h H istory and A ncient Monuments of the N orth- W est D esert of Egypt and L ake Mareotis. London o ert, h , 2002, ar otide m di ale es do ins et des hr tiens n h o ert ed , A l ex andrie Mé dié v al e 2, É tudes A l ex andrines 8 127-167 e aire esanges, , 1978, ri le d se dola 90 107 ri les li i es, 7 n esanges ed , R eh h ss a i i s i a sa s s de l ’ A frique, Col l ection de l ’ É col e F ranç aise de R ome 38 404-405 ome res her, ed , 1946, A pa Mena. A Sel ection of Coptic T ex ts rel ating to St. M enas airo m ere r, - , i on, , 1998, es ateliers d amhores d la ario t n m ere r ed , Commerce et artisanat dans l ’ A l ex andrie hel l é nisiq ai s q h s a is a a a i a i y a ais h s 75-91 thens, aris l- alaki, ahmo d e , 1872, Mé moire sur l ’ antique a i s s a s i s s a s i s s a s i a s h ches o enhagen. l- akharani, , 1983, e ent a ation at area in g t n rimm, einen inter eds , y ia a sia as is h y a i ische Ä gypten. A kten des international en Symposions 26.30. Se ptember 1978 i n T rier, 175-186 ain raser, , 1972, ai a ia I. Oxford. oma , , 1980, s om el- dris area , L ex icon der Ä egyptol ogie 647 rossmann, , 1993, ie ers hiff asilika on a war a- area nd die rigen a ten dieses s in g ten als e r sentanten der erlorenen fr hchristlichen Architektur Alexandreias. In N. Swelim ed , I n Memoriam D aoud A bdu D aoud ( B ul l etin de l a Societé A rché ol ogique d’ A l ex andrie 45 107-121 Grossmann, P., 2002, Christl iche A rchitektur in A egypten. Leiden. rossmann, , 2003, o hmals area nd hilo enite B ul l etin de l a Societé d’ A rché ol ogie Copte 42 13-29 aas, h , 2001, A l ex andria and the Mareotis R egion. In rns adie eds , U rban Centers and R ural Contex t in L ate A ntiquity 47-62 nn r or, ees, , 1930, s area, areotis R eal Encycl opedie 2 1676-1678 har k, , 2008, lass finds from the aths and sāqiya n ma ska a ra eds , B yz antine Marea. Ex cav ations in 20203 and 206 , a a 129-143 rak w i ho ka, , 2008, he earl antine oins n ma ska a ra eds , B yz antine Marea. Ex cav ations in 20203 and 206 , a a 145-152 rak w

a herek, , 2008, he otter assem lage from the baths and sāqiyah n ma ska a ra eds , B yz antine Marea. Ex cav ations in 20203 and 206 , a a 105-127 rak w alar k, , 2008, mma ad oins n ma ska a ra eds , B yz antine Marea. Ex cav ations in 20203 and 206 , a a 153-154 rak w artindale, ed , 1980, T he P rosopography of the L ater R oman Empire, am ridge ller- iener, , 1967, iedl ngsformen in der areotis. A rchäol ogischer A nz eiger 82 2 103-117 ielska- owgia o, , oronko, , 2008, ol tion of the nat ral en ironment in the region of area n ma ska a ra eds , B yz antine Marea.E x cav ations in 20203 and 206 , a a V ol . 1 17-26 rak w od iewi , , 1983, le andria and istri t of Mareotis. G raeco- A rabica II: 199-216 od iewi , , 1988, emarks on the omesti and onasti r hite t re in le andria and rro ndings n an den rink ed , T he A rchaeol ogy of the i a i s h i a h i ai 19- 2 O ctober 1986 267-276 msterdam od iewi , , 1995, o- r haeolog of n ient Alexandria and Mareotis. In A.A. Hussein, M. Miele & iad eds , P roceedings of the Seminar on G eosciences and A rchaeol ogy in the Mediterranean Countries, ai / 127-139 airo od iewi , , 1998, rom le andria to the est. In m ere r ed , Commerce et artisanat dans l ’ A l ex andrie hel l é nistique et romaine. A ctes du col q h s a is a a oratoire de cé ramol ogie de L yon et l ’ Ecol e franais h s 93-103 Athens, Paris. od iewi , , 2002, Mareotis H arbours n h o ert ed , A l ex andrie Mé diev al e 2, É tudes A l ex andrines 8 1-22 e aire od iewi , , 2003, hilo enit ilgrimage arbor of Abu Mina. B ul l etin de l a Societé A rché ol ogique d’ A l ex andrie 47 27-47 ma ska, , a ra , , 2004a, area o rth season of e a ations, 2003 P ol ish A rchaeol ogy in Mediterranean 53-63 ma ska, , a ra , , 2004 , he an ient ort of area, g t o r seasons of e a ations the olish r haeologi al ission Minerv a ril 2004 26-28 ma ska, , a ra , , 2005, o illes ar h ologies de area en g te aisons 2002-2003 A rcheol ogia 119-130 ma ska, , a ra , , 2005a, area ifth eason of a ations, 2004 P ol ish A rchaeol ogy in Mediterranean 43-54

84

K. BABRAJ & H. SZ Y MANSKA: MAREA OR PHILOXENITE? arne, , tanle , , 1993, ate aternar ol tion of the orthwest ile elta and d a ent oast in the le andria egion, g t J ournal of Coastal R esearch 9 1 26-64 inni ki, , 2006, er li s he tamm der akaler in haraonis hen, ersis hen nd tolem is hen g ten. A ncient Society 36 135-142 i s ka, , 2008, aq s s i i ai P il ox enité à l a l umiè re de fouil l es ré centes. Paper deli ered at the hird onferen e on edie al le andria at the entre lt rel fran ais in le andria, 8-10 o em er, 2002 n

ma ska, , a ra , , 2006, Polish Excavations in the asili a at area g t B ul l etin de l a Societé A rché ol ogique d’ A l ex andrie 45 107-117 ma ska, , a ra , , 2008, B aths n ma ska a ra eds , B yz antine Marea. Ex cav ations in 202 03 and 206 , a a 27-83 rak w ma ska, , a ra , eds , 2008, B yz antine Marea. Ex cav ations in 20203 & 206 , a a rak w imm, , 1984, D as christl ich- koptische Ä gypten in arabischer Z eit. ies aden imm, , 1988, eil 4 - , eihe eisteswissens haften o 41/4, ies aden

85

LAKE MAREOTIS: R

PAST

86

M.-F. BOUSSAC & M. EL AMOURI: LAKE STRUCTURES AT TAPOSIRIS

he ake tr t res at a osiris Marie-Franç oise Boussac & Mourad El Amouri

Introduction Among the harbours of Mareotis, which in recent years have been the subject of renewed attention, Taposiris on the north shore is of particular interest. Amongst many features of interest, the city offers the only example, besides that of Alexandria, of a closed basin which allows ontrolled maritime traffi management ig 1 hese str t res were riefl mentioned at the eginning of the 19th ent r a ho, who in 1824 de oted a few lines to the dam r nning from east to west ilt in the so th of the it , a str t re intended, he said, to re ent floods a ho 1827 and sket hed oste1 in 1820 ig 2 hereafter, re ia 1914 , e osson 1935 110-111 ig 3 , hsens hlager 1979, 1999 and od iewi 2002 , amongst others, ha e ointed o t the e liarities of the closed basin, speculating on its chronology and its connection with the wall of the Barbarians which closes, to the west of the city, the very narrow spit of land that separates the editerranean from ake areotis at this oint, and its relation to the a sewa whi h lo ked the lake to the south. At some point in their use, both systems acted as lo ks and seem related, t we do not know if the are art of the same feat re, nor do we know d ring whi h eriod they were used.

When the French Archaeological Mission of Taposiris3 was la n hed in 1998, it was determined that an nderstanding of the harbour system would be a research priority. What was the chronology of the currently visible structures? To which phases of the site did they correspond? e a se of the dimensions of the har o r asin the eastwest artifi ial le ee or a sewa is a o t 1,700 m long , areas that were deemed likel to ro ide some answers with limited resources were selected for excavation, supplemented with an environmental study and geophysical survey. Since most of the data acquired has already been the s e t of arti les o ssa 2007, 2009 , this a er will riefl resent the har o r str t res and the onclusions reached concerning the southern causeway. The east side of the system will be the main focus as this gives the latest ossi le date for all the st died lake onstr tions. Ove rvi ew of the Remains The ancient city of Taposiris, located on the south side of the taenia ridge, is organised into three se tors linked a network of north-so th ro tes the er town with the Brescia terrace and the temple, the middle town, and the lower town on the shores of ake areotis o the west, the Wall of the Barbarians forms the western limit of the it see ig 1

To answer these questions, the American mission from rookl n ollege arried o t se eral so ndings in 1975, during a one month campaign. Anxious to locate the harbour area,2 hsens hlager worked on two se tors, north and north-west of the dug-out channel and the causeway. either ga e the anti i ated res lts in the north one , an ele ated area, inter reted as a lake front, ro ed to e an a m lation of r le waste from am horae worksho s, as shown thereafter m ere r i on 1998 n the north-west one , meri an ar haeologists concentrated their efforts on what initially seemed to be a warehouse along the shore. They described it as a latform whi h had een redesigned se eral times the so-called pl atform buil ding , the did not nderstand its f n tion, t the dated it to the 3rd ent r he also n o ered a so histi ated s stem of water tanks arr ing water towards the harbour from a terrace further north. No research was published to specify the chronology, and no inter retation was s ggested n lished

The topographical survey of the lower town area carried o t in 1999 and 2000 ig 4 , and s lemented in 2009, extends from the Wall of the Barbarians in the west to Plinthine in the east, and highlights the irregular outline of the lake shore artiall anthro ogeni s it of land romontor di ides the lake shore into two marsh plains, one of which, to the east, forms a bay which is interpreted as a harbour basin.4 This basin covers an area of approximately 8 ha and is closed off to the south by a a sewa a ro imatel 1,700 m long , whi h e tends east-west in the shape of a ridge, but it is not straight as the sket hes of oste 1820 see ig 2 or e osson 1935 110 see ig 3 im l t is also interr ted at least si openings that may be deliberate or may be related to the disappearance of the retaining walls which in places supported the north slope of the causeway.

3 in e 1998 a ission ran aise in a osiris agna has een s ported by the Commission des fouil l es of the French Ministry of Foreign and ro ean ffairs thorisation has een kindl gi en and wee appreciate their assistance. 4 nal sis la , h ni i - arseille, in 2008, in a h thesis on the geomor holog of ake Mareotis.

1

hanks to the i lioth e ni i ale o ation gionale in arseille, es e iall o lle olom i , who allowed s to re rod e two sket hes as al oste from their ar hi es 2 hanks to the meri an mission, es e iall enit, who shared their n lished do mentation es P rel iminary R eport of the 1975 s eason

87

LAKE MAREOTIS: RECONSTRUCTING THE PAST

F ig. 1. G eneral map of the site and l ocation of areas under study at T aposiris Magna and P l inthine. © A rchaeol ogical Mission in T aposiris.

88

MA F T , F rench

M.-F. BOUSSAC & M. EL AMOURI: LAKE STRUCTURES AT TAPOSIRIS

F ig. 2. D raw ing by P ascal Coste, m ade in 1820. 2a: G eneral pl an of A l ex andria to A bousir. 2b: D etail of pl an show ing the T empl e of T aposiris ( 1) , t he T ow er of A rabs ( 2) , br idge ( 3) , e astern j etty ( 4) and causew ay across the l ake from the l ev ee to the south shore ( 5) . 2c : P l an and isometric v iew ( south) of the bridge of T aposiris. © B MV R F ig. 3. M ap of the remains of the ancient city of T aposiris by A . D e Cosson, 1935 .

89

LAKE MAREOTIS: RECONSTRUCTING THE PAST

F ig. 4. L ocation of the different sectors of the site of T aposiris that w ere studied al ong the l ake shore. © MA F T expands to the nortwest to form a loop against which is aligned by a series of buildings, whose different orientations and especially superposition indicate different phases his isi le state has signifi antl altered the re io s interpretation of the phases: the channel was dug in the first half of the 2nd ent r o ssa 2009 7 in a location where a densely occupied area was previously developed whose architectural design suggests prosperity. This area was used for storage and trade, judging from the quantity of imported amphorae material found there o ssa 2009 8 n the north shore e tor 1 there is a row of shops, abandoned at the end of the Hellenistic period n-re o ered items left in la e erha s as a res lt of flooding,9 before the digging of the channel. On the other side of the hannel e tor 3 , earth that had een dredged from the lake o ered ildings last o ied at the end of the 1st century BC and which had been in use since at least the 2nd century BC. The excavated building has the same

Topographic mapping and geophysical survey revealed the densit of str t res in the lower town, riefl mentioned as walls and r ins oste 1820 and ignored re ia 1914 5 ltho gh the a sewa that lo ked the lake has irt all disa eared onl a stret h of a o t 10 m o ld e n o ered , its o tlet was lo ated near a warehouse built on the causeway. Most importantly, the geophysical survey revealed in the West Bay and south of the bridge many ruins which are currently silted over.6 These results underscore the extent of changes in the landscape, the mobility of the shoreline and the extent of the harbour area. T he E lements of the L ak e S ystem T he Causew ay and the Channel e eral o erations 1998, 2000-2005 were ond ted at the west end of the causeway, on both sides of the channel and near the bridge, where the basin forms a bend: it

5 re ia 1914 e onl mentions the 1 km-long d ke, r nning arallel to the hills, and a fine ridge whose onstr tion ertainl dates to oman times t seems lear that the lake e tended to a osiris, and that the d ke lo ked the water in a kind of har o r 6 ros e ting ene h, ee the ma in o ssa 2007 452

7 8

ee the anal sis ar i in o ssa 2009 o ssa 2009 see the a endi ar i and the st d enol 9 t d s ggests d ring the 1st century BC: see the analysis by S. Mari in o ssa 2009 137

90

M.-F. BOUSSAC & M. EL AMOURI: LAKE STRUCTURES AT TAPOSIRIS ar hite t ral feat res as the north sho s fo ndations and ar hite toni elements in stone, m d- ri k walls o ered with a oating t was s lied with drinking water rainwater olle ted and dis harged into a r iform tank located to the north. The angle of the buildings on both sides of an east-west axis indicates the prior circulation of water in this area, which was later replaced by the channel. he le el of the tank indi ates that the site hosen to dig the hannel was originall an area not lia le to flooding

har o r traffi to ass, while minimi ing the si e of the h m a k ridge his wo ld artl e lain the as mmetr of the two ar hes, with a h m a k assage whi h est fits the to of the a lts, whi h are all of different widths and heights a ro imatel 4 5 m and 4 m Furthermore, the construction of the bridge can only be explained by the existence of the causeway and channel system limited in estigation 6 2 m 3 2 m ond ted in the fo ndation e le ed radier of the north ier of the ridge, to the flood le el of the lake, onfirms that the onstruction of the bridge is contemporary with the digging of the channel, however it does not allow precise dating. Indeed, the stratigraphic study of this area shows that the ridge was ilt o er a le el that dates a k to the m erial oman eriod it onfirms that the hannel, the ridge and the closed harbour system can be dated to the Imperial Roman period.

he whole area was disr ted ma or works after a eriod of stagnation and neglect. The objective was to create a navigation canal by digging a channel, and ensure acess strengthening and shoring the anks formed he channel was dug and the waste was deposited to the north and south creating two sedimentary causeways, of unequal length t of a similar onfig ration the a sewa to the north e tends for a o t 150 m, to the so th for 1,700 m

T he East P art of the L ake System he lake asin is losed to the east a ilt jetty, which was the s e t of two am aigns in 2005 and 2006, following a preliminary phase of topographical mapping in 1999 and 2000 he an ellation of the following two am aigns 2008 and 2009 , d e to an e e tional rise in the water level, meant that certain points remain to be determined, in particular the chronology: while the date of abandonment is well attested by the coins and ceramics first half of 7th ent r , n ertainties remain on erning the previous phases.

n anthro i em ankment was formed s essi e hea s of a kfill iled , sta ilised a gypsum screed. he most e tensi e works were arried o t to the west, where the edge of the romontor needed to e t a k his e lains wh there are onl two artifi ial em ankment slopes in this sector, to the north and the south, and why the causeway ends near the bridge at a height of some 5-6 m and then dro s in irreg lar ste s towards the east The two sides are asymmetrical: the north slope, which follows the dug-out channel is steeper and supported by a series of retaining walls to prevent the collapse of the earth and therefore seal the passage, the south side has a shallower slope.

he ett is lo ated a ro imatel 1,700 m to the east of the ridge west entran e of the har o r asin elow the Temple of Taposiris. Between the end of the causeway and the tip of the jetty, there is an opening which is currently a o t 100 m wide his ro a l allowed oats to enter the harbour basin from the east. We have not yet observed a str t re that o ld a t as a he k oint like the ridge that loses a ess to the west or like the a sewa r nning north-so th whi h lo ked the lake etween the har o r om le and the so thernmost ank o ssa 2009

T he B ridge A dressed stone10 bridge11 guards the west entrance of the channel. Its present appearance has hardly changed since oste s sket h in 1820 see ig 2- 47 m long in l ding the ram s, and 9 m wide ts ma im m height is 2 55 m It consists of two massive abutments of equal length 6 02 m for the so th a tment and 6 55 m for the north a tment he s a e of a ro imatel 8 35 m etween the two abutments is divided by a narrower intermediate ier 1 20 m his ier is off- entre om ared to the a is of the abutments, thus forming two passages of different widths 4 10 m to the so th and 3 05 m to the north whi h onl allowed a ess to the hannel to small oats o gia 1996 t d of the str t re leads to se eral on l sions First of all, the bridge was probably supported by masonry a lts igs 5-7 12 he flatness of the nat ral terrain reired the reation of a assage s ffi ientl high for the

a ations ha e identified se eral str t res the ett , whi h is interr ted three o enings two fl shes and a water s l hannel whi h ross its entire width two basins, one square, the other in the shape of a hoop, located inside the do k and atta hed to the ett , etween the north fl sh and the water s l hannel T he E ast Je tty The east jetty extends north-south and has a visible length of 230 m and a width ar ing of 5-6 m 13 It ends in the lake and it is ossi le to follow its e tent for a f rther

10 od iewi 1998 102 onl mentions a str t re n fa t, the bridge was necessary to cross the channel for those coming from the so th of the lake sing the a sewa and following the road north 11 s is the ase for most stone ildings in the region, it is al arenite, white dune limestone, extracted from the quarries on the taenia. 12 his ar hite t ral anal sis, and the model ig 7 are the work of st dents at 2 ro ni on , es e iall r nier and imene , nder the s er ision of o rnet , /

13 t is longer than the etties ilt on the so th shore at area 60 and 40 m long at the site referred to as i otos l e amses 2005 10 and is similar to str t res noti ed d ring ros e ting on an island of Mariout, at a site which played a major role in the management of lake traffi ibid. , ites 23, 12

91

LAKE MAREOTIS: RECONSTRUCTING THE PAST

F ig. 5. V

iew of the T aposiris bridge from north. ©

MA F T

F ig. 6. ( abov e) P l an and el ev ation of the north pil e of the bridge, N ov ember 204. L ay out, dr aw ing and D A O by M. El A mouri. © MA F T

F ig. 7. ( l eft) Model and reconstruction of the bridge in T aposiris Magna ( students of the Master P ro- L yon I I ) . © MA F T 92

M.-F. BOUSSAC & M. EL AMOURI: LAKE STRUCTURES AT TAPOSIRIS

F ig. 8. O v eral l pl an and detail ed outl ine of the eastern j etty. L ay out, dr aw ing and D A O by M. E l A mouri. © MA F T

20 m or so nder water ig 8 o the north it grad all disa ears first towards the west and then the east his asymmetry could indicate that the point of attachment of the causeway on dry land is different than at the southern lake end from that at the lake end rther north, the silting of the area and recent constructions prevent the correct reading of its ro te it ated on the a is, a o t 50 m to the north, is a masonr str t re 900114 whi h ma or may not be related to the jetty.

F ig. 9. T he southern end of the eastern j etty. V iew north. © MA F T

he ett ig 9 is ilt in lo ks of reg lar si e 0 50 0 20 0 25 m on a erage arranged in headers, and the design is comparable to that of the harbour structures present at the neigh o ring sites at amal and seir igs 10 11 a ated se tions of the eastern ett re ealed fo r well preserved courses, and two additional courses which have now disappeared. Over its entire length, both sides of the ett are n t ated with ttresses ig 12 hese 27 ttresses 16 on the east side and 11 on the west side are maintained on se eral o rses three or fo r Every second course has two headers while the alternate lo ks are stret hers on their edge he are on a erage 0 50-0 60 m wide and rotr de a o t 0 30 m he stabilised the entire jetty, while encouraging micro areas of

14 n in estigation was arried o t in 2006 on this str t re he report on surface ruins shows an angle similar to that of the jetty. If it belongs to the jetty this would alter our interpretation of the extent of the closed harbour basin.

93

from

LAKE MAREOTIS: RECONSTRUCTING THE PAST

F ig. 10. J

i A mouri. ©

etty at G amal . V iew from north- w est. ©

a a MA F T

ai

h i

s

h a

MA F T

F ig. 1 1. J

h

h

94

sh a

etty at Q useir. V iew from south- east. ©

ss

ay

a i

a

MA F T

y

M.-F. BOUSSAC & M. EL AMOURI: LAKE STRUCTURES AT TAPOSIRIS F ig. 13. V iew of the h sh i ex cav ation, f rom east. © MA F T

n addition, the east art of the fl sh is designed in order to ir late as m h s rfa e fresh water as ossi le ann 2003 32 16 towards the interior of the harbour basin: a base of grey mortar17 at the base of the projection formed by the paving of the western part, has a slope to the east which hinders the entry of sediment into the harbour basin, while allowing the passage of a stream of fresh water. he western art of the ottom of the fl sh, a ed in worked stone, is th s lo ated more than 0 40 m a o e the mortar.

sediment a m lation, th s red ing the risk of sa ing erosion due to currents and saline seepage.15 It is telling that particular attention was given to the construction of the eastern face of the jetty where the buttresses are more numerous. Indeed one notices a difference in treatment between the east side fa ing the o en lake the o tside of the asin that is subject to currents, and the west wall facing the interior of the asin n estigation of the north fl sh re ealed fo r o rses of lo ks to the east whereas there are onl two to the west row of worked stones and a stone lo kage line the east facing, but do not exist on the west side. In addition, the le el of the er eds of the lo ks on the east side is slightly higher than that of the west side. These differences in the construction show that the jetty is a rampart against the o en waters of the lake lo ated to the east o the west, the water in the basin is calmer and the construction of the jetty does not require these precautions.

ar ed into the fa es of the fl sh ond it on the a ed western half, there are several vertical notches with corres onding hori ontal groo es in the a ing, whi h allow the re reation of the s stem of losing and filtering of the fl sh t wo ld a ear that three of these lo king systems could operate simultaneously. They probably held filter sl i e gates or sim le martel iè res type gates, which allowed the ontrol of the flow of water a ording to rrents and seasons while limiting the entry of sediments to the basin.18 Two twin notches have no corresponding groo es whi h im lies the redesign of the a ing and lo king systems. A lightweight and retractable system to cross the fl sh ro a l e isted for assage a ross the ett , t no archaeological evidence shows its design.

T he N orth F l ush here are two fl shes, designed to ring as m h r nning water as ossi le into the asin and th s re ent silting one near the south tip of the jetty, the other to the north, more than 70 m from the resent shore line, onl the se ond fl sh has een the s e t of in estigation igs 12 13 ts east o ening is sha ed like a f nnel with two facing walls built in a quadrant. The west opening of this assage is straight and 1 2 m wide he a is of the fl sh is not er endi lar to the ett t shifted 15 degrees to the north, which probably gives an indication of the direction of the wind and/or dominant rrents in the region d ring antiquity.

16 ann 2003 32 the a thor states that the onstant flow of water helps control the temperature, oxygenation and salinity of the water, adding that the proximity of a source of fresh water is an advantage. 17 re mortar seems to ha e een hosen for the ed of the east entran es of the north fl sh and the water s l hannel owe er, ink h dra li mortar is sed aro nd the er areas of the north fl sh and water supply channel. 18 oes this s stem indi ate seasons with high le els of sedimentation re these seasons onne ted to ile floods ndeed, the ile is oined to ake areotis ia anals t ndergro nd water m st also ha e raised the le el of the lake d ring floods, ringing a great deal of sediment.

15 here is also a risk of erosion of the sides of the a sewa fa ing the hannel this was re ented the stone onstr tions, the ram and the retaining wall.

95

LAKE MAREOTIS: RECONSTRUCTING THE PAST F ish F arming S ystems ne of the most interesting findings is the dis o er of an original s stem for reeding/kee ing fish t the rrent stage of excavations, this consists of a water supply channel which crosses the jetty, a hoop-shaped basin and a fish-tank atta hed to the ett inside the har o r asin see ig 8 his om le , whi h has ossi l not een entirely uncovered, and which shows signs of repairs, indicates intensi e fish farming afon 2001 161, 165 , i e reeding in artifi ial ools rom a t ologi al oint of iew, these installations ilt on the lake are one of the t es of fish farming installations anal ed in ario s re ent st dies, mostl de oted to the oman world ron 2008 asins ilt digging in the sand and masonr work e see elow afon 2001 171 lato tells s that fish farming was ra ti ed on a large s ale on the anks of the Nile and papyri from the Hellenistic and Roman eras reveal much information about ichthyotropheia, piscinai and other apodocheia ho liara- a os 2003 rewer riedman 1989 of whi h o r s stem is an e am le he region of Canope is home to several piscinae dugout of the ro ks, some of whi h ha e long een known re ia 1926 d l- ag id forth oming t the installation uncovered in Taposiris is unique to this day around the shores of ake areotis afon 1998 573 19 However, the asso iation of har o r installations and fisheries is not surprising and the facilities found at Taposiris are similar for example to those found at Kenchreai dating to around 80 ranton, et al 1978 20 basins are created near the ett otha s, et al 2008 afon 2001 162 21 and connected to one another and with the sea through channels.

tion asin s with the fl sh, the groo es holding the filter sl i e gates or lo king martel iè res were modified d ring the rehabilitation of the water supply channel. he walls of the water s l hannel are ilt in worked stones o nd with ink mortar he lo ks whi h form the entrance to the channel are chamfered to facilitate the entry of water. It is a simple funnel system. At the entrance, the ottom of the hannel, like the ottom of the fl sh, is made of grey mortar sloping towards the east. he water s l hannel, like the fl sh, might ha e een o ered a rossing s stem, t we did not find an tra e of it, although we noticed the last course on its north wall was slightly battered. The western part of the south interior face of the channel ig 14 a ears to ha e ndergone re air with small nworked stones whi h ontrast with the s al lo ks of worked stone of the walls of the hannel ther re airs are visible in the south wall of the hoop-shaped basin. The meticulous cleaning of the water supply channel led to the discovery of seven coins wedged between the paving stones of the channel, and thus protected from the cleaning work ield nit 9240 i e of them elong to the ater oman m ire 4 and two are antine dodekanoummia of the 6th-7th ent r , in l ding one dated to era li s 610-641 23 Thus, the water supply hannel m st ha e f n tioned at least from the 4th century to the 7th ent r

T he W ater Suppl y Channel The water supply channel crosses the jetty perpendicularly o er its entire width igs 14 15 his hannel is 0 70 m wide, 6 m long and 0 85 m dee at the oint where the third course is still preserved. It consists of a small decantation basin, a ledge and three closing systems with notches ar ed o osite one another, 0 50-1 m a art the were sed to reate a kind of lo k, like in en hreai ranton, et al 1978 22

ts eriod of a andonment is s ggested tra es of lo king hamfered lo ks the same si e as those still in la e, and a lo k with a groo e identi al to that fo nd on the lo ks still in la e, were dis o ered d ring e aations hese lo ks th s ome from the o rses of the water s l hannel he hamfered lo ks were la ed at the entran e of the water s l hannel to lo k it the fit erfe tl in the entran e of the water s l hannel while kee ing one la er hori ontal, indi ating that the were ro a l deli eratel la ed there he a kfilling of the channel thus seems a deliberate action indicating the sto ing of the water s l or e en of fishing, t ens ring the continued functioning of the jetty.

One of these systems of notches still has its corresponding groo es t in the ottom of the a ing lo ks ig 16 his is the first a ing lo k to the east and a ts as a dam to stop the sediment accumulated in the small decanta-

T he H oop- shaped B asin he west end or e it of the water s l hannel flows into a hoo -sha ed str t re a o t 6 3 m the ase of whi h is a ed igs 8, 17 18 t is ilt in lo ks of worked stone, arranged in a single row and preserved to two or three o rses he lo ks are arranged in stretchers and their bases are covered with a hydraulic coating that forms a slight slo e on the inside edge of the asin see ig 18 This poorly preserved coating is only applied at the junction between the base pavement and the elevation.

19 afon 1998 573, n 3, notes that oastal s e lation has a sed the disappearance of many installations in the western Mediterranean o er the ast fift ears ake Mareotis is also threatened by intensive ro ert s e lation d ring o r first works on the east ett area, it was di ided into lots of land, shown o ndar makers 20 ranton, et al 1978 heir onstr tion dates a k to aro nd 80 21 lso see afon 2001 162, n 89, for osa the a thor notes that in the port of Cosa, “ canals joining the lagoon which has various dams, flow dire tl into the ort, and form art of the same ro e t 22 ranton, et al 1978 ote the resen e of se eral sl i es er hannel, ne essar to form intermediar asins and fish tra s the dimensions of the asins 1 0 7 m are similar to those fo nd at a osiris

23

96

t d

a her,

endi 2, infra.

M.-F. BOUSSAC & M. EL AMOURI: LAKE STRUCTURES AT TAPOSIRIS

F ig. 14. L ay out, e l ev ations and sections of north and south w al l s of the w ater suppl y channel through the eastern j etty. L ay out, dr aw ing and D A O by M. E l A mouri. © MA F T signed so as to reate a slo e down towards the fish-tank basin.

The hoop-shaped basin, fed with water through the supply channel, is closed by a double sluice gate system, to the north-east and south: to the north-east, vertical and hori ontal groo es were t into the lo ks forming the junction between the channel and the basin. To the south, the f nnel-sha ed end is fitted with the same s stem, indi ating the method of ontrol of the flow of water into the fish-tank, a large s are str t re en losed walls with do le fa ing ig 19 he n tion was also de-

The north wall of the basin is an extension of the north wall of the water supply channel and seems to be a renovation or addition ndeed, this wall is not linked to the hannel and is built with smaller, roughly hewn stones, some of whi h are re sed erha s the link with the hannel did not initially exist, at least in this form. 97

LAKE MAREOTIS: RECONSTRUCTING THE PAST T he F ish- tank he fish-tank is a s are asin whose interior dimensions are about 8 m x 8 m. It features walls angled north-south and east-west, all of different form. The west face of the east jetty serves as its east wall. The northeast corner of the basin corresponds to the end of the west face of the jetty. Investigations carried out on the extension of this face showed that no repairs had been made at this location and that the jetty had its original form. It appears that the south and west walls of the basin were entirely built under water. The south wall of an average width of 1 5 m, onsists of two arallel fa es ilt in stone and filled with green la sediment 9221 to the east and 9225 to the west his wall has an opening, built on at least two courses and equipped with a sl i e gate s stem to ontrol the flow of water he ro imit of the fl sh installed in the ett ro ides a fresh water s l to this so thern sl i e gate eatham ood 1958-59 265 24 The position of this pool is therefore also hosen e a se of the flow of fresh water into he harbour basin. Sl uice G ates and W ater F l ow s stem of fresh water flow was set etween the ario s str t res om osing this fisher he flow or the lo king of water, or at least of fish in the ase of filter sluice gates, is ensured by a system of sluice gates or martel iè res type gates. These are all designed in the same way: the erti al groo es t in the fa ing lo ks allow their ositioning a ross assages, while hori ontal groo es t in the paving slabs ensure a good grip. Some of these gates m st ha e let water filter thro gh to ens re a good regeneration of water in the fish tank owe er, no ar haeologi al evidence indicates their construction material.25 he si e etween 0 10-0 20 m wide for the different groo es in the fa ing or a ing lo ks s ggests that at least the e ternal framework of the sl i e gates was made of wood he may have been solid and entirely made of wood, or with openings, made with a wooden frame and metal grid.

F ig. 15. W ater suppl y channel through the eastern j etty opening into the hoop- shaped basin. V iew from east. © MA F T F ig. 16. G roov es and notches in the w ater suppl y chanel . V iew from north. © MA F T

The water from outside enters from two openings located to the south and north of the whole system: n the so th, fresh water arri es thro gh the fl sh located in the jetty. It enters the harbour basin and indire tl feeds the fish tank the assage lo ated in its south wall. It is closed by a sluice gate installed stream of the a ing in lined towards the fish-tank n the north, the water s l hannel feeds fresh water into the hoop-shaped basin. The arrival of water is regulated by four successive sluice gates, which lea e water to flow towards the hoo -sha ed asin while re enting fish from es a ing he fre en of these sl i e gates also fa ilitates maintenan e work leaning, re airs, et rearrangement ma e lain the position of the sluice gate located to the west end of the water supply channel, which opens directly onto the northeast corner of the hoop-shaped basin. The space partitioned off by this sluice gate and the one before it, is the largest in the channel. The last sluice

24 eatham ood 1958-59 265 n 8, on erning the need for a flow of fresh water into the fish tanks the a thors ite ortensi s riti ising l s whose fish tank does not ha e eno gh 25 iallano 1997 17 notes ron e sl i es with holes, like those re ommended ol mella iii 17,6 e a se the are fine nets whi h allow water to ass thro gh while tra ing the fish eatham ood 1958-59 265, n 9 t other s stems e ist in wood or om osite wooden frame and metal grid for e am le , or in stone like at o hlos eatham ood 1958-59 275 and late 63 a

98

M.-F. BOUSSAC & M. EL AMOURI: LAKE STRUCTURES AT TAPOSIRIS F ig. 17. V iew of the hoopshaped basin, t he w ater suppl y channel and the sh a k i h © MA F T

F ig. 18. T he pav ement in the hoop- shaped basin. V iew from north. © MA F T

F ig. 19. T he opening betw een the hoop- shaped basin and h sh a k i s h w est. © MA F T

99

LAKE MAREOTIS: RECONSTRUCTING THE PAST gate is fitted a ross and at an angle to the a is of the channel and is therefore directed towards the centre of the loop, which may also be due to a repair. This all suggests that this sluice gate dates from a later date.

9216/9220 filling the hoo -sha ed asin, rin i all no ered storage ontainers m horae 1, 4 and 5/6 , indi ate a final hase of se of the fish tank d ring the first half of the 7th century. This hypothesis is suported the a sen e of g tian am horae gloff 167, well documented on Alexandrine and Mariout sites from the mid-7th century. Similarly, among the coins wedged between the paving stones on the channel bed, there are two antine dodekanoummia from the 6th-7th century , in l ding one dated to era li s 610-641 see endi 2

he assage etween the hoo -sha ed asin and the fishtank, whi h is also lo ked a sl i e gate, is e i ed with the same system. Thus, all these adjustments help to prevent the basins from silting up while encouraging the onstant flow of water, whi h is a ke element for the s ri al of fish in a ti it iallano 1997 17 rthermore, the shallow and muddy bed is the perfect environment for fish farming afon 2001 159 26

This last phase of use of the system also relates to the last hase of the wareho se e a ated in 2000 on the a seway, strategically located near the causeway which crosses the lake o ssa 2009 129 he finds in l ded three Heraclius dodekanoummia and late am horae 1, 4, 5/6 and 7

T he F ish he fish s e ies identified in the ield nit related to the se of the fish tank 9215 and 9220 in arti lar at this stage of the st d are mainl marine fish or fish whi h li e in ra kish water afon 1998 575 27 A preliminary diagnosis carried out using photographs to visually identify the species28 in advance of a comprehensive study, indi ates the resen e of st rgeons and garfish B el one bel one t the resent stage of work, all the e a ated str tures seem too small to have housed real breeding farms est- a amanoli 1986 301 es ose it was for the tem orar storage of li e fish, like at en hreai

he intensit of lake a ti ities d ring late anti it orresponds to what the texts and archaeology tell us of the role of a osiris d ring this eriod the identifi ation in 2009 of late antique thermal baths in the middle town, compara le to those at area, f rther strengthens these findings owe er, nothing is et known a o t the ellenisti installations, although we are collecting more data on the it from the 2nd - 1st century BC, and even from the end of the 3rd ent r the tem le, a ording to r awass, was fo nded tolem , s ggesting that the it s deelo ment is linked to or a elerated this ro al interention awass 2008 29 owe er, it im lies nothing a o t lake onstr tions e ertheless, in the west, the area o ered the a kfill from the digging of the hannel in the first half of the 2nd ent r o ssa 2009 137-141 , re ealed man im orted am horae hodians and nidians whi h show the intensit of trade from the o tside ia le andria, at least etween the mid-2nd cent r and the earl 1st century BC. Although their fragmentation prevents us from being precise, the site where these containers were found, close to the shore, suggests that we are not dealing with land transport, but transport on the lake ernard 2009 his storage area s ggests s eifi installations of whi h nothing remains

Chronology In the east sector of the harbour system, the various phases of installation of the fish farming fa ilities and the hronologi al link etween the isi le s stem and the onstr tion of the ett are rrentl diffi lt to s e if hereas the west har o r s stem ridge and hannel an e dated to the 2nd century, the state of visible installations in the east remains unclear: the traces of repairs in the system between the hoo -sha ed asin and the hannel do not make it possible to date its construction, nor that of the visible jetty. All that can be said about the jetty is that it cannot e later than the 4th century. he resen e in the filling of the so th wall of the fish tank, of am horae remains 1 and fragments of ontainers 4, some of whi h are dated etween the last third of the 5th century and the middle of the 6th century,29 ro ides an a ro imate dating for its onstr tion terminus post quem

othing is known a o t the str t res from the eginning of the m erial era in des ri ing a osiris as the gate eisbol è ) of Egypt in his L etter to the A l ex andrians in 41, la di s refers to the fis al and militar role of the it , whi h was ro a l asso iated with s e ifi fa ilities o ssa 2001 owe er, the oldest hase of the har o r that we were able to identify is that which includes the digging of the hannel, whi h is not earlier than the first half of the 2nd ent r o ssa 2009 30 If one considers that the wareho ses and the fish farming fa ilities o erated until the 7th century, then one understands to what extent the intermediate stages of evolution still escape us.

However, ceramic and monetary data provide consistent dates for the last phase of use of the system: ceramic material dis o ered in the last ar haeologi al la er

26 afon 2001 159 reeding re ires er sheltered water with a shallow sandy or silty bed: lagoon shores provide the most favourable environment” . 27 afon 1998 575 he a thor insists that ra kish water is fa o ra le for reeding as it hel s the de elo ment of o ng fish and attra ts fish 28 ra hino, ers omm h ni i - arseille 29 t d i ne f, ee endi 1, infra.

30

100

ee

ar

i s st d in o ssa 2009

M.-F. BOUSSAC & M. EL AMOURI: LAKE STRUCTURES AT TAPOSIRIS

Appendix 1 T he Ceramics in the H arbour Area: the E astern Je tty ( S ector 9)

F ig. 20. Main types of ceramics found in T aposiris Magna ( Sector 9) during the L ate R oman era.

Archaeological excavations carried out in the area of the har o r asin, and more re isel on the eastern ett e tor 9 n o ered a few erami s often er fragmentar and eroded. They amount, except for fragments of bodies, to more than 243 sherds, or 119 indi id als 31 However, most of them, i.e. more than 66% of the ceramic assemlage 79 , onsist of storage and trans ort ontainers, both locally produced and imported, characteristic of the ate oman eriod, dating whi h is onfirmed the st d of oins homas a her see endi 2 this a er

ere r i on 1989 236-243 ,33 and were used for the storage and transport of wine, and possibly olive oil. Fragments fo nd in rea 9 ig 20, a are related to the 1 t e a ording to the lassifi ation of ate astern mhorae esta lished omini e i ri 2005 his form, generally attributed to the 6th and 7th centuries, is characteri ed thi k rolled rim an edge to mid-ne k re alls the mo lding of the re io s t e i ri 2005 75 he handles are now almost unribbed. The clay is rather dense and a general trend emerges marl la , fine te t re and a red to light brown section. In l sions are of large si e, and onsists mainl of grains of art , some white and sometimes red.

el hine i ne f,

, airo

1. T he Imported Amphorae and Ceramics h h a 32 mong the im orted ontainers at the site, 1 am horae are the se ond most ommon t e 10 his e iden e finds arallel in all ate oman g tian sites, from the se ond half of the 4th century until the mid-7th century, or slightly beyond. These jars were produced on the so thern oast of rke , r s and hodes m-

h

h a 4 rod tions from so thern alestine re resent 32 7 of , i e 39 indi id als he general form of these am horae ig 20, is mor hologi all lose to Pieri’ s type B and dated between the middle of the 6th and the 7th ent r i ri 2005 106-107 an of these amphorae have a sandy and gritty marl clay, with a medium dense fa ri and ontain se eral grains of art , some gre and white arti les of ario s si es he se tion is usually homogeneous, from buff to orange. Shaping and surface treatment are quite coarse, and clay accretions are fre entl o ser ed on the ne k he ma ros o i hara -

31 ll the sherds are anal sed sing the riteria of the la and sha e to determine the inim m m er of ndi id als onte t first, then for the whole se tor ims, handles, ottoms and odies fragments were counted. NMI or the global minimum, estimated for the sector, is the highest number among the different morphological features, except the bodies. Regarding the amphorae, the number of handles is divided by two. 32 he term ate oman m horae 1 to 7 refers to the t olog of ate astern am horae esta lished ile 1981 85-122 for the material found in Carthage.

33 owe er, onl two worksho s ha e een e a ated in r s a hos and igg ee emesti ha i haelides 2001 289-296 emesti ha 2003 469-476

101

LAKE MAREOTIS: RECONSTRUCTING THE PAST

i a i s h of the hoop- shaped basin ( F U 9216/ 920) .

teristics help us to locate the region of production in southern alestine, es e iall in the area of a a, shkelon and Ashdod.34 However, it is interesting to note the similarity between the clay of Palestine amphorae and ceramics manufactured from Mareotic clays.

e tor 9 are mor hologi all lose to i ri’ s “ bag-shaped am hora, t e 4 i ri 2005 117, fig 76 t is a glo lar am hora of small si e the ne k is short and lindri al or on e , ending with a small ro nd and/or flared rim, with a sin o s rofile ig 20, n the ellia site, ranç oise Bonnet-Borel sets the appearance of these containers during the second quarter of the 7th ent r onnet 1983 442 , howe er it is not im ossi le that the started as early as the late 6th century. They are still present during the Fatimid era, with no major change in the shape ogt 1997 256, 258 s regards the identifi ation of site productions, only one group of clay could so far be identified marl la , medi m densit fa ri , sand the se tion is usually brown to buff. The inclusions consist of several grains of art , some gre and white arti les of ario s si es he ma ros o i e amination of the am horae, which can logically be assumed to be local or at least regional productions, questions the distinction between Palestinian and Mareotic productions. Thus, two hypotheses can be put forward: they are either Egyptian or Palestinian ontainers e orted with the 4 m horae nl macroscopic and petrographic analysis would answer this question.

n addition to trans ort ontainers of t es 1 and 4, a few fragments of am horae and erami s o ld not e identified with the e e tion of a fragment of the od of an African sigil l ata and two elements of Cypriot sigil l ata. It is a ring-base low bottom and a rim which seems to corres ond to a es 9 sha e and date from late 6th to late 7th ent ries a es 1972 378-382 2.

E gyptian Production y ia i i a ay ith the e e tion of 12 resid al fragments 6 of 3 am horae, marl la rod tions are mainl illustrated by globular or ovoid containers, more commonly referred to as “ bag-shaped” , which were produced in many eastern editerranean sites, more s e ifi all in the eant, alestine and g t 15 ontainers fo nd in

34 or a state of estion on i ne f 2005 54-62

i

Q uite logically, common ware is illustrated primarily by marl clay products, some with alluvial content, manufact red witho t an do t in areoti worksho s, on the site

4 am horae and i liogra h , see

102

M.-F. BOUSSAC & M. EL AMOURI: LAKE STRUCTURES AT TAPOSIRIS

Appendix 2 T he Coins from S ector 9

or at least near a herek 2001 60-64 he sha e re ertor in l des dishes with ro nded rim ig 20, d , some of them with a high-placed carination in their external wall, s with flattened arination ig 20, e , fragments of storage ars and ases for li ids and some ooking ots ig 20, f

Thomas Faucher, Sorbonne, Paris A number of coins whose study is currently in progress were fo nd d ring the e a ations in e tor 9 fter some reliminar restoration, it is ossi le to make a few remarks here

y ia i i ia ay hese rod tions from the elta and/or the ile alle are onl a small art of the erami s 17 indi id als, i e 14 2 of all erami s identified art from some fragments of three AE amphorae, a bottom and a carinated sho lder of a wine 7 ontainer ig 20, g ha e een identified ommon wares are ill strated owls ig 20, h ,35 dishes with ro nded rims, and two ooking ots with a C shaped rim dated at the Kellia to the 7th century gloff 1977 103, l 52 no 4 6, t e 138 and to the first half of the 7th century in Baouit.36

ost of the oins elong to the ate oman and antine eriods o oin dating to earlier than the 4th century was fo nd, whether in or o t of stratigra hi al onte t ll the ate oman oins 12 items are small denominations ca 10 mm ommonl alled 4 for ron e and 4 for the si e he identifi ation of these oins is not ossi le ntil the restoration is om leted t it is nlikel that more information about the type and date of these coins will be gained, for two reasons. First, the proximit of the lake water has fa ilitated the orrosion of these coins which are often in poor condition. Secondly, these oins, ommonl dated to the 4th-5th centuries, were often molded oins, nlike str k oins, rod ed more or less legall to o er ome the la k of rren his te hni e allowed for the production of a large number of coins in a short time, t makes oins almost illegi le, e en when they are just produced. Therefore, it is logical that their circulation and subsequent deposition in the ground has erased any information.

3. T he Assemblages h ish y he so thern wall of the fish-tank onsists of two fa es with a filling of green la 9221/9225 ew erami s were fo nd in the filling t the ro ide a hronolog for the onstr tion of the fish-tank and ro a l the whole om le he onsist of odies of some 1 am horae and fragments of 4 ontainers in l ding a ottom and a rim lose to i ri s 1 form this form is dated from the last third of the 5th to the middle of the 6th ent r i ri 2005 105-106

In addition to these Roman coins, there are also three dodekanoummia a oin of 12 noummion easil re ognisable by the letters IB on the reverse, produced using the same molding ro ess mong the latter, were identified one oin from the reign of stin 518-527 and a se ond from the reign of era li s 610-641 t is diffi lt to give a precise date for molded coins since the coin might have been molded later than the date of production of the model.

The ceramic material discovered in the last archaeologial la er 9216/9220 of the hoo -net sha ed asin onsists mainl of storage ontainers, i e 1, 4 and 5/6 m horae, and fragments of ommon wares whi h might e linked to food ons m tion the fishermen ig 21 h s, this material attests to a final hase of se of the asin d ring the first half of the 7th century. his h othesis is onfirmed the a sen e of g tian gloff 167 am horae, well do mented on le andrine and Mariout sites from the mid-7th century onwards.

The Alexandrian excavations emphasised the importance of this phenomenon whose magnitude was often underestimated in the past. This manufacturing technique is attested as early as the Ptolemaic era and continues in Roman and antine eriods des ite its a arent restri ted se d ring the arl oman m ire in an ase it is diffi lt to sa if it was outlawed by the state and successfully repressed.

To conclude, the study of Late Roman pottery discovered at e tor 9 offers some interesting insights into the ommercial activities and trade in the harbour area, which was the western stoms of le andria on the areotis ake since Roman times if not before. In addition to local marl clay productions, the repertory includes several imports, mainly of wine containers from the eastern Mediterranean and some erami s from the elta and/or the ile alle , refle ting the e onomi italit of the region and fishing activities.

hen f ll restored, the oins from e tor 9 will ho ef lly provide a more precise datation and help discern more accurately the different periods of use of the eastern jetty area.

35 ne sho ld tentati el om are this rim fragments with a series of dishes fo nd at ellia and dated to the first half of the 7th ent r however these are of alluvial clay and with a painted decoration on white sli ee onnet 1994 378-379, fig 231 o 138 36 n lished material

103

LAKE MAREOTIS: RECONSTRUCTING THE PAST Bibliography ernard, , 2009, le andrian T ainiai and and raffi atterns ote on the mnest e ree P . T eb 5 in Light of the Topography. Z eitschrift für P apyrol ogie und Epigraphik 168 265-270 l e, , amses, , 2005, ake areotis esear h ro e t reliminar e ort from ilot r e g st 2004 N ew sl etter of the A rchaeol ogical Society of A l ex andria 15 ( F ebruary) 5-16 onnet, , 1983, oteries, erres, monnaies, d ors et ins ri tions n asser l s eds , Surv ey arché ol ogique des K el l ia ( B asse- É gypte) . R apport de l a campagne 198: 423-480 o ain onnet, , 1994, e mat riel ar h ologi e r olt en 1977, 1982 et 1983 a o o r er- o i t n asser onnet eds , Ex pl orations aux Q ouç oût er- R oubâ’ î yât . R apport des campagnes 1982 et 1983 349-406 o ain o gia, , 1996, A ncient bridges in G reece and coastal A sia Minor h thesis, ni ersit of enns l ania o ssa , - , 2001, e illes en ar otide a osiris Magna et Plinthine. B ul l etin de l a Socié té franç aise d’ é gyptol ogie 150 42-72 o ssa , - , 2007, e her hes r entes a osiris agna et linthine, g te 1998-2006 Comptes rendus des sé ances de l ’ A cadé mie des inscriptions et bel l es- l ettres ( j anv ier- mars) : 445-479 o ssa , - , 2009, a osiris agna la r ation d ort artifi iel n mas e rel eds , A rché ol ogie et env ironnement dans l a Mé diterrané e antique: 123-142 aris re ia, , 1914, A l ex andria ad A egyptum. Bergamo. re ia, , 1926, Monuments de l ’ É gypte gré co- romaine, I : L e rov ine e i monumenti di Canopo. Bergamo. rewer, , riedman, , 1989, ish a shi i ancient Egypt. Warminster. ho liara- a os, , 2003, L a pê che en É gypte à l a l umiè re des papyrus grecs. Ioannina. e osson, , 1935, Mareotis: B eing a Short A ccount of the H istory and A ncient Monuments of the N orthW estern D esert of Egypt and L ake Mareotis. London. emesti ha, , 2003, m hora rod tion on r s d ring the ate oman eriod n akirt is ed , D e R ome à B yz ance; de F ostat à Cordoue. É v ol ution des faciè s cé ramiques en Mé diterrané e ( V e – I X e siè cl es) , actes du V I I e congrè s international sur l a cé ramique mé dié v al e ( T hessal onique, 1 1- 16 octobre 19) : 469476 th nes emesti ha, , i halides, , 2001, The Excavation of a ate oman 1 m hora kiln in a hos n illene e atson eds , L a cé ramique byz antine et proto- isl amique en Syrie- J ordanie ( I V e - V I I I e centuries A D . J .- C.) . A ctes du col l oque d’ A mman ( 3- 5 dé cembre 194) 289-296 eir t i ne f, , 2005, rod tion et ir lation des iens a a d rant l nti it tardi e le t moignage des am hores n alio ed , G az a dans l ’ A ntiquité T ardiv e. A rché ol ogie, rhé torique et histoire: 54-62 Salerne.

gloff, , 1977, K el l ia. L a poterie copte. Q uatre siè cl es d’ artisanat et d’ é changes en B asse É gypte. Gene m ere r, - , i on, , 1989, es r gions de rod tion d am hores im riales en diterran e orientale. A mphores romaines et histoire é conomique: dix ans de recherché . A ctes du col l oque de Sienne ( 224 m ai 1986) : 236-243 ome m ere r, - , i on, , 1998, es ateliers d amhores d a ario t n m ere r ed , Commerce et artisanat dans l ’ A l ex andrie hel l é nistique et romaine. A ctes du col l oque d’ A thè nes, organisé par l e CN R S, l e L aboratoire de cé ramol ogie de L yon et l ’ Ecol e franç aise d’ A thè nes, 1 1- 12 dé cembre 198: 75-88 thens, aris est- a amanoli, , 1986, r h ologie, ethnogra hie o ethnoar h ologie des resso r es marine de sites cotiers. Le cas de la pê che aux muges dans les lagunes de la r e identale n erard ed , V I è mes rencontres international es d’ arché ol ogie et d’ histoire d’ A ntibes. L ’ ex pl oitation de l a mer de l ’ A ntiquité à nos j ours. II: L a mer comme l ieu d’ é changes et de communication : 281-303 al onne awass, , 2008, n sear h of leo atra and ark Antony. H orus g st/ e tem er 26-29 a es , 1972 L ate R oman P ottery. London. ron, , 2008, e onstr ting the e hni es and otential Productivity of Roman Aquaculture in the Light of e ent esear h and ra ti e n ermon ed , V ers une gestion inté gré e de l ’ eau dans l ’ empire romain 175-185 ome afon, , 1998, is inae et is i lt re dans le assin o idental de la diterran e J ournal of R oman A rchaeol ogy 11 573-582 afon, , 2001, V il l a Maritima, recherches sur l es v il l as l ittoral es de l ’ I tal ie romaine ( 3 è me s. av . J .- C. / 3 è me s. ap. J .- C.) . Rome. eatham, , ood, , 1958-59, marine loration in rete, 1955 T he A nnual of the B ritish School at A thens 53-54 263-280 El-Maguid, A., forthcoming, Maamourah Surveys: Preliminar re ort n alas ed , T ropis V I I I : Eighth conference on ship construction in antiquity, H ydra, G reece, A ugust 20 . Athens. a herek, , 2001, area 2001 ote on the otter P ol ish A rchaeol ogy in the Mediterranean 13 60-64 McCann, A.-M., 2003, h a a sh y Cosa, a s hort guide. Rome. hsens hlager, - , 1979, a osiris agna 1975 season. F irst I nternational Congress of Egyptol ogy: 503-506 erlin hsens hlager, - , 1999, a osiris agna n ard ed , Encycl opedia of the A rchaeol ogy of A ncient Egypt 759-761 ford a ho, - , 1827, R el ation d’ un v oyage dans l a Marmarique et l a Cyré naï que. Paris. i ri, , 2005, L e commerce du v in oriental à l ’ é poque byz antine ( V e – V I I e siè cl es) . L e té moignage des amphores en G aul e. Beyrouth. 104

M.-F. BOUSSAC & M. EL AMOURI: LAKE STRUCTURES AT TAPOSIRIS ile , , 1981, otter from the isterns 1977 1, 1977 2 and 1977 3 Ex cav ations at Carthage 19 7 conducted by the U niv ersity of Michigan 6 85-122 od iewi , , 1998, rom le andria to the est land and waterways. n m ere r ed , Commerce et artisanat dans l ’ A l ex andrie hel l é nistique et romaine. A ctes du col l oque d’ A thè nes, organisé par l e CN R S, l e L aboratoire de cé ramol ogie de L yon et l ’ Ecol e franç aise d’ A thè nes, 1 1- 12 dé cembre 198: 93-103. Athens, Paris. od iewi , , 2002, areoti har o rs n oert ed , A l ex andrie mé dié v al e 2 1-22 e aire otha s, , einhardt, , oller, , 2008, arth akes and siden e at en hreai sing e ent arth akes to e onsider the r haeologi al

and iterar

iden e n araher, ones all oore eds , A rchaeol ogy and H istory in R oman, Mediev al and P ost- Mediev al G reece: Studies on Method and Meaning in H onor of T imothy E. G regory 53-66 ldershot iallano, , 1997, P oissons de l ’ antiquité . Catalogue d e osition, 1er a ril-31 ao t 1997 s e d stres ranton, , haw, , rahim, , 1978, K enchreai Eastern P ort of Corinth, ol e de ogt, , 1997, es rami es omme ades et a assides d sta l ntar- ostat traditions m diterran ennes et infl en es orientales n mians d r him a d ed , L a cé ramique mé dié v al e en Mé diterrané e. A ctes du V I e congrè s de l ’ A I ECM2 ( A ix - en- P rov ence, 13- 18 nov embre 195) 243-260 i -en- ro en e

105

LAKE MAREOTIS: RECONSTRUCTING THE PAST

106

M. BERGMANN, ET AL.: SCHEDIA ON THE CANOPIC NILE

Schedia, Alexandria’ s Harbour on the Canopic Nile. Interim Report on the German Mission at Kom el Giza/Beheira (2003-2008) Marianne Bergmann, Michael Heinzelmann & Archer Martin H istorical S ources and Research H istory Shortly after the foundation of Alexandria, the new metro olis was onne ted to the ano i ile an artifi ial, 30 km long channel. At the point where the canal branched off the Canopic Nile a new town, Schedia, was founded (see in general: RE 1921: 401-403; RE 1899: 2030; Fig. 1).1

to display them. At the same time they are an indication of the formation of a prosperous citizenry with an understanding of itself as inhabitants of Schedia. In Roman times Christianity seems to have established itself early under le andrian infl en e hedia re ei ed its own e is o al seat. In spite of its importance the site as well as the Nile Branch, is still virtually unknown. The river and the canal have now vanished from the landscape. Schedia itself was still a prominent stretch of hills at the beginning of the 20th century with a length of about 2.5 km and a width of 1.5 km. Intensive sebakh extraction and the increasing encroachment on the ancient settlement site by the villages of Kom el Giza, Kom el Hamam and Kom el Nashwa have, however, made the topography so unclear that the place is hardly to be noticed in the landscape. Only in a few places an the flat remains of the koms and single str t res e recognized. The greatest height in the territory of Kom el Hamam is occupied by the present cemetery.

In its important river harbour all goods coming from Upper Egypt had to be transferred to smaller vessels. It was also a customs station wherefore a pontoon bridge was installed (σχεδία), which gave its name to the place. Schedia seems to ha e een a flo rishing reek-dominated pol is throughout Ptolemaic and Roman times and was inhabited until the early Byzantine period. Ancient sources suggest the existence of ample harbour and storage installations with the personnel and administrative infrastructure that belonged to them (Strabo, G eogr. 17,1,16; Procopius, A ed. 6,1). The inscriptions found in the area of Schedia provide further evidence of the character of the place (Bernand 1966: 45 ff., 1970a: 329 ff.; Botti 1902). Thus, cults of Z eus Soter and Athena Polias are attested, along with a ship and military station already in the late 4th century BC. About 115 BC the garrison station at Schedia dedicated a Kleopatreion. These cults, as well as early Ptolemaic grave inscriptions with Greek names, appear to indicate a high percentage of Hellenic inhabitants or at least of strongly Hellenized people in the city founded ex nov o. There was, however, a Jewish colony as early as the time of Ptolemaios III Euergetes (246-221 BC) with one of the oldest known synagogues attested by an inscription – evidence of the early multicultural character of this harbour city. In the Roman Imperial period Schedia calls itself a pol is in inscriptions and apparently has its own magistrate (archon). Various honorary statues and bases for private citizens and emperors found in the area of Kom el Giza suggest that there must have been appropriate public spaces in which

fter the first ag e attem ts to lo ali e hedia in the 18th century, it was the Napoleonic expedition of 1798-1802 that first on erned itself with hedia and the le andrian canal (Jomard & Jacotin 1818/25; Bernand 1970b: 422). In this onne tion the la e is shown on the first a rate ma as one of largest tells in the Nile Delta. A participant in the e edition, ean i hel e re, was the first to ro ose the localization of the beginning of the canal at El-Karyun and the identifi ation of the gro of mo nds of ashwa and Kom el Giza as Schedia. Subsequently, Ptolemaic-Roman finds egin to e re orted in the area later identified as hedia, mostl han e finds farmers or eo le engaged in quarrying sebakh. These objects, which include a series of honorary and funerary inscriptions and a Nilometer, finished in art in ri ate olle tions t mostl in the museums of Cairo and Alexandria.

1. This project is being carried out by the Institute for Classical Archaeology of the University of Goettingen in cooperation with the Archaeological Prospection Service of the University of Southampton. Further support came from the German Archaeological Institute, the Department of Geodesy at the University of Stuttgart, the Geology Department of Mansoura University and the Centre d’ É tudes Alexandrines. he ro e t has een f nded first the rit h ssen o ndation and since 2004 by the German Research Foundation. We thank the Supreme Council of Antiquities and its Director General, Dr. Z ahi Hawass, for the permission to work at Schedia. For help and support we thank the head of the Antiquities Service for the Department of Beheira, Fawzi el Choulani, and the director of the SCA’ s excavations at Schedia, Ahmed Abd el Fattah.

t was onl in the 1980s, tho gh, that the first reg lar archaeological work was done at the site, when the Egyptian Antiquities Service carried out some rescue excavations because of the growing threat to the ancient site through modern construction (Abd el Fattah 1988, 1998, 20002001). Several areas of varying sizes were investigated on the eastern edge of Kom el Giza and at Kom el Hamam, that is in the eastern part of the ancient settlement (see Fig. 4). Numerous building structures of the late Hellenistic, Roman and Byzantine periods were uncovered. A high number of Hellenistic objects were found, among them numerous coins. The excavated areas were restored and protected by the Egyptian Antiquities Service. 107

LAKE MAREOTIS: RECONSTRUCTING THE PAST

F ig. 1:

A l ex andria and its hinterl and in 1925,

Survey of Egypt, s heet 4 ( detail ) .

T he N ew E v idence Since 2003 thanks to the Supreme Council of Antiquities (SCA), the Institute of Classical Archaeology of the University of Goettingen has conducted an interdisciplinary project consisting of a topographical survey, large scale geophysical surveying, geological investigations, selective stratigraphic excavations and a comprehensive study of the finds ergmann ein elmann 2007

boundaries can also disclose characteristics of the ancient topography, because some of them seem to document ancient boundaries, like the former edge of a river branch. G eophysical and G eo- archaeol ogical I nv estigations Three seasons of geophysical surveys have been conducted at the site since 2003 by the University of Southampton, to assess the archaeological remains of the area surrounding the Schedia excavations (Figs. 2 & 3). The surveys were carried out using magnetometry. Furthermore, in 2005 a series of deep drillings and sedimentary analyses was carried out in collaboration with the Geology Department of the University of Mansoura, which will give further evidence of the course of the Canopic Nile and the canal.

T opographical Surv ey In order to integrate old maps and the results of the excavations, geomagnetic and geological investigations into a nified s stem, it was ne essar to esta lish a raster of fi ed oints for the whole site in estigated fter the introduction of this geodetic net, relevant points of the newly esta lished fi ed oint s stem were meas red sing a real time GPS to determine the parameters of transformation for the implementation into the Egypt Blue-Belt-system. Subsequently an up-to-date cadastral map was created on the basis of recent Satellite images (Fig. 2). This mapping of the modern surroundings of ancient Schedia was indispensable for several reasons. With regard to the planning of the geomagnetic surveying and the stratigraphic sondages, detailed knowledge of the modern topography es e iall of the anals, streets and field orders is of great importance. Furthermore, it is possible to draw some conclusions about the ancient landscape from its current shape. This holds true for the relief of the terrain, which reveals areas of settlement as well as the course of the proba le har o r asin, t to a ertain degree modern field

At present these results seem to suggest that the modern Kanubiye Canal follows the south-western embankment of the ancient Canopic Nile which may have passed between Kom el Giza and Kom el Hamam. In that case, its breadth amounted to 150 m. As there is also clear evidence that there was once a lagoon or a second branch of the river to the north of Kom el Giza, it can be assumed that the settlement of Kom el Giza actually was isolated on a kind of long and narrow island. Maybe it was this topographical situation which gave the stimulus to construct a pontoon bridge here. The investigations have also revealed a possible localisation of the Alexandrian Canal with a strict east-west orientation between Kom el Nashwa and Kom el Sherif and a 108

M. BERGMANN, ET AL.: SCHEDIA ON THE CANOPIC NILE

F ig. 2: Schedia. T opographical map based on quickbird satel l ite image ( 205) ( A ) and corings ( H ) . G erman Mission at K om el G iz a/B eheira.

indicating ex cav ated areas

F ig. 3: Schedia. M ap w ith reconstruction of the ancient topography. G erman Mission at K om el G iz a/ B eheira.

109

LAKE MAREOTIS: RECONSTRUCTING THE PAST

F ig. 4. Sc 20306. G

hedia. P l an of the southeastern settl ement ( K om el G iz a, K om el H amam) w ith ex cav ation areas 19 erman Mission at K om el G iz a/B eheira.

possible harbour basin immediately north of it. A second canal could have existed between Kom el Hamam and Kom el Sherif, branching at a right angle from the Canopic Nile and then turning in a westward direction. We thus appear to have a complex system of canals.

80-

92 and

the ar hite t ral remains do mented as a first ste e ondly, some stratigraphic trenches were excavated in order to learn more about the chronology and function of the structures. Furthermore, a series of new stratigraphic sondages have been carried out in previously untouched areas.

As for the extension of the ancient city, it is quite clear that the ancient settlement once occupied the whole area around the possible harbour basin (today’ s Nashwa), the whole southern bank of the former Canopic Nile for a length of ca. 2.5 km (today’ s Kom el Sherif, Kom el Hamam, el Karyun), as well as parts of the northern bank (today’ s Kom el Giza). However it seems that the built up areas did not rea h a signifi ant de th t st k rather lose to the edges of the waterways, because of the necessities of a river harbour with its ongoing transhipment of goods. Thus, the picture that emerges is that of a rather amorphous and unhomogeneous settlement, consisting of a series of elongated narrow strips along the waterways, which also separated the different parts of the settlement from each other. Its nucleus has to be assumed in the vicinity of the harbour, from where it probably grew step-by-step without over-all planning along the river embankment. Unfortunately, the former kiman in the area of the harbour have suffered most from sebakh extraction, while we have to assume that the better preserved areas of the former settlement at Kom el Giza and Kom el Hamam were actually on the outskirts.

Summing up the results the following picture appears. In the western part of the older excavations (Area 3), there is a double-tholos-bath with bath-tubs to sit in (Fig. 5). It belongs to a Hellenistic type, which is well attested in the Nile Delta, the Fayum and other parts of Egypt, where it was used as a Hellenistic heritage much longer than anywhere else in the ancient world (see Boussac, et. al. forthcoming) A special feature of the Schedia bath is the visible division between the baths for men and women, otherwise attested in the ontem orar a ri oin finds and the deep foundations suggest that the bath was built in Hellenistic times, but it is one of the rare examples whose long duration can be shown. Provided with additional rooms with bath tubs for immersion and with new high quality terra o floors, it was sed at least ntil aro nd 200 (Bergmann & Heinzelmann forthcoming). In its immediate vicinity, still in Area 3, the remains of a huge foundation are visible. Its dimensions seem to indicate that they belonged to a big public building, maybe a temple. Its facade was probably oriented westwards to the Canopic Nile. Unfortunately, however, this part is inaccessi le e a se of modern onstr tions gain all findings in this area, including the coins, indicate a (late? ) Hellenistic date. Both buildings, the round baths and the possible tem le fo ndation, onsist of fired ri ks, whi h is ite unusual for Ptolemaic architecture outside Alexandria.

Ex cav ations at K om el G iz a ne im ortant aim of the field ar haeologi al a ti ities was to explore further the parts of the town found during the SCA’ s rescue excavations in the 80s and early 90s (Fig. 4). Therefore, these areas were cleaned and all 110

M. BERGMANN, ET AL.: SCHEDIA ON THE CANOPIC NILE

F ig. 5:

K om el G iz a, A rea 3. P

l an of H el l enistic baths. G erman Mission at K om el G iz a/B eheira.

F ig. 6: K om el G iz a, A rea 1. P l an of R oman v il l a, t ombs and l ater basins. G erman Mission at K om el G iz a/B eheira.

111

LAKE MAREOTIS: RECONSTRUCTING THE PAST F ig. 7: 203-

206. G

K om el H amam. P l an w ith ex cav ations areas erman Mission at K om el G iz a/B eheira.

F ig. 8: K om el H amam. P l an of A rea 6 w ith R oman storage buil ding and l ater phases. G erman Mission at K om el G iz a/B eheira.

112

M. BERGMANN, ET AL.: SCHEDIA ON THE CANOPIC NILE

F ig. 9: K om el H amam. R econstruction of the R oman storage buil ding ( J . Sc humann) . G erman Mission at K om el G iz a/B eheira.

While these two structures of Area 3 seem to be in a certain sense public buildings, Areas 1-2 farther to the east reveal a completely different picture (Fig. 6). In both areas the earliest traces of use were numerous constructions that are possibly to be interpreted as funerary monuments dated on typological grounds to the Roman period. Especially one type of monument is very characteristic: a stepped pyramidal structure of ashlar blocks, probably with a pillar on it, which has close parallels in the necropoli of Alexandria and Marina el-Alamein (Daszewski 1998).

particular interest is a large building measuring 70 x 40 m, which is visible on the surface of Kom el Hamam. Three stratigraphic sondages were carried out down to the foundations of the building (Figs. 8 & 9). They show that it was constructed around AD 100 above earlier mud-brick structures, which were demolished for the purpose. The building was erected in an interesting building technique: the supporting structure consisted of 10 x 8 rows of massive brick pillars with cores of opus caementicium, which bore massive cross vaults of caementicium, collapsed fragments of whi h were fo nd t was an enormo s fired brick building, particularly by Egyptian standards, that also demonstrates an in-depth technical knowledge of Roman buildings from central Italy of the period. However, two features indicate that the responsible architect was not really familiar with this building technique: contrary to the Roman prototypes, all pillars were constructed without foundations, instead of which huge mudbrick pylons were added already in the original phase of the building on the outside in order to absorb the pressure of the vaults. Furthermore, mud-brick walls running in a north-south direction were constructed between the pillars, which means that the building was separated into a series of nine parallel naves each ca. 5 m wide, 40 m long and 5.5 m high. All these naves opened on the south side thro gh h ge doors he floor onsisted of a sim le mud-brick pavement. In the middle of the naves long low benches (ca. 0.25 m high, 1.50 m wide) made of mudbricks, seem to have been installed. Unfortunately, no speifi finds were fo nd that o ld gi e an indi ation of the original function of the building; obviously it was completely emptied before the changes of the second phase. However, judging by the typology of the building, it most probably formed a large storage facility for some kind of precious goods.

One of the less well preserved pyramidal structures has een in estigated ts ore onsisted of a e of fired bricks. Around and underneath these bricks a stratum containing ashes, bones and pottery has been found. All bones belonged to animals of different types, such as at least four jaws of sheep as well as a complete skeleton of a young goat missing the head. They do not show signs of burning. However, a certain portion of the bones concentrated in one area was burnt and was in a very fragmentary state of conservation. They have yet to be analyzed, but it seems likely that they belong to a human cremation, while the unburnt animal bones might be connected to some funeral rites. Because of a Trajanic coin found within the ash layers this structure probably belongs to the 2nd century AD. The area must be regarded as suburban cemetery. At a somewhat later time, an isolated villa was built apparently in the midst of the sparse graves of Area 1 (Fig. 6). This villa of the middle Imperial period, was organised around a central atrium-like courtyard and was f rnished with mosai s and o s se tile floors t was in use for a long time and repeatedly rebuilt and renovated. During the 4th century AD the villa was abandoned, and numerous basins were installed over its ruins and the neighbouring graves, as well as in most of the other areas excavated up to now. Undoubtedly, these were production units that were used intensively and were repaired repeatedly up to the 6th century.

Later, the building underwent several different phases of use and renovation. Only a short time after its erection, additional mud-brick walls were added during the later 2nd and 3rd centuries AD in order to create simple apartments. These habitations show two main phases of use. In one of the sectors several rooms were examined, some of them containing cooking installations. They revealed a huge amount of pottery, mainly amphorae, as well as coarse and

Ex cav ations at K om el H amam New archeological examinations were carried out at Kom el Hamam and other parts of the ancient city (Fig. 7). Of 113

LAKE MAREOTIS: RECONSTRUCTING THE PAST F ig. 10: K om el H amam, A rea 9. R oman encl osure w al l . G erman Mission at K om el G iz a/B eheira.

cooking wares, mostly of local production. In the late 4th or early 5th century AD these habitations were abandoned and levelled in order to install a granary. On a level ca. 1 m above the pavements of the prior habitations, a special ri k floor was onstr ted s orted some do ens of small parallel vaulted chambers – a typical Roman system for keeping grain dry. This grain storage facility was in use until the late 5th or 6th centuries AD, when it was destroyed fire n its r ins sim le dwellings and agrarian installations were installed, which were in use until the Arab conquest.

Once again in the 4th century, a radical change took place. As in the case of the v il l a suburbana at Kom el Giza, most of the earlier buildings at Kom el Hamam were destroyed or changed their function. Instead of the dwellings or monumental buildings, a whole series of simple working installations occur. Among these are several types of basins, which seem to have served various purposes. Some have the characteristic arrangement for wine production with a higher platform for treading and a lower catchment basin accompanied by further basins for fermentation (Fig. 11) (Rodziewicz 1988). Others are isolated or have furnaces nearby and appear to have served other productive processes that annot e identified at resent side from their function, the basins testify to an intensive agricultural use of the territory in late antiquity. These agricultural installations were in use until the beginning of the 7th century, when Schedia was completely abandoned at the time of the Arab invasion.

Immediately south of this huge building, the remains of another monumental structure was uncovered (Figs. 7 & 10). At a distance of ca. 4.5 m a huge wall runs parallel to the southern front of the above-mentioned storage building. The wall has impressive dimensions: with a thickness of about 2 m its foundations consist of a massive foundation of opus caementicium and three layers of huge ashlar blocks, each with a height of 0.45 m, while the elevation consisted of two more layers of ashlar blocks and a massive mud-brick wall on top. The stratigraphic evidence shows that the wall was constructed shortly before the h ge storage ilding, ro a l in the first half of the 1st century AD. Its function is not yet clear. The dimensions as well as the massive construction technique indicate that the wall must originally have reached an impressive height. Therefore, it seems probable that the wall actually formed a high enclosure belonging maybe to a temple or some other kind of public building. Further investigations showed that the wall actually belonged to a much bigger structure, consisting of at least two rectangular courts. Strangely enough the northern half of it was abandoned shortly after its construction and substituted by the huge pillar building, while the southern part was continuously in use until late antiquity.

A nal ysis of the F inds The detailed analysis of our excavated material, especially coins and pottery, will give further information in the future concerning Schedia’ s economic development. Howe er, e a se of the h ge amo nt of finds we are st at the beginning. Coins Around 4,000 coins found at the ancient site of Schedia have been analyzed by Ch. Noeske. Among them are 1,400 stra and stratified finds from the re ent e a ations n the whole they cover the period from ca. 320 BC (Ptolemy as a satrap) to the time of Arab conquest (Heraclius, AD 641). Hellenistic coins from the SCA excavations at Kom el Giza are numerous, while in the excavations on the northern side of the Kanubiye Canal (Kom el Hamam, Kom el Sherif) the Roman Imperial and late-antique coins 114

M. BERGMANN, ET AL.: SCHEDIA ON THE CANOPIC NILE prevail. The analysis of these coins has different aims. The dates they offer for the stratigraphic units are important, but they are also analyzed in their own right, as sources for minting policy, monetary circulation and its regions, and for economic change. Comparison to the few other places where coins have been noted, such as Elephantine, Abu Mina and parts of Alexandria, is informative in this respect.

phorae make up approximately 2/3 or 3/4 of the assemblages, which is a characteristic percentage for Mediterranean sites that engage actively in commercial exchange, while in the latter amphorae are attested at about 1/3. onsideration of some of the s e ifi f n tional gro s onfirms the differen e etween the earlier and the later la ers he hara teristi s of the fine wares and am horae vary markedly between the two periods. There are so few lamps that it is not possible to speak of trends in them. The utilitarian wares (coarse and cooking wares, which at Schedia appear in the same fabric) are essentially of local production throughout, and their typological development remains to be determined.

P ottery At Schedia the programme of pottery study aims at as complete a documentation as possible. The protocol used is predicated on the idea that all fragments, even typologiall nidentifia le od sherds, ha e some information to give. Therefore, all the sherds per stratigraphic unit are attributed to a functional group and fabric, counted and weighed. For rim sherds the percentage of the diameter preserved is also registered, in order to calculate estimated vessel equivalents. Diagnostic fragments are, of course, lassified t ologi all , and hronologi all sensiti e items are used for dating purposes. This degree of quantifi ation ermits o r material to e om ared with other assemblages throughout the Roman world, where various standards prevail. It also allows us to make statistical considerations at various levels (Martin 2008; Martin forthcoming).

mong the fine wares in the earlier la ers it is noti ea le that hedia did not arti i ate in the fine-ware oom of the Augustan period, when most parts of the Roman world wanted to im ort or rod e red-gloss wares refle ting the repertoire of Italian sigil l ata, that is to present themselves as Roman at the table (Martin 2008: 268). There are some very occasional imports of Eastern Sigillata A and less frequently yet of Italian sigillata, but for the most part Schedia seems to continue to employ red and black-slipped wares in the Hellenistic tradition. In the late-antique layers a certain quantity of Cypriot and also of African Red-Slip Ware appears, and a larger amount of Egyptian products inspired by them is attested. It is only then that one can say that Schedia participates in the Mediterranean-wide trends in fine wares

first le el on erns the o erall om osition of the assemblages by functional groups. There is some evidence that the percentage in particular of transport amphorae with res e t to fine ta le wares, lam s, ooking and oarse wares, refle ts the degree of integration of a site into trade networks. Here Schedia shows a decided difference between the late-antique layers up to the abandonment no earlier than the 7th century and the earlier ones. In the former, am-

In the earlier levels Egyptian amphorae greatly outnumber imported ones – the majority were the Amphore Egyptienne 3 t e whi h were first de elo ed at the eginning of the Empire. Other examples included the earlier Amphore

F ig. 1 1: K om el H amam. P l an of A rea 8 w ith basins for w ine- production. G erman Mission at K om el G iz a/B eheira.

115

LAKE MAREOTIS: RECONSTRUCTING THE PAST Egyptienne 2 and the contemporary Amphore Egyptienne 4 types (Empereur 1998: 79). They appear most frequently in the same fabric as the utilitarian vessels, decidedly more rarely in the northwest coastal fabric associated with production around Maryut, and only occasionally in the Nile silt fabric. Imported amphorae are represented by a diffuse scatter of various Aegean and Levantine pieces. In late antiquity, on the other hand, the provenience of the amphorae is much more focused and the percentage of imported vessels much higher. Egyptian vessels appear in the Nile silt fabric associated with Egloff 172 and the Carthage Late Roman Amphora 7, while the imports come almost exclusively from Cilicia with the Carthage Late Roman Amphora 1 and Gaza with the Carthage Late Roman Amphora 4.

During the 2nd and 3rd centuries the city reaches its maximum extension. That the brick-pillar building investigated here was transformed for residential purposes contrary to its original destination, can be considered a further indication of population pressure and of an increasing process of urbanization. Parts of the city reveal a strong Greek tradition (the Hellenistic bath and the stepped bases of gra e mon ments and oman infl en e v il l a suburbana). Other monuments, like the pillar building with its ontem orar se of fired and m d- ri ks show an interesting mixture of imported and local building traditions. At the same time, Schedia appears to have been a city with interesting cultural contrasts: while a prosperous upper lass li ed in ho ses of learl oman infl en e, a large number of inhabitants lived in simple dwellings of local Egyptian character.

In short, Schedia in the Imperial period presents, paradoxically for the major river harbour of Alexandria of which the written sources inform us, the picture of a site little integrated into the exchange networks and cultural trends of the time, while late-antique Schedia appears to participate more actively in them. How to explain this situation remains to be seen. Could it be that in the earlier period little needed to be offered in exchange for Egyptian products (e.g. grain collected in tax, stone from the Imperial quarries), while late-antique Egypt was in a similar situation to other provinces, with imports and exports. In any case, merit is due to the pottery study that serves to highlight a line of research to be pursued that might not have been highlighted in any other circumstance.

In late antiquity the explored parts of the city underwent a fundamental change. They seem to lose a large part of their urban character with the transformation of the older buildings into productive installations, of rural and other determination. At the same time the new installation of a granary in the pillar building coincides with the undiminished export of grain to Constantinople. How these changes are to be interpreted against the background of new discussions a o t the flo rishing agri lt re and renewed trade and imports of late antique Egypt (Banaji 2001; Kingsley & Decker 2001) and against the situation in Alexandria itself, remains to be seen. Although the new transformation of the pillar building into a large granary points to the still uninterrupted importance of the city as a supply base, it changes into a centre for agricultural production that possibly still has regional importance in supplying Alexandria but appears to a large extent to lose its role as an emporium, perhaps because of the overall decrease in trade. This hypothesis, which for now can take only the late Ptolemaic and Roman settlement of Schedia into account, needs to be checked by further investigation and especially to be complemented by a greater consideration of the Ptolemaic phases.

Conclusions The following preliminary hypotheses concerning the development of Schedia seem possible. After the foundation of Alexandria and the construction of the Canopic Canal the early Ptolemaic settlement of Schedia existed probably aro nd an artifi ial har o r asin at the starting oint of the anal e a se of the artifi ial ridge, the schedia’ , and a very dynamic development, the settlement, extended on both sides of the Canopic Nile towards the southeast. Here, on its southern borders, monumental and public buildings (baths and temples) were erected not later than middle and late Ptolemaic times. The settlement, therefore, already had urban character, which is also documented by the inscriptions.

Bibliography Abd el Fattah, A., 2000-2001, Brief report on excavations carried on at Kom el Giza during the season 1989. A nnal es du Serv ice des A ntiquité s de l ’ É gypte 76: 9-12. Abd el Fattah, A., 1998, Q uelques nouveaux moules alexandrins à Kom Giza. In A. Abd el-Abd el Fattah & P. Gallo (eds.), A egyptiaca A l ex andrina. Monuments pharaoniques dé couv erts ré cemment à A l ex andrie 1: 65-73. Alexandrina. Abd el Fattah, A., 1988, Recent discoveries in Alexandria and the Chora. In J.-Y . Empereur (ed.), Commerce et artisanat dans l ’ A l ex andrie hel l é nistique et romaine. A ctes du col l oque d’ A thè nes, organisé par l e CN R S, l e L aboratoire de cé ramol ogie de L yon et l ’ Ecol e franç aise d’ A thè nes, 1 1- 12 dé cembre 198: 47-53. Athens, Paris.

In the course of the Late Hellenistic and Early Roman periods a substantial necropolis with monumental tombs developed along the southeastern outskirts of the city, but because of the continuous growth of Schedia, luxurious residences for a prosperous upper class soon penetrated the periphery previously used for burials. At the same time on the southern embankment a monumental building, maybe a temple complex, as well as large storage facilities like the pillar building, arise, apparently as a consequence of massive investments. On one hand, they attest to the economic attractiveness of Schedia in the Imperial period, and, on the other, their distance from the harbour suggests that the more favourable zones were already built up. 116

M. BERGMANN, ET AL.: SCHEDIA ON THE CANOPIC NILE Banaji, J., 2001, A grarian Change in L ate A ntiquity. G ol d, L abour and A ristocratic D ominance. Oxford. Bergmann, B., & Heinzelmann, M., 2007, Schedia, Alexandrias Hafen am kanopischen Nil. Z wischenbericht zu den Arbeiten 2003-2007. H efte des A rchäol ogischen Seminars B ern 20: 65-77. Bergmann, B., & Heinzelmann, M., forthcoming, The tholos-bath at Schedia. In M.F. Boussac, T. Fournet & B. Redon (eds.), L e bain col l ectif en Egypte. D es bal aneia antiques aux hamams contemporains: origine, é v ol ution et actual ité des pratiques. A ctes du col l oque B al né orient d’ A l ex andrie 1- 4 D é cembre 206 . Alexandria. Bernand, A., 1966, A l ex andrie l a G rande. Paris. Bernand, A., 1970a, Alexandrian Canal. L e del ta é gyptien d’ aprè s l es tex tes grecs s s i yq s: 329-380. Cairo. Bernand, A., 1970b, Schedia/Menelais. L e del ta é gyptien d’ aprè s l es tex tes grecs s s i yq s: 381-442. Cairo. Botti, D.G., 1902, Studio sul III° nomo dell’ Egitto inferiore. B ul l etin de l a Socié té d’ A rché ol ogique d’ A l ex andrie 4: 41-61. Boussac, M.F., Fournet, T., & Redon, B. (eds.), forthcoming, L e bain col l ectif en Egypte. D es bal aneia antiques aux hamams contemporains: origine, é v ol ution et actual ité des pratiques. A ctes du col l oque B al né orient d’ A l ex andrie 1- 4 D é cembre 206. Alexandria. Daszewski, A.W., 1998, La nécropole de Marina el-Alamein. In S. Marchgay, M.-Th. Le Dinahet & J.-F. Salles (eds.), N é cropol es et pouv oir: idé ol ogies, pratiques et interpré tations. P roceedings of a conference L yon 195 : 229-241. Lyon. Empereur, J.-Y ., 1998, Les ateliers d’ amphores du lac Mariout. In J.-Y . Empereur (ed.), Commerce et artisanat dans l ’ A l ex andrie hel l é nistique et romaine. A ctes du col l oque d’ A thè nes, organisé par l e CN R S, l e L aboratoire de cé ramol ogie de L yon et l ’ Ecol e

franç aise d’ A thè nes, 1 1- 12 dé cembre 19 8: 75-91. Athens, Paris. Jomard, E.F., & Jacotin, P., 1818/25, Carte T opographique D e L ’ É gypte Et D e P l usieurs P arties D es P ays L imitrophes (1818/25), M 1: 100 000, Blatt Nr. 37. Kingsley, S., & Decker, M., 2001, New Rome, new theories on inter-regional exchange. In S. Kingsley & M. Decker (eds.), Economy and Ex change in the East Mediterranean during L ate A ntiquity. P roceedings of a conference at Somerv il l e : 1-27. Oxford. Col l ege. O x ford 29 th May 19 Martin, A., 2008, Pottery from Schedia near Alexandria (Egypt). R ei Cretariae R omanae F autorum A cta 40: 263-269. Martin, A., forthcoming, The pottery from a late-antique settlement at Schedia (Western Delta, Egypt). In S. Menchelli, M. Pasquinucci & S. Santoro (eds.), T he 3 rd I nternational Conference on L ate R oman Coarse W are, Cooking W are and A mphorae in the Mediterranean: A rchaeol ogy and A rchaeometry ( L R CW 3) ( P arma – P isa, 2530 M arch 208) . Oxford. Procopius, i iis, Vol. VII, H.B. Dewing (transl. & ed.), 1961. Cambridge, MA. RE 1899, Wissowa, G. (ed.), P aul ys R eal - Encycl opädi e der cl assischen A l tertumsw issenschaften, Vol. I 1, 2030 s.v. Χαιρέου (Sethe). Stuttgart. RE 1921, Wissowa, G., Kroll, W., & Witte, K. (eds.), P aul ys R eal - Encycl opädi e der Cl assischen A l tertumsw issen- schaft, Vol. II A 2, 401-403 s.v. Schedia (Kees). Stuttgart. od iewi , , 1988, lassifi ation of wineries from Mareotis. In J.-Y . Empereur (ed.), Commerce et artisanat dans l ’ A l ex andrie hel l é nistique et romaine. A ctes du col l oque d’ A thè nes, organisé par l e CN R S, l e L aboratoire de cé ramol ogie de L yon et l ’ Ecol e franç aise d’ A thè nes, 1 1- 12 dé cembre 19 8: 27-36. Athens, Paris. Strabo, G eography, Vol. VIII, H.L. Jones (transl. & ed.), 1959. Cambridge, MA.

117

LAKE MAREOTIS: RECONSTRUCTING THE PAST

118

P. WILSON: SURVEY WORK IN MAREOTIS

Recent Survey Work in the Southern Mareotis Area Penelope Wilson

Introduction The ‘ Sais and its Hinterland Project’ based at Sa el-Hagar, Sais in Gharbiyah Province, was designed partly to determine the impact of river distributary systems on settlement in the north and north-west of Egypt. In order to compare the Canopic with the Rosetta riverine systems, 70 sites were surveyed and visited in Beheira and Kafr esh-Sheikh Provinces. This paper describes and discusses sites in the southern Mareotis area, which formed an interesting group in themsel es he re resent la es infl en ed another type of water-system, an inland lake located at sea-level, that is fed by a series of distributaries from the Canopic s stem he lo ation and e tent of ake areotis fl t ated o er time and this ma ha e een an infl en ing factor in the life-span recorded at some of the sites in the southern Mareotis area.

sides. The lower parts of its sides have been left largely untouched, but are used for animal husbandry. The village of Kom Ishu was formerly on the western, southern and eastern sides of the hill but has developed in linear streets radiating to the east. It is possible that the original heart of the village was sheltered from the north wind by the hill he field o ndaries e ond the illage s ggest that the kom may have covered a larger area at one time. The whole of the upper surface of the hill is now covered by a modern cemetery. Pottery gathered at the site is mostly of Late Roman date.

K om I shu and K om el - H agg The two sites share a similar geological nature as they both consist of rocky limestone outcrops at the edge of a limestone ridge formation and seem to have been the focus of human activity in Lake Mareotis (Embabi 2004: 260-2).

Kom el-Hagg (SCA 100155) is surrounded by a small ez bet, called Ezbet el-Bank. The mound has a sloping, rounded edge to the east and the other sides are slightly steeper at the edges. The area is used for stabling animals and as a manure store. The southern edge of the kom has een t awa , lea ing a metre-high se tion a o e field le el ontaining otter fragments he fields to the south of the main mound are covered in sherds and may have been part of the ancient site, perhaps representing a settlement against the rocky outcrop. It is also possible that the sherds to the south were washed or swept down off the rocky hill. It would be interesting, therefore, to define the e a t nat re of the site at this oint e a se of its location and the apparent use of the local geological features as at Kom Ishu and also at Kom Lemsan, situated to the north.

Kom Ishu is approximately 12 km west-south-west of Kafr ad Dawar. The rocky outcrop has steep, almost sheer,

The pottery sample from the site gave a broadly consistent date range from the Ptolemaic through to the Early Roman

T he S ites (Figs. 1-2) The sites are discussed in an anti-clockwise direction around Lake Mareotis. Although the pottery collection and survey are of a preliminary nature, some interesting observations can be made arising from a study of the results.

F ig. 1: A ncient sites in the D el ta, hi ghl ighting the Mareotis region ( draw n by P . W il son) .

119

LAKE MAREOTIS: RECONSTRUCTING THE PAST

F ig. 2: 1: 50,

Map of the sites surv eyed, w ith 0 m etre sea- l ev el contour area highl ighted in grey. B y R . D ickinson, af ter Survey of Egypt maps ( 197) .

120

P. WILSON: SURVEY WORK IN MAREOTIS periods. The different types of vessel from the site included both amphorae and domestic vessels and so did not indicate a particular function for the site, but there certainly seems to have been some kind of settlement here. Most of the pottery was Egyptian in origin, and presumably was produced locally, with only one of the amphorae of Early Ptolemaic date perhaps from Rhodes, possibly a reused wine or water container. Both Kom el-Hagg and Kom Ishu are situated on the western edge of the main body of Lake Mareotis and could have had strategic value in guarding the lake shores or in watching over the desert areas to the west, the river to the so th and marshes or fields on all sides ernand 1970 866). Although the two places are within signalling distance, the pottery represents two distinct phases, with that from Kom el-Hagg covering the Ptolemaic and Early Roman periods, while that at Kom Ishu is of the Later Roman period. There may be a number of reasons for this, mostly due to the nature of the collection of the pottery sample. But it could be that Kom Ishu replaced Kom el-Hagg as a watching place or simply that it continued for a longer period of time, perhaps because it was situated upon a more long-lived canal which later became the Tirat Abis. In such a case, Kom Ishu had a better strategic value in the later period because of its water channel communication.

F ig. 3: by P . W

T he main mound at K om el - Mahar ( photograph il son) .

2nd century. Most of the pottery sample consists of tablewares, with some im orted finewares and glass, and ma have come from domestic contexts. Kom el-Mahar could have been a small town or village focussed on a series of elite houses and villas on the southern shore of the lake, taking advantage of the elevated land (Botti 1902: 55).

K om el - Mahar and Sidi G haz i Kom el-Mahar (SCA 100167) lies isolated in farmland to the east of the Nubariyah Canal and south-west of Sidi Ghazi village. The main mound is covered in pottery, inl ding finewares ig 3 he name of the site in ra i means ‘ shell’ and shelly deposits are apparent all over the mound and in the sections. The shell derives from the fact that the mud-brick used to build the houses of the town was made from silt mixed with crushed shell from degraded, local oolitic limestone. When the mud-brick disintegrated, a large amount of shell was left lying upon the surface.

The modern village of Sidi Ghazi (SCA 100142) lies approximately 10 km south of Kafr ad Dawar and to the north of the Masraf al-Umum Canal which runs to the west of the asla nin gro of tells here is now no tell at the location of Sidi Ghazi, but modern maps indicate that the village lies upon a small area of raised ground. It is likely that the original tell has been subsumed under the village. A site here may have formed one of the islands inside Lake Mareotis. South- w estern Sites Kom el-Gel (SCA 100217) is the site at Ezbet Abd el-Q adir at-Tawil and lies approximately 17 km north-west of atamir he site onsists of a mo nd with a flat to and gently sloping northern side and a secondary mound to the east. The top of the mound is covered with a modern cemetery, while the sides are used for animal husbandry. he ho ses on the western side of the tell are ilt on flat land at the ase of the hill he fields to the so th of the mo nd ontain noti ea le antities of otter and some limestone fragments, suggesting that this was once a more sizeable mound. Satellite imagery suggests that the area to the south may have been part of the mound because of the sha e of the fields

The lower slopes of the eastern and south-eastern side of the mound are covered by a modern cemetery. The northern and north-western side of the mound has sheer sides, where they have been dug away and it is likely that the original site extended in all directions away from the mound. It is now surrounded by deep irrigation ditches and fields, whi h are en roa hing on the o ter arts of the site. One of the northern sections contains a substantial, stratified se en e, in l ding ashlar limestone lo ks, a wall made of sand m d- ri ks and shell-filled, m dbricks. There are also fragments of red granite on the mound and glass fragments lying upon the surface of the site he so thern area of the site has een flattened and there are building outlines visible on the surface. There are some lower, smaller mounds covered in scrub to the northwest and south-east sides of the site.

he otter sherds from the site om rised fine wares, domestic material and amphorae. The pottery dated mainly to the Late Ptolemaic period, with some Early and Late Roman sherds included in the collection.

The main pottery component from the site ranges from Late Ptolemaic to Early Roman in date (Fig. 4), suggesting that it may be a Ptolemaic foundation continuing as a town into the Roman period, before being abandoned around the

The original site at Kom el-Farag lies under a modern town, a ro imatel 13 km north-west of atamir t may be the location of Kom el-Hanache (Toussoun 1922: 121

LAKE MAREOTIS: RECONSTRUCTING THE PAST F ig. 4: L ate P tol emaic to Earl y R oman pottery sampl e from K om el - Mahar ( draw ings and anal ysis by D . G rigoropoul os) .

lake next to the village which may be a relic of an ancient river distributary or ox-bow lake, extending for about 500 m in a north-west to south-east orientation. The lake may link with some ancient depressions in the landscape to the west or have formed one of the distributary channels from the Canopic Branch. If there was once an ancient settlement here, it may have been located upon the river, lying along a raised levee as with other tells in this part of Beheira.

Pl.II), but as the Ezbet Farag grew to the south, the tell seems to have been subsumed underneath the houses and removed by the Delta Light Railway. The cemetery part of Kom el-Farag lies to the north-west and may be a remnant of the older mound, being a maximum of 1.5 m above ground level. The old part of the village upon the mound has a less organised street plan than the more recent areas of the town with their orthogonal layout. The interest of this location is that it lies on the Tirat Abdel al-Fatah and Tirat al-Hajar drainage channels and, perhaps, on the channel of one of the earlier Canopic distributaries on the western side of Beheira. Kom Farag is also at the centre of a cluster of small villages and ez bets with the word kom in their name, such as Kom es-Siwayd to the south, Kom Kifri, as Sa’ ayidah and Jaradah to the south-west and to the north-west Ezbet Abd el-Q adir atTawil (Kom Makhboura), which has a cemetery upon a kom. They may well be part of a town cluster focussed on an irrigation basin network dating at least to the medieval period.

Abu Guduur (SCA 100212) is 2.5 km north-west of Abu Matamir and lies to the south of the main road, upon a low ridge of silt along an ancient distributary course, at the edge of the Mareotis basin. It once covered an extensive area, but now consists of some low, sandy mounds covered in s r and a larger flat area with s r and shes Most of the site is now used as a municipal rubbish dump, having once also been a gravel extraction and processing area. The highest mound is only about 2 m above the level of the surrounding land and the whole area is very sandy. There was pottery lying on the tell as well as some bronze o e ts, in l ding a stirr ring and a three-flanged arrowhead. Both could date to the Late Dynastic period. A small diagnostic set of pottery fragments were collected from the site. As far as can be judged, they mostly date to the Late Ptolemaic period, but there were some Late Roman sherds and some Late Dynastic pottery as well.

South of the L ake Kom el-Adda 100230 or om l- ida lies 8 km north-west of Abu Matamir. Although there is a cemetery mound to the south-east of the main village, the main interest of the area is the curious crescent-shaped, enclosed 122

P. WILSON: SURVEY WORK IN MAREOTIS The place may have been a strategic Dynastic location for a small military establishment related to the other western delta fortified sites and to other mo nds r nning from Matamir to the south east. Southern P ort on the N aukratis Canal Kom Trugi or Turuga (SCA 100251) lies 6 km due north of Abu Matamir, close to several other places with kom in their name and on the former edge of Lake Mareotis. Turuga is known to have been an important inland port on the lake for people going from Alexandria to the Wadi elNatrun and for goods and supplies coming out of the Western Desert and delta towards Alexandria for trade (Timm 1984-1992 2545-6 e osson 1935 79, 151 t is ossible that the site had two halves on either side of the waterway into the lake, one known as Psenemphaia, according to a stela from the site ernand 1970 899-913 and the other known as Theroge or Therange, which lends its name to the whole site after the Ptolemaic period. Roman period material was noted at the site in 1895 by D.G. Hogarth en er 1999 305, 308-11 and otti noted that mar le column capitals could been seen there and that a headless white marble statue was found at Kom Trugi (Botti 1902: 58). Kom Trugi suffered greatly from sebakhin digging (De osson 1936 , with the a thorities onl inter ening at the discovery of important objects or remains such as a Ptolemai - oman ath-ho se l- hasha 1957 127-139

F ig. 5: R ed brick structures at K om T rugi, e x cav ated bel ow foundation l ev el . T he scal e is 2 m etres high ( photograph by P . W il son) . F ig. 6: Section through the cemetery mound at K om el Q adi ( photograph by P . W il son) .

It is clear that this was once an extensive site and perhaps a high mo nd t now onsists of flat, sand areas interspersed with several excavated red brick structures and limestone upon mounds. To the north of the area, there is a modern village with an adjoining cemetery and both may have been built upon part of the original site. Amongst the extant red brick structures there is part of a pavement left standing upon its sand-brick foundation and a pillar of earth, as well as red brick and plastered tanks, perhaps from a winery. There is also a fragment of a limestone wall consisting of fi e o rses of ashlar lo ks for a length of 4 75 m, with some plaster still adhering to one side (Fig. 5). All of this area seems to ha e een d g o t to elow the floor level of these buildings and the site seems very denuded. On one of the mounds to the south there are the foundation walls of a limestone ilding he flat, sand area etween the mounds shows building plans in wet weather, according to the local people, but is now mostly used as a football pitch. There are also a number of red granite grindstones at the site, l ing on the gro nd ltho gh it is diffi lt to make sense of the standing remains, they do suggest that an impressive town with large red brick and limestone buildings once stood here.

the medieval period. The visible levels date perhaps to the Early and Middle Roman periods and information about the later levels may be lost. The height of the mound suggests that there is still buried material at the site, but that it is disappearing fast. N orth- east of the L ake Kom el-Q adi (SCA 100165), south of the main Cairo to Alexandria road, and Kafr ad Dawar may be part of a cluster of sites aro nd l asla n, forming a townshi or irrigation basin group of sites. The sprawling area at Kom el-Q adi has several distinct zones and a main archaeological area to the south of Kom el-Q adi village. It consists of a high mound, around 8 m high, the southern and northern faces having been cut away to form sheer sections. The so thern se tions show signifi ant stratified se en es, including sand-brick walls, pottery deposits and human rials in oth otter offins and ri k a lts k lls of adults and children can be seen in the sections and on top of the mound (Fig. 6). The remains may be interpreted as

The pottery sample collected from the site dates mainly from the Early Roman period, but there are also some Ptolemaic, Middle-Late Roman and Medieval Arab period sherds in the collected sample. Because of the disturbed nat re of the site, howe er, it is diffi lt to asso iate arti lar otter with s e ifi areas of the site he range of the pottery agrees well with the historical material, suggesting that the site was used from the Ptolemaic through to 123

LAKE MAREOTIS: RECONSTRUCTING THE PAST S ummary (Table 1) Lake Mareotis created its own special topography and lands a e, with the shallow flooded area ordered reeds and rushes forcing people onto the high ground around the lake edge, islands in the lake and between the distributaries to the south and the Canopic and Alexandria Canal to the north. The strange, thin tongues of land, with their dykes and paths, must have made the desert routes seem attractive to travellers and so created opportunities on the Western Desert edges for monastic hostels and stopping places. The topography of the area shows that the presence of the lake had a profound impact upon social organisation and economic management in the North-West Delta. Pharaonic period attempts to control the Western Desert edges may always have led to limited settlement in the north-west area, as suggested by Abu Guduur and the fortress line in the Ramesside period (Habachi 1980). The foundation of Alexandria seems to have acted like a magnet, attracting service settlements and industries upon a huge scale, as well as enabling the development of agricultural lands around the city and bringing settlement in favourable locations. The preliminary dating evidence from the sites in Beheira shows that the earliest material at the surveyed sites in the majority of cases comes from the Ptolemaic period, contemporary with the development of le andria ilson rigoro o los 2009 42-43 he sites at Kom el-Mahar, Kom el-Gel and Kom Trugi may all be part of this wave of development, with the creation of farming towns, a process which continued into the Early Roman period. This seems to have been a busy time for the lake, with many settlements established either as internal trans-shipping ports for the movement of goods and people

a reuse of the abandoned town mound as a cemetery. This may be the Roman cemetery from which three inscriptions were published (Botti 1902: 57). On the eastern side of this mound a substantial mud-brick wall is visible in the section. In the village to the west, there is a Roman bath-house, which has been restored recently and includes art of the on rete floor of a thol os, some possible underfloor f rna es and re tang lar tanks ore n o ered red brick chambers as well as surface building plans, can be seen on a sandy area to the east of the village and main mound. A further sandy area lying to the east of this part of the village and a patch of extensive low scrub to the north-east of the village may have been part of the original site. There were some larger fragments of limestone lying in the mounds and the village. The pottery collected was mostly from the mound. It consisted of Early and Middle Roman material, including some im orted fine wares, man ooking ots whi h showed evidence of burning on the outside and some Egyptian amphorae. The material seems to indicate the domestic nature of the site in the Roman period, but the relationship of the cemetery remains to the settlements was not clear from a preliminary examination of the mound sections. The site seems to be a small town and its associated cemetery, perhaps dating from the height of Roman management of this area. T abl e 1: Summary tabl e of sites surv eyed by the ‘ Sais and its H interl and P roj ect’ in the region of southern L ake Mareotis. Site & current dimensions

Early Roman Ptolemaic Early Late 1st c. BC rd st 3 -late 1 c. BC to 2nd c. AD

Middle Roman Late Roman 3rd-4th c. AD 5th-end 7th c. AD

Kom Ishu 150 x 200 x 10 m

Early medieval Arab 8th-16th c. AD

X

FUNCTION Watch post

Kom el-Hagg 120 x 200 x 4 m

X

X

Watch post

Kom el-Mahar 440 x 240 x 12 m

X

X

Main Ptol-Roman town and villa site

Sidi Ghazi Under town Kom el-Gel 270 x 220 x 4 m

Island village X late

X

Town/villa

Kom el-Farag Under town

Farming village

Kom el-Adda 220 300 3 m

Riverside town

Abu Guduur 300 250 2 m

X late

Kom Trugi 350 300 3 m

X

X late X

X

Kom Hassan nder fields Kom el-Q adi 370 250 8 m

Fort? , riverside town X

Inland harbour town Unclear

X

X

Main Roman town

124

P. WILSON: SURVEY WORK IN MAREOTIS Bibliography SCA numbers refer to site registration numbers of the Supreme Council of A ntiquities, E gypt.

to and from Alexandria, or as points of control around the lake, such as at Kom Ishu, Kom el-Hagg and perhaps Sidi Ghazi. The apparent abandonment of places, such as Kom el-Mahar and Kom el-Gel, may be part of the ‘ Antonine Plague’ effect with depopulation of some areas (Bagnall rier 1994 173-8 n the later eriods, the la k of ate Roman and Early Arab continuity at the sites suggests that, in common with other parts of Beheira province, once the Canopic Branch of the Nile and its distributaries had silted up, the sustainability of some of the settlements was not possible and that people moved away from earlier settled land. They may have gone to a few major centres such as Abu Matamir or Dilingat or Damanhur or Alexandria itself, rather than carrying on trying to eke out an existence on the shores of the Mareotis Lake.

Bagnall, R.S., & Frier, B.W., 1994, T he D emography of R oman Egypt. Cambridge. Bernand, A., 1970, L e D el ta é gyptien d’ aprè s l es tex tes grecques, I-III. Cairo. Botti, G., 1902, Studio sul IIIo nomo dell’ egitto inferiore: e più specialmente sulla Regione Mareotica. B ul l etin de l a societé royal e d’ arché ol ogie - A l ex andrie 4: 4184. e osson, , 1935, Mareotis: B eing an A ccount of the H istory and A ncient Monuments of the N orth- W est D esert of Egypt and L ake Mareotis. London. e osson, , 1936, l arn gi, B ul l etin de l a societé royal e d’ archaeol ogie - A l ex andrie 31 113-116 Daressy, G., 1929, Ménélais et l’ embouchure de la branhe ano i e R ev ue de l ’ É gypte ancienne 2: 20-51, Pls. I-II. Embabi, N.S., 2004, T he G eomorphol ogy of Egypt. L andforms and Ev ol ution. Vol. I. T he N il e V al l ey and the W estern D esert. Cairo. El-Khachab, A.M., 1957, Les Hammams du Kom Trougah. A nnal e du Serv ice des A ntiquité s de l ’ Egypte 54: 117-139 Habachi, L., 1980, The Military Posts of Ramesses II on the Coastal Road and the Western Part of the Delta. B ul l etin de l ’ I nstitut franç ais d’ arché ol ogie oriental e 80 13-30 Spencer, P., 1999, Hogarth’ s 1895 Report to the Committee of the Egypt Exploration Fund. In W.J. Tait & A. Leahy (eds.), Studies on A ncient Egypt in H onour of H .S. Smith 302-311 ondon Surv ey of Egypt 1997. 1:50,000 Maps; “ Abu al-Matamir” , “ Kafr ad-Dawwar” , “ Ikinij Maryut” & “ al-Iskan dariyyah” . Survey of Egypt/Ministry of Finances, Cairo. Timm, S., 1984-1992, D as christl ich- koptische Ä gypten in arabischer Z eit: Eine Samml ung christl icher Syät ten in Ä gypten in arabischer Z eit, 6 Vols. Wiesbaden. Toussoun, O., 1922, Mémoire sur les anciennes branches du Nil. Mé moires de l ’ I nstitut d’ É gypte 4, Premier fascicle, Cairo Wilson, P., & Grigoropolous, D., 2009, T he W est N il e D el ta R egional Surv ey, B eheira and K afr esh- Sheikh P rov inces. London.

Of the sites surveyed here, Kom Trugi was the largest and most important, perhaps straddling the distributary of the Canopic Branch from Naukratis and providing a secondary direct link from the south to Alexandria and from the Western Delta area to Alexandria. Kom el-Q adi seems to have been a typical Early Roman town with a bath-house, similar to other new towns in the Roman delta, and a substantial mound used as a cemetery. It may have been founded in the Early Roman period at around the same time as hedia and art of the l ster aro nd asla nin, on e s ggested to e the lo ation of enelais aressy 1929). Further detailed work in any or all of the Mareotis catchment area would be highly desirable to explore riverine, lacustrine and human interaction during periods of profound cultural change. Ack now ledgements I thank Dr Dimitrios Grigoropoulos of the Deutsches Archä ologisches Institut – Athens who studied the pottery from the sites and Dr Roger Dickinson for the mapping. am also gratef l to the reme o n il of nti ities in Damanhur under Fawzy Fawzy el-Kholi and Naama Suleiman, as well as Z ahi Hawass and Magdy Ghandour in airo or the fieldwork, thank r oanne owland, Mohamed Osman, Ashraf abd el Rahman and Ahmed Bilal. The Department of Archaeology, Durham University and Egypt Exploration Society also supported the work. Funding from the Arts and Humanities Research Council, UK is acknowledged. The survey is now published by the Egypt Exploration Society’ s Delta Survey Project under the title T he W est N il e D el ta R egional Surv ey, B eheira and K afr esh Sheikh P rov inces.

125

LAKE MAREOTIS: RECONSTRUCTING THE PAST

126

D. DZ IERZ BICKA: WINERIES OF THE MAREOTIC REGION

Wineries of the Mareotic Region Dorota Dzierzbicka

Sea, it was by far the best in the Nile Valley. Sources from the Pharaonic and Persian periods testify to wine production in this area long before the coming of the Greeks (cf. Forbes 1955: 70-124). The wineries date from the Roman and Byzantine periods. Some are of considerable size and testif to e tensi e lti ation of gra es and large s ale wine production in this area, especially in correlation with the local pottery kilns which indicate a mass production of amphorae (Empereur 1993: 39-47).

Wineries or remains of wine making installations have been found in various parts of the Nile Valley: the coastal areas and the Delta, the Fayum and Bahariya Oasis, as well as Upper Egypt (Dzierzbicka 2005: 11-24). However, their concentration in the Mareotic region, especially along the southern shore of the lake and around Abu Mina, is unmatched in other parts of the country (Fig. 1). n addition to the identified wineries, there are also elements of wine-making installations that are no longer in situ – primarily lion headed spouts through which freshly ressed or trodden m st flowed into the at mero s e amples, including those found around Lake Maryut, can be seen in le andria in the re o- oman se m or at the Kom el-Shukaqa necropolis (Rodziewicz 1998: 28, n. 6).

L iterary S ources In literary sources wine produced in the Mareotic region during the Greco-Roman Period is well attested. It was known to Horace (O des I 37.14), who wrote that the mind of Cleopatra was made mad by Mareotic wine (mens l ymphata Mareotico). According to Strabo (XVII 1,14) the Mareotic region produced good wine and in such quantity that it was racked to be matured. The white Mareotic grape variety (Mareotides al bæ ) is mentioned by Virgil in G eorgica (G eor. ii. 91). Athenaeus (D eip. I 33 d-e), in turn

Many factors contributed to the large concentration of wineries in the vicinity of Lake Maryut. Although in antiquity the limate in the elta ma ha e not een as enefi ial for vineyards as on the northern shores of the Mediterranean

F ig. 1: A map of the Mareotic region w ith marked l ocations of w ineries. P l otted on the map are ex cav ated and publ ished structures, as w el l as unex cav ated ones mentioned by schol ars in other w orks or featured on SCA maps as protected archaeol ogical sites. T he w ineries designated as N N w ere pl otted by R odz iew icz ( 198: 28, F ig. 1) , but the name of their l ocation w as not prov ided ( D . D z ierz bicka) .

127

LAKE MAREOTIS: RECONSTRUCTING THE PAST laims that areoti wine was e ellent it was white and pleasant, fragrant, easily assimilated, thin, did not go to the head and was diuretic. According to this author better yet was the wine from Taenia, from the same region. This was pale yellow and had an oily quality, which disappeared after it had been diluted with water. Besides being sweet, this wine was rather aromatic and mildly astringent.

There are three other wine making installations in the vicinity of the modern village of Huwariya. One of them is a simple Early Roman Type I unit built of dressed stone lo ks o nd with mortar ne lored, mentioned Rodziewicz 1998: 31). A winery in southern Huwariya (reported by Rodziewicz 1998: 34), built of irregular stone blocks, clay and waterproof plaster, consisted of one press nit a treading latform and at onne ted two flights of steps. The round base of a mechanical press was set in a niche in the wall of the treading platform. A channel in the floor of the latform led from the ress dire tl to the vat. A third installation, located 3 km southeast of Huwariya (Rodziewicz 1998: 35), consisted of a treading platform, a vat and a mechanical press separated from the treading floor a low, thin wall hannel nder the floor led from the me hani al ress to the at he walls of the structure were built of irregular stone blocks bonded with mortar and covered with plaster reinforced with potsherds.

W ineries1 T he South Coast of L ake Maryut The wineries in the immediate vicinity of Lake Mareotis are located on the strip of land between the lake and the Mediterranean Sea, and along the southern shore of the lake. The majority are found on the southern shore of the lake. A large winery near Burg el-Arab (30o55’ 37” N 29o 31’ 48.50” E),2 is located in the northeast corner of a large building (El-Ashmawi 1998: 55-73). A mechanical press was set up on the treading platform and closed off from the surrounding area by a thin wall. The must from the me hani al ress en los re and from the treading floor flowed to the ad a ent olle tion asin thro gh two se arate channels.

A small, presumably Roman winery of George Nesim (30° 57’ 18.30” N 29° 36’ 15.20” E) was found by the Lake Mareotis Research Project (Blue & Khalil forthcoming). It appears to have consisted of a treading platform and at e od iewi 1998 31 , t the resen e of a mechanical press unit with a separate, small collection asin within the same om le annot e e l ded, although it requires further investigation.

The winery located at Taher el-Masry (30o56’ 59.40” N 29o34’ 58.20” E)3 was built of stone blocks and stone rubble embedded in mortar. The uncovered press unit consisted of a s are treading latform 3 75 m 3 75 m and a at 2 37 m 2 00 m , 1 40 m dee tlines of str t res s rrounding the unit suggest that the winery may have been art of some kind of a larger om le

T he N orth Coast of L ake Maryut To the north of the lake, at Abu Talaat (unpublished, mentioned by Rodziewicz 1998: 29), an Egyptian archaeological mission uncovered a winery with two connected treading platforms and one collecting vat. The wine-making nit was entirel hewn o t of the edro k e e t for one wall, which was built of stone blocks.

A restored wine-making unit near Marea (30o58’ 49.10” N 29o40’ 12.90” E) consisted of a large treading platform, a smaller room with a base for a mechanical press, and a collection vat (El-Fakharani 1983: 182-184). Two separate channels led from the two rooms to the vat. The rooms where the pressing took place were raised higher than the room of the at and two flights of ste s onne ted the two levels. Nearby, on the same site another unpublished winemaking unit was found.4

The winery at Site 215 recorded by the Lake Mareotis Research Project (30° 53’ 30” N 29° 25’ 22” E) has two parallel treading platforms visible on the surface. The channels in their western sides indicate the presence of two collection basins, which have not yet been uncovered. Thus, this winery represents a compound of two Type I wineries according to the typology established by Rodziewicz (1998: 31-35). Further investigation is needed to identify its other features.

1. This part of the paper includes elements previously published in Dzierzbicka 2005. 2. GPS coordinates of the wineries as well as some additional information – di mensions, architectural details – w ere collected on my study tour of the Mareotic wineries in November 2005 during my scholarship at the Polish Centre of Mediterranean Archaeology in Cairo, Egypt. The study tour included the wineries near Burg el-Arab (also called Abu Sir), at Taher el-Masry and Marea, two of the installations at Abu Mina, and the wineries at Karm el-Shewelhy and Karm el-Baraasi. I am thankful to Mr. Adli Rushdy from the West Delta Inspectorate and to Mr. Saber Selim from the Islamic Inspectorate of the Supreme Council of Antiquities for allowing me to see the sites. For GPS locations and photographs of the wineries at George Nesim, Kom Trouga and Site 215 I am greatly indebted to Dr. Lucy Blue and Dr. Emad Khalil of the Lake Mareotis Research Project. 3 lored r stafa shd of the g tian nti ities rganisation in 1998/99 (as it was communicated to me by Mr. Maged Ahmed, Inspector of the SCA), unpublished. The survey I conducted in 2005 yielded the GPS coordinates of the site, as well as some preliminary observations based on the remains visible on the surface. 4. I am thankful to Prof. Mona Haggag for bringing it to my attention.

A rea to the East of L ake Maryut Much fewer wineries have been located in the eastern part of the Mareotis region. A winery was uncovered in 1917 aninos asha near ir a east of le andria (published as a private bath in Breccia 1926: 47-49). The Ptolemaic dating established by Breccia on the basis of the quality of plaster used in the structure seems rather uncertain large treading latform 3 25 m 5 80 m o ered with water roof laster was raised 0 30 m a o e the floor level of the building. Two channels led through a wall separating the platform from a collection basin 1.60 m deep, 4 20 m long and 2 20 m wide win flights of ste s onnected the two parts of the winery. 128

D. DZ IERZ BICKA: WINERIES OF THE MAREOTIC REGION There is also a Roman Period winery at Kom Trouga (30° 57’ 46.90” N, 30° 10’ 27.10” E). This unpublished winery mentioned by El-Fakharani (1983: 184) and Rodziewi 1998 31 was a om le that onsisted of se eral scattered wine making units, today badly damaged due to quarrying activity in the area. Remains of basins and adjacent raised treading platforms are visible. All of the units appear to have consisted of a single treading platform and vat, covered with waterproof plaster. The best preserved one was built of stone blocks, while others were brick structures. Possible wine-making installations have also been found at Schedia.5

two treading platforms, two vats, one mechanical press, and storage rooms. The structure was built of mud-brick and re-used stone blocks. The winery located at Karm el-Shewelhy (30o49’ 37.70” N, 29o38’ 65” E) consisted of a treading platform and vat, two storage rooms, and a courtyard. The wine-making unit was hewn in ro k e e t for the eastern wall, whi h was ilt of limestone lo ks he floor of the treading latform and the walls of the vat were covered with waterproof plaster (Abd el-Aziz Negm 1998: 73). A sizeable winery at Karm el-Baraasi (30o50’ 41.40” N, 29o40’ 49.50” E) consisted of several large treading latforms with ats and me hani al resses he e avations published by Abd el-Aziz Negm (1998: 63-70) brought to light a large treading platform and vat. In the same room as the vat there were two small enclosures (4 m2 each) for mechanical presses, each equipped with a small vat of its own. In November 2005 another treading platform was visible semicircular in shape, with an adjacent vat. This unit was located further to the west and separated from the first latform a room he floor le el of this room was lower than the two treading platforms and onne ted to them two short flights of ste s rther to the west traces of three more treading platforms were isi le not et e a ated he om le was ilt of m dbrick with corners reinforced with limestone blocks. The entran e to the nit n o ered the e a ators led from a o rt ard he winer was art of a larger om le across the courtyard from the winery there was a house, and the courtyard itself was surrounded by rooms. Around 300 m to the north-west of this om le 30o50’ 50.70” N, 29o40’ 40.20” E) there are probably more installations yet to be uncovered, as one can tell from the outlines of basins covered with waterproof plaster that are discernible on the surface.

T he R egion of A bu Mina A number of wineries were found in the vicinity of Abu Mina, not far from Lake Mareotis. The region was a thriving grape-growing and wine-making centre. There are three wineries within the town area. A large winery uncovered in 1907 by Kauffmann to the east of the double bath (Grossmann, et al 1984 148 rossmann, et al 1997 87-90 ehind the portico of the colonnaded street (30o50’ 52.90” N, 29o39 76 10 was e anded in fi e hases ntil it e ame a om le of fi e treading floors with fo r mechanical presses. It was built of mud-brick and stone blocks covered with waterproof plaster. Another winery was located behind the apse of the basilica (30o50’ 45.80” N 29o39 82 30 t onsisted of two rooms, the first ho sing one large treading latform flanked two me hanical presses, a collection vat, and one smaller raised platform n the se ond room there was another treading floor and at rossmann 1999 82-83, fig 9 he third winery (Mül ler-Wiener & Grossmann 1967: 468-473) within the area of Abu Mina was uncovered in the so-called K umring A , in three rooms of Building 17, entered from a courtyard. Two of these rooms each contained a set of one large and one small treading platform and vat. The third room on the opposite side of the courtyard was used for storage.

E lements of M areotic W ineries The Mareotic evidence provides an abundant and diverse range of evidence for wineries, a brief overview of the basic elements of a winery (the treading platform, the collection basin, and the mechanical press), and their function, are provided here.

The winery at Izbat Mohamed Farid, 1 km north of Abu Mina, had two phases of use (Grossmann 2002: 31). In the first hase, at the eginning or first half of the 5th century AD it was a simple pressing unit consisting of a treading platform and vat, with walls built of baked brick and covered with waterproof plaster. When it was no longer in use, another winery was built (later part of 6th century AD) partly on top of the old one. This new structure comprised a treading platform, a vat, and two mechanical presses. ring oth hases of se the winer stood ne t to a m ltistory country house.

aking the winer at area as an e am le, it an e seen that a basic press unit was a compound consisting of a raised treading platform and a collection basin, built within an enclosed and most probably roofed, space (El-Fakharani 1983 182-184 he treading floor was se arated from the vat by a low wall. It was covered with waterproof plaster, preserved, for instance, in the Burg el-Arab winery (El-Ashmawi 1998: 62), and was often quite large, as in the case of the Karm el-Baraasi winery (Abd el-Aziz Negm 1998: 66). The room in which the press unit was located was sometimes entered from a courtyard, or had a large window gi ing eas a ess to the treading floor Grapes were thrown through this window, or simply carried through the door, to the treading platform, where they were crushed.

hree wineries near ina were e lored in 1986 the Supreme Council of Antiquities (Abd el-Aziz Negm 1998: 65-73). The winery at Karm Gadoura consisted of

5. For this information I am indebted to Prof. Dr. Marianne Bergmann and Prof. Dr. Michael Heinzelmann of the Swiss-German mission e a ating at hedia

129

LAKE MAREOTIS: RECONSTRUCTING THE PAST F ig. 2: A schematic draw ing of a T ype 4 w inery w ith names of its el ements used in G reek papyri from Egypt ( D . D z ierz bicka) .

W ineries in Papyri r haeologi al finds are not the onl a aila le so r es on wineries (for more details on wine production in papyri see Dzierzbicka 2005: 24-89). Greek papyri of the Hellenisti and oman eriod allow for the identifi ation of words used to call the elements of a winery. The reconstruction drawing of Type 4 winery (Fig. 2) shows the basic parts of a winery and gives the terms by which they were called in papyrus documents. Matching the architectural features with their Greek names in papyri makes it possible for the resear her to se h ndreds of te ts that ro ide in al a le information on the function of the winery as a whole and on its elements, permitting a better understanding of the way these structures were used and provoking more questions to ask oneself in the field

Most wineries were also equipped with a mechanical press either a mo a le ag ress, or a more so histi ated, fi ed screw press. Their preserved remains were found in Abu Mina (Grossmann, et al. 1997: 87-90) and in Burg el-Arab (El-Ashmawi 1998: 60-64), where one can still clearly see a raised, circular base of the screw press. These devices were used to squeeze the remaining liquid out of the trodden gra e l either dire tl on the treading floor, or in an area separated from it, as in Burg el-Arab. In the latter case, a channel connected the place of the mechanical press with the main collection basin, or with a smaller, se arate at he m st flowed into a lastered at or ats d g into the gro nd t remained there thro gh the first turbulent phase of fermentation, during which the basins were covered with wooden planks for protection against contaminants.

One of these questions, especially important in the case of the Mareotic wineries, concerns their surroundings. Papyrus documents containing references to buildings and other immovable property found in vineyards, provide interesting information on other installations forming part of a wine-making om le a le 1 he ta le elow is a om arati e hart of the fa ilities mentioned in e tant te ts, showing how fre entl the a ear and in what combinations. The documents mention elements already discussed: the treading platform (l ê nos) and collection basin (pithos), in a later period referred to by a common term – l ê nopithos, a mechanical press (stemphyl ourgikon organon or mê chanê ), as well as a sunning yard (hê l iastê rion), a storeroom (thesauros) and a cellar (oinothê kê ). In fact, spaces that may have been storerooms were uncovered adjacent to wineries, for instance in Abu Mina (Grossmann, et al. 1997: 89). Large wineries where mass production took place even had a potter’ s workshop (keramikon ergastê rion) on the premises. There was indeed a large pottery kiln in the vicinity of the winery near Burg elArab, suggesting this may have been such a more developed production estate (El-Ashmawi 1998: 62).

enerall s eaking, wineries e hi it a rather niform str ture, dictated by practical needs, that hardly changed regardless of whether the winery was located on the Mareotis lakeshore, in the ahari a asis n lished wine ress photo in Hawass 2000: 164), in Nubia (Gardberg 1970: 41-44), or even on the Crimean Peninsula (Michalowski 1958 49 areoti wineries ro ide an e ellent han e to study the characteristic traits of these structures. What is striking about the typology elaborated by Rodziewicz (1998: 31-36) is the diversity of wine making installations whi h led to the disting ishing of as man as si sim le and two composite types in just over a dozen wineries. However, the distinguished types are in fact very useful as they have strong similarities with wineries in other parts of Egypt and beyond (e.g. Michalowski 1958: 49). As it was noted by Rodziewicz (1998), the wineries consist of two or three essential elements: the treading platform, the collection basin, and, in types 3-8, a mechanical press. The convenient uniformity of this category of structures makes them fairl eas to re ognise in an ar haeologi al e a ation. 130

P . T ebt. III.1 814, Sale, after 227 BC, Tebtynis

X

P . L ond. II 401, Petition, 115–1 10 BC, Thebaid

X

P . de m. G ieben 2, Sale, 107–3 0 BC, Sebennytos

X

P SI VIII 918, Sale, AD 38/39, Tebtynis

X

X

X

P . R oss. G eorg. II 28, Lease, after AD 163/164, Arsinoites

X

P . F l or. III 385, 2nd–3 rd century AD, Hermopolites

P . O x y. LI 3638, ession, 220, inar ,

rh n hites

X

X

X

X

X

X

X

SP P XX 58 (7) Col. II, 3rd century AD, Hermopolis

X

X

X

SB XX 14291, Lease, 3rd ent r

X

X

X

X

O ikopedon

P yrgos

K el l a X

X X

X X

X

X X

X

ermo olis

SP P XX 218, Lease, early 7th century AD, Hermopolites

X

X

P . V ind. Sal . 12, Lease, AD 334/335, Hermopolis

P . V at. A phrod. 25 fr. A, Division of property, 6th century AD, Aphrodito

Epoikion

X

X

P . V ind. T and. 28, ease, 576/577

Epaul is

X

X

P . C airo Masp. I 67097 (r), Contract, AD 571/572 (? ), Aphrodites Kome (Antaiopolites)

X

X

P . O x y. XXXIV 2723, , Sale, 3rd ent r

rh n hos

X

X

X

,

X

X

X

P . F l or. I 50, Division of property, AD 269, Hermopolis rh n hos

X

X X

rh n hos

T hesauros

X

P . O x y. XLIX 3491, Marriage contract, AD 157/158, rh n hos

P SI XIII 1328, etition, 201,

O inothê kê

H ê l iastê rion

K eramikon ergastê rion

P ithos

Stem. organon

Mê chanê

L ê nô n; l ê nos

DOCUMENT

L ê nopithos

D. DZ IERZ BICKA: WINERIES OF THE MAREOTIC REGION

X

X X

X

X

X X

X

T abl e 1: I nstal l ations forming part of a w ine- making compl ex mentioned in G reek papyri containing references to immov abl e property found in v ineyards or to the appurtenances of a w inery. 131

LAKE MAREOTIS: RECONSTRUCTING THE PAST P . dem. G ieben 2 = Vandorpe, K., & Clarysse, W., 1998, A Greek Winery for Sale in a Fayum Demotic Papyrus. In A.M.F.W. Verhoogt & S.P. Vleeming (eds.), T w o F aces of G raeco- R oman Egypt. G reece and D emotic and G reek- D emotic T ex ts and Studies P resented to P . W . P estman: 127-139. Leiden, Boston, Köl n. P . F l or. I 50 = Vitelli, G. (ed.), 1906, P apiri greco- egiz ii, P apiri F iorentini I . D ocumenti pubbl ici e priv ati del l ’ età r omana e biz antina. Milan. P . F l or. III 385 = Vitelli, G. (ed.), 1915, P apiri grecoegiz ii, P apiri F iorentini I I I . D ocumenti e testi l etterari del l ‘ età r omana e biz antina. Milan. P . L ond. II 401 = Kenyon, F.G. (ed.), 1898, G reek P apyri in the B ritish Museum II, 13-14, no. 401. London. P . O x y. XXXIV 2723 = Ingrams, L., Kingston, P., Parsons, P.J., & Rea, J.R. (eds.), 1968, T he O x yrhynchus P apyri XXXIV, no. 2723. London. P . O x y. XLIX 3491 = Bül ow-Jacobsen, A., & Whitehorne, J.E.G. (eds.), 1982, T he O x yrhynchus P apyri XLIX, no. 3491. London. P . O x y. LI 3638 = Rea, J.R. (ed.), 1984, T he O x yrhynchus P apyri LI, no. 3638. London. P . R oss. G eorg. II 28 = Krüge r, O. (ed.), 1929, P apyri russischer und georgischer Samml ungen I I . P tol emäi sche und frühr öm ische T ex te, no 28 iflis P SI VIII 918 = Vitelli, G., & Norsa, M. (eds.), 1927, P apiri greci e l atini VIII, no. 918. Florence. P SI XIII 1328 = Vitelli, G., & Norsa, M. (eds), 1949-1953, P apiri greci e l atini XIII, no. 1328. Florence. P . T ebt. III.1 814 = Hunt, A.S., & Smyly, J.G. (eds.), 1933, T he T ebtunis P apyri III.1, no. 814. London. P . V at. A phrod. 25 = Pintaudi, R. (ed.), 1980, I P apiri V aticani di A phrodito, no. 25. Rome. P . V ind. Sal . 12 = Salomons, R.P. (ed.), 1976, Einige W iener P apyri, no. 12. Amsterdam. P . V ind. T and. 28 = Sijpesteijn, P.J., & Worp, K.A. (eds.), 1976, F ünf unddreissig W iener P apyri, no. 28. Z utphen. SB XX 14291 = Bilabel, F., Kiessling, E., & Rupprecht, H.-A. (eds.), 1997, Sammel buch griechischer U rkunden aus A egypten XX, no. 14291. Wiesbaden. SP P XX 58 = Wessely, C. (ed.), 1921, Studien z ur P al aeographie und P apyruskunde X X . Catal ogus P apyrorum R aineri. Series G raeca. P ars I . T ex tus G raeci papyrorum, qui in l ibro „ P apyrus Erz herz og R ainer– F ühr er durch die A ustel l ung W ien 18 94 “ descripti sunt, no. 58. Leipzig.

The winery was not just a press unit, but a building or a set of buildings, located either in the urban area or in the vicinity of a vineyard. According to some documents the om le was also e i ed with farm ildings epoikion, pyrgos, epaul is), storehouses (kel l a) and building plots (oikopedon). Indeed, the treading platform and vat at Abu Q ir Bay was part of a sizeable square structure built of limestone blocks. The layout of rooms in the building is only partly preserved (Breccia 1926: 47-49). It has not been determined where the entrance to the winery was located and whether there were passages leading to the ada ent rooms he wine-making om le in this str t re may have not been limited to only one unit, consisting of a treading platform and vat, especially that the other rooms also had an industrial function, perhaps related to wine and oil production. As we learn from papyri (Dzierzbicka 2005: 24-89), after the turbulent phase of fermentation was over, the wine was poured into vessels that had been obtained from a pottery worksho lo ated within the om le or a ired from outside the winery. Wine that was paid for in advance was distributed straight from the vat and carried off from the winer in essels ro ided the er he filled ars that remained in the winery were placed in storage rooms or in the sunning yard (hê l iastê rion) for maturing. Both of these facilities were usually spaces or buildings within the wine-making om le and the were e i ed with doors and locks. Maturing wine was occasionally controlled to check if it had not gone sour or spoiled. After the process of fermentation had ended, the vessels were sealed. There are still many uncertainties about the broader conte t of wineries he immediate s rro ndings of a winemaking structure can provide information on the building that housed the installation and determine whether it was located in a rural setting or in a larger industrial or even residential area. Further archaeological investigations, as well as the study of literary and documentary sources, may provide insight into industrial buildings and structures found in the vicinity of a winery and improve our understanding of the character of wine production in the Mareotic region. Ack now ledgments I am grateful to the Polish Centre of Mediterranean Archaeology of the University of Warsaw for assistance granted to me during my stay in Egypt during the conference.

A ncient Sources Athenaeus, T he L earned B anqueters, books I-III.106e, S.D. Olson (transl. & ed.), 2006. Cambridge. Horace, O des I , Carpe D iem, West, D. (transl. & ed.), 1995 ford Strabo, G eography, H.L. Jones (ed.), 1996. London, New Y ork. Virgil, Ecl ogues. G eorgics. A eneid 1- 6 , Vol. 1, H.R. Fairclough (ed.), 1974. Cambridge.

Bibliography P apyri P . Cairo Masp. I 67097 (r) = Maspero, J. (ed.), 1911, P apyrus grecs d’ é poque byz antine. Catal ogue gé né ral des antiquité s é gyptiennes du Musé e du Caire I, no. 67097. Cairo.

132

D. DZ IERZ BICKA: WINERIES OF THE MAREOTIC REGION Gardberg, C.J., 1970, Scandinav ian J oint Ex pedition to Sudanese N ubia, V ol 7: L ate N ubian Sites: Churches and Settl ements. Astrom. Grossmann, P., Kosciuk, J., Severin, G., & Severin, - , 1984, ina lfter orl figer eri ht Kampagnen 1982–1 983. Mitteil ungen des D eutchen A rchäol ogischen I nstituts A bteil ung K airo 40: 148. rossmann, , rnold, , o i k, , 1997, avations at Abu Mina 1995. B ul l etin de l a Socié té d’ A rché ol ogie Copte 36: 87-90. rossmann, , 1999, e ort on the e a ations at Mina in spring 1998. B ul l etin de l a Socié té d’ A rché ol ogie Copte 38: 75-84. rossmann, , 2002, e ort on the e a ations at Mina in spring 2001. B ul l etin de l a Socié té d’ A rché ol ogie Copte 41: 25-31. Hawass, Z ., 2000, T he V al l ey of the G ol den Mummies. Cairo. Michalowski, K., 1958, Mirmeki I . Warszawa. Mül ler-Wiener, W., & Grossmann, P., 1967, Abu Mina. 6. orl figer eri ht A rchäol ogischer A nz eiger 82.4: 468-473. od iewi , , 1998, lassifi ation of wineries from Mareotis. In J.-Y . Empereur (ed.), Commerce et artisanat dans l ’ A l ex andrie hel l é nistique et romaine. A ctes du col l oque d’ A thè nes, organisé par l e CN R S, l e L aboratoire de cé ramol ogie de L yon et l ’ Ecol e franç aise d’ A thè nes, 1 1- 12 dé cembre 19 8: 27-36. Athens, Paris.

Secondary Sources d el- i egm, , 1998, e ent e a ations aro nd Abou Mina. In J.-Y . Empereur (ed.), Commerce et artisanat dans l ’ A l ex andrie hel l é nistique et romaine. A ctes du col l oque d’ A thè nes, organisé par l e CN R S, l e L aboratoire de cé ramol ogie de L yon et l ’ Ecol e franç aise d’ A thè nes, 1 1- 12 dé cembre 1 98: 65-73. Athens, Paris. El-Ashmawi, F., 1998, Pottery kiln and wine-factory at Burg el-Arab. In J.-Y . Empereur (ed.), Commerce et artisanat dans l ’ A l ex andrie hel l é nistique et romaine. A ctes du col l oque d’ A thè nes, organisé par l e CN R S, l e L aboratoire de cé ramol ogie de L yon et l ’ Ecol e franç aise d’ A thè nes, 1 1- 12 dé cembre 1 98: 55-64. Athens, Paris. Blue, L., & Khalil, E. (eds.), forthcoming, A Mul tidiscipl inary A pproach to A l ex andria’ s Economic P ast: the Mareotis Case Study. ford Breccia, E., 1926, L e rov ine e i monumenti di Canopo, a a i i i s. Bergamo. Dzierzbicka, D., 2005, Wineries in Graeco-Roman Egypt. J ournal of J uristic P apyrol ogy 35: 9-89. Empereur, J.-Y ., 1993, La production viticole dans l’ Egypte ptolémaï que et romaine. In M.-C. Maouretti & J.P. Brun (eds.), L a production du v in et de l ’ huil e en Mé diterrané e: 39-47. Athens. l- akharani, , 1983, e ent a ations at area, Egypt. A egyptiaca T rev erensia 2: 175-186. Forbes, R.J., 1955, Studies in A ncient T echnol ogy, Vol 3. Leiden.

133

LAKE MAREOTIS: RECONSTRUCTING THE PAST

134

E. KHALIL: MAREOTIS’ WATERFRONT

Waterfront Installations and Maritime Activities in the Mareotic Region Emad Khalil

Introduction Lake Mareotis represents one of the most distinctive geomorphic features in the north-west coastal region of Egypt. In antiquity, it was fed by means of a number of canals, which bifurcated off the Nile’ s defunct Canopic Branch, and flowed into the so thern and eastern sides of the lake (Fig. 1). Some of these canals were navigable, which enabled merchandise to be transported to and from the hinterland. By the Greco-Roman period the lake was also connected to the sea through a navigable canal that debouched at Alexandria (Strabo 17.1.7). Its connection to both the Nile and the sea resulted in Mareotis becoming a vital conduit of communication in Egypt’ s internal transport system. Moreover, it supported around its shores various agricultural activities and embraced major production centres for different ind stries, whi h ontri ted signifi antl to the economy of Alexandria and to Egypt as a whole. Accordingly, this paper will look at the role that Lake Mareotis played in the ‘ maritime’ 1 transport system of GrecoRoman Egypt. It will also examine the types and nature of the maritime and waterfront installations that were recorded along the shores of the lake and the possible spatial and functional relationship between the different sites.

(Strabo 17.1.14; Pliny 5.11.63). It comprised a main rectangular body of water which merged to the east and south with the Nile Delta Plain, and a narrow arm that extended westwards parallel to the northern coast. However, during the past two millennia Lake Mareotis has undergone dramati hanges whi h signifi antl affe ted its si e and nature. Nonetheless, the western extremities of the lake refle t the original e tant remains, and form an arm, known as the Mareotic Arm, that extends some 40 km west of Alexandria and is 2-3 km wide and is separated from the lake’ s main body by causeways and shallows. It also contains an island, Mareotis Island, which is 3.7 km long and about 680 m at its widest point (see both Blue and Hopkinson this volume).

T he L ak e M areotis Research Project Much evidence indicates that Lake Mareotis extended in antiquity for about 50 km south and west of Alexandria

Since 2004 the Centre for Maritime Archaeology (CMA) of the University of Southampton, in collaboration with the Department of Underwater Antiquities (DUA) of the

Archaeological investigation of the western Mareotic Arm has been ongoing for several decades; however, previous resear h has een largel limited to s e ifi areas and s eifi iss es s h as work on the antine ort of area/ Philoxenité (Petruso & Gabel 1982; El-Fakharani 1983; od iewi 2003 , and work on am horae and wine roduction (Empereur & Picon 1986, 1998) and on the wineries of the areoti region od iewi 1998

F ig. 1: T he approx imate ancient and present l imits of L ake Mareotis ( E. K hal il ) .

1. The term ‘ maritime’ is used in this sense to denote all aspects of waterborne activity and communication from the sea, across the lake, along the canals and on the Nile River.

135

LAKE MAREOTIS: RECONSTRUCTING THE PAST F ig. 2: T he L ake Mareotis R esearch P roj ect surv ey area al ong the shores of the w estern Mareotic A rm ( L ake Mareotis R esearch P roj ect) .

W aterfront S ites in the M areotic Region The waterfront sites that were recorded along the shores of the western Mareotic Arm and on Mareotis Island can be lassified into fo r ategories

Egyptian Supreme Council of Antiquities (SCA), has conducted a comprehensive archaeological survey along the shores of western arm of Lake Mareotis (see Blue this volume). The survey revealed a wealth of archaeological sites including numerous settlements of a maritime and industrial nat re, whi h refle t the e onomi a ti ities that took place in the Mareotic region (Fig. 2). Most of the archaeological sites recorded along the shores of the Mareotic rm are lo ated etween area/ hilo enit and a osiris Magna, with a concentration on Mareotis Island. The sites identified in l de maritime str t res, s h as har o rs, jetties and quays, in addition to what appear to be waterfront warehouses and storage facilities. Sites pertaining to industrial activities are also evident and include amphora kilns, ceramic slag and kiln wasters, as well as a number of structures relating to water management, such as cisterns, wells and water wheels (sakkia). This correlates with much archaeological and textual evidence for viniculture and wine rod tion in the region od iewi 1998

1 - H arbours in the form of har o r om le es of significant design and constitute substantial structures. However, the onl two sites that fit this des ri tion are area/ hiloxenité and Taposiris Magna. The complexity and magnitude of their structures are unrepresented elsewhere in the entire Mareotic region. The two harbours date, however, to quite different periods. Taposiris Magna is essentially Hellenistic in date (see Boussac & El Amouri this volume), while area/ hilo enit mostl dates to the ate oman eriod see aggag, i hot, od iewi and a ra ma ska this ol me owe er, the two har o rs are associated with relatively large towns, and much historical and archaeological evidence indicates that these two towns where probably among the largest and most active along the shores of Lake Mareotis in antiquity.

The dating of the sites discovered relies primarily on the ceramic assemblages collected during the survey. Accordingly it was realised that the majority of sites date from the Hellenistic period until the 7th century.

T aposiris M agna has extensive archaeological remains that date from the Hellenistic to the Late Roman period, including evidence for thriving maritime and commercial 136

E. KHALIL: MAREOTIS’ WATERFRONT activities (El-Fakharani 1974; Empereur 1998: 225-7; od iewi 1998a o ssa l mo ri this ol me The town which is located on the northern shore of the western sub-basin of the lake has one of the best preserved harbours in Lake Mareotis (see Boussac & El Amouri this volume). Commercial activities in Taposiris Magna were mainly focused on handling products transported across the lake as well as receiving goods arriving from the west through overland routes, and shipping them to Alexandria (Empereur 1998: 225). Under the Romans, Taposiris Magna was a customs station where duties were levied on products coming from the Mareotic region and from Cyrenaica heading east towards Alexandria or to the Nile Delta (Empereur 1998: 225-7; Vör ös 2001: 15-6). Alternatively, river vessels could have travelled on the lake through the harbour of Taposiris Magna to the west as far as the lake extended.

ings to allow water to fl sh awa the silt and sediments that might accumulate in the harbour basin. This arrangement was supplemented by the construction of a 1,700 m long wall that extends southwards from the artifi ial ridge to the so thern shore of the lake t was also supplemented by the construction of a solid limestone wall that extended from the northern shores of the lake to the seashore. This wall ensured that even caravans travelling overland had to go through the town of Taposiris Magna (see Fig. 1 in Boussac & El Amouri this volume) e osson 1935 111 od iewi 1990 72-4 Another prominent structure in Taposiris Magna is a 17 m high tower that stands on the coastal ridge to the north of the harbour over looking the Mediterranean coast to the north and the lake to the south (Kadous 2001: 457-60). The function of the tower and its relation with the town and harbour of Taposiris Magna is disputed (El-Fakharani 1974; Vör ös 2001: 37). Nonetheless, it is generally accepted that the tower, which represents a 1:4 or 1:5 replica of the Pharos lighthouse of Alexandria, was in fact a funerary monument for a Hellenistic necropolis that occupied the area around and below the tower (Empereur 1998: 225). The utilisation of such a structure as a landmark by navigators on both the sea and the lake is a possibility that cannot be overlooked (Fig. 3)

The harbour of Taposiris Magna was constructed to control the movement of vessels travelling through it. This was achieved by digging a channel c. 1,700 m long and 50 m wide parallel to the northern shore of the lake. The spoil res lting from the digging was iled to form an artificial ridge which delimits the channel from the south. At the western end of the channel stood a limestone structure that took the form of a double-opening gate or bridge, through which all boats wishing to go through Taposiris agna had to ass m ere r 1998 225-7 od iewi 1998a: 102, n. 32; Vör ös 2001: 15-6) (see Figs. 5-7 in Boussac & El Amouri, this volume). The total width of the gate is about 8.3 m, however, it is divided by a 1.2 m thick wall into two openings; one is 4.1 m wide and the other one is 3 m wide, thus indicative of the maximum possible width of the vessels that passed through. The eastern entrance of the channel is partially obstructed by a quay which is c. 230 m long extending from north to south perpendicular to the shoreline. The distance between the southern end of the quay and the eastern end of the artifi ial ridge, 100 m, forms the eastern entran e of this semi-closed harbour basin of Taposiris Magna. The eastern quay of the harbour includes at least two de-silting openF ig. 3:

As a result of the recent excavation of the area, it is now believed that the digging of the channel as well as the construction of this harbour system, took place the during the Early Roman period rather than during the Hellenistic period, as was previously believed (El-Fakharani 1974; Boussac & El-Amouri this volume). Although the northern shoreline at Taposiris Magna was occupied during the Hellenistic period, as evidenced by houses and shops from the 2nd and 1st century BC, it seems that it was abandoned by the end of the Hellenistic period as a result of a rise in the lake le el he flooded area was then e a ated in the Roman period to create the closed harbour system (see Boussac & El Amouri this volume). M area/ Philoxe nité is located about 15 km east of Taposiris Magna, on the southern shore of the lake. El-Falaki 1966 96 identified this settlement and its asso iated harbour as the town of Marea, the capital of the Mareotic region. According to Herodotus (2.31), Marea was a post of Egyptian soldiers guarding the Libyan border during the time of King Psammetichus of the 26th Dynasty od iewi 2003 27 n the tolemai and oman eriods, Marea functioned as a major trade centre, second onl to le andria n the antine eriod, in addition to its involvement in commercial activities and internal trade, area flo rished as a stopover for pilgrims heading to the hol antine shrine of t inas ina , 20 km so th of the lake a el etr so 1980 har k 2002 od iewi 2003 ntil re entl , most ar haeological research carried out in the area has revealed no evidence earlier than the 5th century AD. However, recent archaeological investigation at Marea has revealed material

T aposiris Magna tow er ( photo: E. K hal il ) .

137

LAKE MAREOTIS: RECONSTRUCTING THE PAST

F ig. 4:

T he middl e quay of the harbour of Marea ( photo: E. K hal il ) . Excavations on the peninsula at the easternmost part of Marea, resulted in the discovery of a large Early Roman building which consists of a courtyard surrounded by numero s rooms of relati el similar si e he ilding is connected to a quay to the north through stairways cut into the rock. Therefore, it has been suggested that the building could have been used for storage and trade. The remains of 1st century BC to 1st century AD workshops for metalwork were also discovered on the peninsula (see Pichot this volume). The recent archaeological discoveries at Marea, particularly of pottery and coins, would indicate that the area was thriving before the 5th century and possibly as early as the Hellenistic period

from the Hellenistic and Roman periods (see both Pichot and a ra ma ska this ol me mongst the most signifi ant ar haeologi al remains in Marea are four quays that extend into the lake and divide the 1.5 km long shoreline into eastern, central and western harbour basins (Figs. 4 & 5). The dimensions of the quays from west to east are: 41 m x 6.5 m, 111 m x 5 m, 125 m 7 m, and 35 m 4 m ma ska a ra 2008 1115). Judging from the construction technique of the quays at Marea, in which large regular limestone blocks (1 m x 0.5 m x 0.03 m) were used for their construction, it seems reasonable to suggest that the four quays were constructed earlier than the antine it owe er, the were ro ably subject to several building phases in subsequent periods sin e e iden e of antine h dra li mortar opus signinum) can still be seen between many of the building blocks of the quays.

2 - The second category of waterfront sites that was recorded along the shores of the western Mareotic Arm consists of different types of anchorage facilities such as quays and jetties, which form the majority of maritime F ig. 5: T he quay of Marea harbour w as made of l arge regul ar l imestone bl ocks w ithout ev idence of the use of mortar ( photo: E. K hal il ) .

138

E. KHALIL: MAREOTIS’ WATERFRONT

F ig. 6:

T he box - shaped harbour on the southern shore of L ake Mareotis ( photo: L ake Mareotis R esearch P roj ect) .

installations in the region. More than ten different anchorage facilities were recorded on the northern and southern shores of the lake and on the northern shore of Mareotis Island. Possibly the most substantial of them is a Kibotos or box-shaped harbour (Site 09) which is located at the end of a promontory on the southern shore of the lake, about 2 km to the west of the Sidi Kerir-Borg El Arab road. The harbour, which is constructed of large limestone blocks, consists of two parallel moles enclosing an area some 60 m long (N to S) and 36 m (E to W) wide (Fig. 6). The eastern mole is 60 m long and at its northern extremity it returns to the west for a distance of 12 m, while the western mole, which is less well preserved, extends for 40 m and returns to the east at the northern end for a distance of some 6 m. A gap of 18 m between the two ends of the two moles equates to the entrance of the harbour on the north side. The moles are constructed of up to three courses of single and double breadth limestone blocks. Although the dating for this harbour is still uncertain, judging from its construction te hni e and from the large si e of lo ks sed 1 10 x 0.7 x 0.5 m), it seems that the harbour is pre-Roman in date. However, the existence of lumps of coarse red mortar opus signinum with lime inclusions between some of the blocks indicate that it remained in use at least until the antine eriod ne ar ed mooring ring was noted on the upper course of one of the blocks, which would have helped facilitate the mooring of vessels to the outside of the harbour (El-Fakharani 1984; Blue & Ramses 2006).

They are Sites 204 (Gamal) and 208 (Q useir). The jetties at each of these sites, which are located approximately 2 km apart, are about 50 m long and 8 m wide. Evidence of red mortar opus signinum is noted between some of the blocks. Also the jetty at Site 208 (Q useir) had mooring stones extending from the upper course of blocks at the eastern side (the lee side) of the quay. Other jetty-like features, although not as substantial, are located along the southern shore of the lake and the northern shore of the Mareotis Island. However, it was realised during the survey that the anchorage facilities along the northern shore of the lake are mostly associated with civic and residential sites, while those along the southern shore of the lake, particularly on Mareotis Island, are associated with sites of a commercial nature. The dating of these sites based on ceramic collections is quite problematic since the jetties are continuously washed by water in the winter, which, in many cases, does not leave any ceramics to be dated. However, judging from the ceramics dated from adjacent sites, it was realised that most of the sites could have been used for quite a long period of time, probably from the ellenisti to the ate oman or antine eriod 3 - he third t e of maritime installation identified an e described as seaw alls or more accurately lak e w alls. Unlike the jetties, which are perpendicular to the shore, lake walls arallel the shore and the were intended to define the shores and protect them from the effects of silting and sedimentation ig 8 t least fi e lake walls were dis o ered in the survey region. These kinds of structures are mainly found along the southern shore of the lake and the northern shore of Mareotis Island, those shorelines most subject to silting and the deposition of sediments as a result of the prevailing northwest winds that would carry sediment from the coastal ridges and deposit it into the lake. Besides acting as a form of protection against silting, the lake walls could also have been utilised as docking facilities for merchant vessels. Another possible function for such structures was to retain rainwater for use in agricultural purposes.

With the exception of this square harbour on the southern shore of the lake, all the other anchorage facilities along the shores of the lake take the form of jetties and quays that extend into the water perpendicular to the shore (Fig. 7). The technique used for the construction of most of the quays was building two parallel single or double breadth iers of limestone and filling the distan e etween them with rubble. With the exception of the substantial struct res alread noted at area/ hilo enit and a osiris Magna the two most substantial anchorages are located on the northern shore of the lake opposite Mareotis Island. 139

LAKE MAREOTIS: RECONSTRUCTING THE PAST F ig. 7: O ne of the quays ex tending into the l ake at the northern shore of Mareotis w estern arm ( photo: L ake Mareotis R esearch P roj ect) .

F ig. 8: ( bel ow ) A l ake w al l ex tending paral l el to the southern shore of the l ake. E v idence for red mortar (opus signinum) can stil l be seen betw een the bl ocks ( photo: L ake Mareotis R esearch P roj ect) .

The longest of these lake walls is located on the north shore at the western end of Mareotis Island (Site 21). It is c. 245 m long and 1 m wide and is constructed of a series of limestone blocks laid as stretchers along the lake edge (see Hopkinson this volume). Other examples include a wall located in the middle of the island on the north shore, the extant remains of which extended some 70 m in length and was made of one course of large limestone blocks of 0.60 m x 0.30 m x 0.25 m dimension, that were arranged as headers facing the shoreline (Site 123). Similarly, on the southern shore of the lake further substantial walls of o er 250 m in length were identified at oth ites 109 and 44. 4 - he fo rth and final t e of waterfront str t res does not necessarily have a maritime function. At a number of sites in the survey region the remains of several multi room buildings were recorded very close to the present waterline. Some structures even extend into the water (Fig. 9). Examples of this type of structure are found in Sites 117, 118 and 119 which are located at the north-eastern shore of the island. At Site 117 there are the remains of a multi-roomed building that measures 12 m EW x 17 m NS which is divided on the inside into at least four smaller rooms. Site 118, about 25 m west of Site 117, contains the remains of at least two multi-roomed structures which measure 18 m EW x 15 m NS and 20 m EW x 20 m NS. Each of them contains the remains of numerous walls which belonged to a number of internal rooms of different sha es and si es About 40 m to the west of Site 118, a further Site 119 contains the remains of a rectangular building that extends from the shoreline southwards for about 40 m and measures about 25 m EW. Limited excavation carried out in the middle section of this building revealed the remains of two 140

E. KHALIL: MAREOTIS’ WATERFRONT

F ig. 9: Some of the mul ti- room square buil dings l ocated at Site 1 18 al ong the northern shore of the Mareotis I sl and. T he buil dings coul d hav e been used for storage purposes ( image: L ake Mareotis R esearch P roj ect) . signifi ant str t res he first str t re is a re tang lar enclosure that measures c. 9 m x 4.5 m that contained two rooms. Both rooms have almost the same dimension c. 3 m EW x 2.7 m NS. The second structure constructed on the same alignment as the first and ad a ent to it to the west, is a rectangular building that measures c. 2.5 m NS and at least 5 m EW, in which remains of imported Hellenistic amphorae were discovered.

Marea and Taposiris Magna, only a distance of some 15 km, but out of a total of more than ninety archaeological sites recorded along the shores of the Mareotic Arm west of Alexandria, only four substantial sites were recorded to the west of Taposiris Magna. Thus, judging from the nature and extent of these sites, it is evident that the navigable limits of Lake Mareotis in antiquity extended west of Taposiris Magna for at least 12 km.

Accordingly, it seems reasonable to suggest that this type of waterfront structure were used as storage facilities for different merchandise and products that were traded along the Mareotic Arm.

Likewise, it is noticeable that maritime installations located on the southern shore of the lake from Marea to the eastern end of Mareotis Island, are in fact located on a ridge that extends for about 800 m from the present southern shoreline to the west. It is noteworthy that no sites were recorded on the southern shore of the lake opposite this ridge. This actually raises a question about the nature of the area between the ridge and the southern shore of the lake, and whether or not it was actually land in antiqit , that has s se entl een artifi iall e a ated or subject to inundation due to the changes in ground water level (see Flaux forthcoming). Similarly, all the archaeological sites of a maritime nature that were recorded on Mareotis Island, where located along its northern shore, with essentially no evidence for sites either on the southern shore of the island or on the northern shore of the lake opposite. This also raises a question about the nature of this island and whether or not it was actually an island in antiquity.

By examining the remains of these structures, particularly on Maerotis Island, it becomes evident that they have undergone construction phases over successive periods. Moreover, it seems that the sections of the structures closest to the waterline were subject to accumulated sediments, and hence had to be rebuilt. In other words the different hases of ilding and modifi ation of str t res o ld e the result of adapting to changes in the waterline. Relations and Significance By looking at the distribution of archaeological sites in general and maritime sites in particular along the shores of the Mareotic Arm, it becomes evident that not only is there an apparent concentration of sites in the area between 141

LAKE MAREOTIS: RECONSTRUCTING THE PAST N avi gation in L ak e M areotis In the 1st century BC, when speaking about the water supply for Lake Mareotis, Strabo (17.1.7) mentions that it is filled man anals from the ile, oth from a o e and on the sides, and through these canals the imports are much larger than those from the sea, so that the harbour on the lake was in fact richer than that on the sea” . On another occasion (Strabo 17.1.22) speaks of “ …s everal canals, which empty into Lake Mareotis” . As a result, it has been assumed that there was intense maritime traffi assing thro gh the lake arr ing ario s rod ts and cargoes to Alexandria. Merchandise which would have been transported to Alexandria for local consumption and for transshipment to other Mediterranean harbours would have included Egyptian products such as papyrus, textile and grain (Rickman 1971: 300-6, 1980: 231-5; Lewis 1983: 165-7), as well as quarried stones from the Eastern Desert (Peacock 1992: 5-7; 2002: 426-7). It would have also included products imported via the Red Sea from India, Arabia and East Africa such as spices, tortoiseshell, frankincense, ivory, cotton, silk and gems (Strabo 2.5.12; Casson 1991: 200-212; Peacock 2002: 432-3). At the same time, Alexandria was receiving from the Mediterranean, for local consumption and for transshipment to the south, various products such as wine, oil, fine otter , glass, tim er, o er and tin et, the role that Lake Mareotis played in this internal transport system is some what unclear.

The different nature of maritime installations located along the northern shore of the lake and those along the southern shore and on Mareotis Island, is also noteworthy. As mentioned earlier, maritime installations such as quays and jetties located the northern shore of the lake are mostly associated with large tell sites of a civic and residential nature. These tells were up to 60,000 m2 in area, and were densely occupied. They are mainly covered with building stones and the foundations of buildings, as well as the remains of several wells, cisterns and red brick basins lined with opus signinum, which could have been used in baths, houses or other urban structures. However, the situation on the southern shore is quite different since the southern shore is where most industrial and commercial sites were recorded, and hence maritime installations were mostly associates with those sites. In antiquity, almost all the amphora and wine production sites in the region, as well as sakkia installations (Emere r i on 1998 od iewi 1998 see also the indi id al ontri tions of l e, o kinson and ier bicka this volume), were located on the southern shore of the lake, the focus of agricultural and industrial activities. The reason that agricultural and industrial activities where concentrated on the southern shores of the Mareotic Arm is mainly due to the difference in the topography between the northern and southern shores of the lake. The western arm of Lake Mareotis is delimited from the north and the south by two limestone carbonate ridges, of average elevation 25-35 m and average width 300 m (Said 1990: 499; Warne & Stanley 1993; El-Raey, et al. 1995: 191; Frihy, et al. 1996: 282). The northern ridge is known as the Abusir Ridge, and to the south a longitudinal depression 3-4 km wide known as Al-Alamein-Maryut Depression, extends roughly E-W, partly occupied by the western arm of Lake Mareotis. This depression is delimited to the south by another coastal ridge, known as Gebel Maryut Ridge, which is located 5-9 km south of the Abusir Ridge.

Although it is well known that Lake Mareotis was fed by means of a number of canals, which branched off the ano i ran h, and flowed into the so thern and eastern reaches of the lake, there is a considerable degree of uncertainty about the exact number, location and the routes of these canals. However, the most important of these canals was hedia anal see ergmann, ein elmann Martin this volume) It bifurcated off the Canopic Branch of the Nile at the town of Schedia, originally a Hellenistic foundation that was later known as Chaereu, currently loated in the region of the illages of om l- i a, om El-Nashw and Kom El-Hamam, some 30 km south-east of le andria ergmann ein elmann 2004 hile the Canopic Branch continued north to debouch into the Canopic Bay (Abukir Bay), the Schedia Canal turned northwest towards Alexandria and followed a course close to the present course of the Al-Mahmoudeyah Canal. In a statement by Strabo (17.1.16) in which he describes the town of Schedia, he mentions that it has “ …t he station for paying duty on the goods brought down from above it and brought up from below it; and for this purpose, also, a s hedia float has een laid a ross the ri er, from which the place has its name” . Accordingly, Schedia was the main Nile emporium, customs harbour and checkpoint east of Alexandria, where custom duties were imposed on imported and exported goods (Empereur 1998: 225; Bergmann ein elmann 2004 oreo er, it seems that the Canopic Branch at Schedia was obstructed by some kind of a pontoon that prevented boats from sailing past it until duties were paid on merchandise travelling both ways. Additionally, it was at Schedia where exported commodities

Accordingly, the distance between the Gebel Maryut Ridge and the southern shore of the lake is far greater than the distance between the Abusir Ridge and the northern shore of the lake. In antiquity, the area south of the lake was a fertile lain that flo rished with agri lt ral a ti ities and was known for the quality of its vines, olives and fruits as well as for the lti ation of fla and a r s thenaeus 1.33.d-e; Pliny 13.22.71; Strabo 17.1.14; Empereur & Picon 1998; Horden & Purcell 2002: 353; McGovern 2003: 121-3). Moreover, the abundance of calcareous clay, particularly suitable for amphora production, resulted in a thriving large-scale amphora industry along the southern shore of the lake (Empereur & Picon 1986: 103-9, 1992, 1998 od iewi 1998 l e amses 2007 h s, the focus of wine and amphora production in Hellenistic and Roman times was the southern shores of the Mareotic Arm. Thus, it is reasonable to suggest that those maritime installations located along the southern shore of the lake were very much involved in commercial activities including the transport of Mareotic products to Alexandria and possibly also to the southern limits of the lake. 142

E. KHALIL: MAREOTIS’ WATERFRONT brought from upriver were transferred from large Nile boats to smaller boats that could travel easily through the canals to Alexandria (Procopius 6.1.3; Hassan 1997: 365 n. 13).

sequently, the lake’ s southern and eastern shoreline was unstable and subject to constant change, and was therefore unsuitable for the establishment of substantial harbours and waterfront installations. A recent survey conducted along the ancient southern and eastern limits of the lake (Wilson 2007; see Wilson this volume) revealed that most settlements established in this area during the Ptolemaic and Roman periods, were located on high ground around the lake’ s edge. Also it revealed that many settlements were involved in agriculture and industrial activities mainly in the service of Alexandria. However, as yet there is no e iden e for s stantial maritime str t res or signifi ant waterfront installations.

As it approached Alexandria, the Schedia Canal bifurcated into two branches in the Alexandrian suburb of El eusis l- o ha he first ran h t rned towards the north-east leading to Canopus, east of Alexandria, while the other branch continued south of Alexandria and parallel to the lake’ s northern shore, until it debouched into Lake Mareotis south-east of Alexandria. According to Strabo (17.1.7), boats also sailed from the Nile to the Canopic Branch and through the network of canals that fed the lake from the south and east, then across the lake northwards to Alexandria. This indicates that navigation on Lake Mareotis was intense and operated in many directions. It also raises a point about the practicalities of sailing in Lake Mareotis from south to north against the prevailing north-westerly wind. The predominant winds along the north coast of Egypt are north-westerly and they prevail more than 40% of the time throughout the year and more than 70% of the time during the summer sailing season (El-Z ouka 1979: 125-7; El-Gindy 1999: 17). Accordingly, merchant vessels sailing in Lake Mareotis from south to north would have faced a direct headwind, which meant that the boats had to tack in order to reach Alexandria. Tacking in Lake Mareotis was possible considering the large area of the water body; however, tacking from the southern limits of the lake to Alexandria would have meant that boats would have to travel several times the direct distance across a water body full of shallows and marshlands and against prevailing winds. In the 5th century St. Palladius (7.1) mentioned that he sailed across Lake Mareotis from north to south, from Alexandria to the monastic settlement of Mount Nitria, a distance of about 50 km, in a day and half. Accordingly, sailing in the lake in the opposite direction would have taken much longer, possi l as long as fo r to fi e da s

Conclusion There were two ways for river vessels to travel to and from Alexandria, either across the lake, or along the Schedia Canal. Considering the arguments outlined above, particularly in relation to the prevailing winds, it seems reasona le to s ggest that the main north o nd traffi ro a l went via the Canopic Branch and the Schedia Canal, rather than across the lake. However, sailing south across the lake would have been conducted with considerable ease. Along those stretches of the canal where boats had to maneouver against the wind, they could have been towed along from the shore, a standard procedure for moving river boats in rivers and canals around the world. In that respect, Strabo’ s statement (17.1.7) about the lake harbour south of Alexandria being richer than the seaport of Alexandria, would still be valid. At the time of Strabo, the Schedia Canal debouched into Lake Mareotis, so all the anal traffi had to ass thro gh the lake oreo er, boats coming from the western arm of the lake also arrived at the lake harbour. Therefore, it is possible that the lake harbour was quite busy receiving river vessels from the south as well as from the west. At the same time, it is not unreasonable to suggest that east–w est commercial traffi along the western areoti rm was ro a l more intense and more reg lar than the north so th traffi that passed through the main body of the lake. Recent archaeological investigation in the Mareotic region have revealed that the number, nature and extent of archaeological sites along the shores of the Mareotic Arm, is unparalleled any where else in the Western Deltaic Region (see Blue this volume). Settlements in this region were located far from the silting effects of the Nile sediments, the coastline was more stable, prevailing winds were more favourable for east-west movement, and settlements were in close proximity to Alexandria. Therefore, it is understandable why so many shoreline settlements and associated maritime installations were established along its shores. Thus, the contribution of the western Mareotic Arm to the economy of ancient Alexandria and hence of Egypt as a whole, was ro a l far more signifi ant than an other area along the shores of Lake Mareotis. Thus, the shores of the western arm of Lake Mareotis appear to have been one of the most active areas of economic activity in the Western eltai region d ring the ellenisti , oman and antine periods.

Furthermore, the extended period of travel across the lake would have laid boats venerable to another challenge that prevailed on Lake Mareotis in antiquity. Achilles Tatius (4.12) in the 2nd century and Heliodorus (1:14) in the 3rd century, spoke of piracy and bandits on Lake Mareotis. The marshes and islands of the lake provided excellent hideouts for groups of bandits and their vessels. Also, the large si e of the lake made it ite diffi lt to g ard and control, therefore, it is possible that sailing across the lake with valuable commodities was quite risky. Moreover, settlements located on the southern and eastern shores of Lake Mareotis were far more susceptible to sedimentation from silts deposited via the nearby Canopic ran h of the ile, arti larl d ring flood seasons, as well as sediment which had been carried by the prevailing winds across the lake from the north-west to the southeast. All this would have contributed to the build up of sediments against the southern and eastern shores of the lake, thus preventing settlements in this region from being as actively involved in across lake transportation. Con143

LAKE MAREOTIS: RECONSTRUCTING THE PAST Bibliography Achilles Tatius, T he A dv entures of L eucippe and Cl itophon, S. Gaselee (transl.), 1969. London. Athenaeus, i s his a C.B. Gulick (transl.), 1953. London. ergmann, , ein elmann, , 2004, Schedia ( K om i ah a a a h ia R eport on the documentation and ex cav ation season. 18 March – 18 A pril 203. Paper presented at conference on Alexandria in Antiquity, University of Oxford, December 2004. Blue, L., & Ramses, S., 2006, L ake Mareotis R esearch P roj ect. Report submitted to the Egyptian Supreme o n il for nti ities on the fieldwork and res lts of the e tem er 2006 field season Blue, L., & Ramses, S., 2007, L ake Mareotis R esearch P roj ect. Report submitted to the Egyptian Supreme o n il for nti ities on the fieldwork and res lts of the a l / g st 2007 field seasons Casson, L., 1991, T he A ncient Mariners. Princeton. De Cosson, A., 1935, Mareotis: B eing an A ccount of the is y a i s h h s D esert of Egypt and L ake Mareotis. London El-Ashmawi, F., 1998, Pottery Kiln and Wine Factory at rg l- ra n m ere r ed , Commerce a isa a a s a i h is iq ai s q h s a is a a a i a i y a ais h s 55-64. Athens, Paris. El-Fakharani, F., 1974, The Lighthouse of Abusir in Egypt. H arv ard Studies in Cl assical P hil ol ogy 78: 257-272. El-Fakharani, F., 1983, Recent Excavation at Marea in Egypt. In G. Grimm, H. Heinen & E. Winter (eds.), y ia a sia as is h y a i ische Ä gypten. A kten des international en Symposions 26.30. Se ptember 1 978 i n T rier, 175-186 ain El-Falaki, M.S., 1966, A ncient A l ex andria. A l ex andria: ), D ar N ashr A l - T haquafa ( M.B. Astronome (transl.), 1872. Copenhague. El-Gindy, A., 1999, Meteorological and Hydrodynamic Conditions in the Eastern Harbour of Alexandria and its i init n alim ed , P roceedings of the ksh h a s i h s ai abl e D ev el opment of the Submarine A rchaeol ogical Sites at Q ayetbey Citadel and Eastern H arbour of a ia s y y a i s hysi s h y a y, V. II: 1-82. Alexandria. El-Raey, M., Nasr, S., Frihy, O., Desouki, S., & Dewidar, Kh., 1995, Potential Impact of Accelerated Sea-Level Rise on Alexandria Governorate, Egypt. J ournal of Coastal R esearch 14: 190-204. El-Z ouka, M.K., 1979, i ai as i h s D el ta: A G eographical Study ( 8´ ), Alexandria (in Arabic). m ere r, - , i on, , 1986, a e her he des o rs d m hores n m ere r arlan (eds.), R echerches Sur L es A mphores G recques. A ctes

du col l oque international organisé par l e Centre ai a a h h s i i q i si s a ais h s h s 10- 12 Se ptembre 1984) : 103-126. Athens, Paris. m ere r, - , i on, , 1998 es telier d mhores d a ario t n m ere r ed , Coma isa a a s a i h is iq ai s q h s a is a a a i a i y a ais h s 75-88. Athens, Paris. m ere r, - , i on, , 1992, a e onnaissan e Maréotide. Ex trait des Cahiers de l a Cé ramique y i 3: 145-152. m ere r, - , 1998, A l ex andria R ediscov ered. New ork Flaux, C., forthcoming, Geomorphological survey of the south-west shores of Lake Mareotis. In L. Blue & E. Khalil (eds.), A Mul tidiscipl inary A pproach to A l ex a ia s i as h a is as y Oxford. Frihy, O.E., Dewidar, Kh.M., & El-Raey, M.M., 1996, Evaluation of Coastal Problems at Alexandria, Egypt. O cean & Coastal Management 30.2-3: 281-295. Gabel, C., & Petruso, K., 1980, An Environmental and Cultural Study at Lake Maryut, Lower Egypt. B oston i si y i a i s ki a s No. 25. Boston. etr so, , a el, , 1982, area antine ort in Northern Egypt. s i si y i a i s ki a s No. 62. Boston. assan, , 1997, he nami s of i erine i ili ation: A Geoarchaeological Perspective on the Nile Valley, Egypt. ha y 29.1: 51-74. Heliodorus. A ethiopica, T. Underdowne (transl.), 1895. London. Herodotus. T he H istories, A. de Sélincourt (transl.), 1971. Middlesex. Horden, P., & Purcell, N., 2002, T he Corrupting Sea. Oxford. Kadous, E., 2001, Monuments of A ncient A l ex andria. ( ) Alexandria (in Arabic). har k, , 2002, area 2001 indow anes and other lass inds n awlikowski as ewski (eds.), P ol ish A rchaeol ogy in the Mediterranean XIII: 65-71. Lewis, N., 1983, L ife in Egypt under R oman R ul e. Oxford. McGovern, P.E., 2003, i i h a h h O rigins of V inicul ture. Princeton. Palladius, T he L ausiac H istory. The Monks of Nitria. htt //www fordham ed /halsall/ asis/ alladi slausiac.html (accessed February 2010). Peacock, D., 1992, R ome in the D esert: A Symbol of P ow er. An Inaugural Lecture delivered at the University of Southampton. Peacock, D., 2002, The Roman Period. In I. Shaw (ed.), T he O x ford H istory of A ncient Egypt: 422-445. Oxford. Pliny the Elder, T he N atural H istory, J. Bostock & B.A. Riley (transl.), 1855. London. 144

E. KHALIL: MAREOTIS’ WATERFRONT a a i a i y a ais h s 27-36. Athens, Paris. od iewi , , 2003, hilo enit , ilgrimage arbour of Abu Mina. B ul l etin de l a Socié té A rché ol oiq a i 47: 27-47. Said, R., 1990, Q uaternary. In R. Said (ed.), T he G eol ogy of Egypt: 487-507. Rotterdam. Strabo, G eography, H.L. Jones (transl.), 2001. London. ma ska, , a ra , , 2008, Marea V .1 : B yz antine Marea – Ex cav ations in 20203 and 206 . Krakow. Vör ös , G., 2001, T aposiris Magna: P ort of I sis. Budapest. Warne, A.G., & Stanley, D.J., 1993, Late Q uaternary Evolution of the Northwest Nile Delta and Adjacent Coast in the Alexandria Region, Egypt. J ournal of Coastal R esearch 9.1: 26-64. Wilson, P., 2007, s a i a y htt //www d r a k/ enelo e wilson/ elta/ atamir html (accessed February 2010).

Procopius, T he B uil dings, H.B. Dewing (transl.), 1940. London. Rickman, G., 1971, R oman G ranaries and Store B uil dings. Cambridge. Rickman, G., 1980, T he Corn Suppl y of A ncient R ome. Oxford. od iewi , , 1990, aenia and areotis r haeological Research West of Alexandria. A cta of the F irst I nternational Col l oquium of the Egyptian Society of G reek and R oman Studies 1: 62-80. Cairo. od iewi , , 1998a, rom le andria to the est and and aterwa s n m ere r ed , a isa a a s a i h is iq ai s q h s a is a a a i a i y a ais h s 1 1- 12 dé cembre 198: 93-103. Athens, Paris. od iewi , , 1998 , lassifi ation of ineries from areotis n m ere r ed , Commerce et a isa a a s a i h is iq ai s q h s a is a

145

LAKE MAREOTIS: RECONSTRUCTING THE PAST

146

S. RAMSES & A. OMAR: OUTRAGE AND DESTRUCTION

Lake Mareotis Research Project. Phases of Outrage and Destruction Sameh Ramses & Ahmed Omar

Interest and Aim of the Project The historical importance of the region of Mareotis during the Ptolemaic and Roman periods, either as an important wine production center, in support of other Egyptian industries (Strabo 17.1.7), or as a resort of the elite (Buttler 1989: 8-12), has attracted a great deal of archaeological attention, as this volume illustrates (El-Fakharani 1983; Empereur & Picon 1998; Rodziewicz 1998). Nevertheless, the region of Mareotis is threatened by modern urban development and many historical sites are subjected to destruction, not only due to its unique location in the vicinity of Alexandria, but also its moderate climate which has established the area as one of the most popular resorts in the region (Abd-Elhady 2008: 177-184). Therefore, the aim of the Lake Mareotis Research Project, conducted between the summer of 2004 until the summer of 2008, was to survey, record and map the platforms, harbours, and any other waterfront constructions that could be found along the northern and southern shores of the western part of Lake Mareotis, and to create a very detailed database for each site, as well as to determine its present state of preservation.

Phases of Outrage and Destruction According to the last survey of 2008 more than 70 archaeological sites have already been noted in the region (Fig. 2). Of course, not all of them had previously been registered, but some had been surveyed or documented.1 However, despite efforts by the Archaeological Department of Alexandria University to discover and protect some of these sites, the full extent of the area, and the variety of remains, both on land and in the water, still remains to be determined and much work needs to be done. This is particularly urgent considering the very immediate threats of modern urban development and various industrial and agricultural activities that are threatening the shores of the lake. Consequently, during the course of the survey, three main fa tors were identified that are elie ed to resent a particular danger to the archaeology of the region. F ish F arms As a consequence of being inundated by water largely of agriculture runoff and drainage water, the lake water qualit has deteriorated making it more diffi lt for a ariet of fish to s r i e im ltaneo sl , man fish-farmers egan to appropriate quite large portions from the lake basin and along the shores n the ro ess of reating fish rod ing basins, a great deal of dredging and destruction of some of the archaeological sites, both offshore and under water, has taken place. Accordingly, the original shores topography has been deformed, and the dredging leads to

The survey project covers an area of about 40 km long and up to 3 km wide within the western arm of Lake Mareotis, to the west of Borg El-Arab airport, on the Sidi Kerir road, and extends towards El-Hammam city, it also includes Mareotis Island some 3.7 km long and 680 m wide (Fig. 1; see Blue, Hopkinson and Khalil this volume). F ig . 1: G eneral pl an from A l ex andria to Mareotis ( L ake Mareotis R esearch P roj ect) .

1. It is noteworthy that only about nine sites had already been registered by the Supreme Council of Antiquities (SCA), and only six are considered to be SCA properties. A tl as A l moqea A l athariaya, SCA, 2002: No. 3: El-Beheira Governorate; No. 4: The Rest of the Lower Egypt Governorates. Egypt.

147

LAKE MAREOTIS: RECONSTRUCTING THE PAST

F ig. 2:

Map of sites inv estigated during the L ake Mareotis R esearch P roj ect ( L ake Mareotis R esearch P roj ect) . I ndustrial A ctiv ities Although not as polluted as some parts of Lake Mareotis, the western arm determines a substantial area of the current lake littoral, and is considerably affected by a huge amount of untreated industrial waste and polluted water whi h flows into the lake on a dail asis he ind strial activities around the shores of the lake not only affect the water quality but in some cases have a direct impact on the

mis-interpretations of its original form. These activities particularly characterise the southern shore of the lake (Fig. 3). It is worth mentioning that some farms have appropriated quite large areas, up to nearly 4 hectares, especially in the vicinity of Mareotis Island, which embraces a great deal of archaeological remains, making access to the island in reasingl diffi lt tting awa the roads and destroying the routes that lead to the island. 148

S. RAMSES & A. OMAR: OUTRAGE AND DESTRUCTION

i

a

s

sh a

ai

F ig. 4: L ime storing and col l ecting at Site 4 ( Mareotis R esearch P roj ect) .

a

h s

h

L ake

sh

h ak

ak

a

F ig. 5: L ake w al l affected at Site 4 ( R esearch P roj ect) .

archaeological sites located in the region. Site 44, located on the southwest shore of the lake, provides clear evidence for this negligence. It is directly adjacent to a cement factory, and is used to store the lime for the cement (Fig. 4). The cement mound is huge and it no doubt, amongst other things, covers the western end of a lake wall that extends along the lake shore (Fig. 5), and disregards the existence of additional archaeological features including a circular structure at the far west of the site near the water’ s edge, and four square probable wine-basins, that were previously excavated by the SCA. Similar scenes are witnessed further west, for example at the site of Naga El-Mawalik (Site 109). This is an example of an endangered site that still displays aspects of its ancient harbour and enclosure walls, together with many other hills and mounds of pottery. Its location close to the cement factory, means that it is being used to supply the factory with limestone. Parts of this site are now totally inaccessible, with unrecorded archaeological features being surrounded by barbed wire.

is

s a h L ake Mareotis

On the north shore of the lake, at the western extent of the survey area, Site 214, immediately opposite Naga El-Mawalik, has recently been purchased by the El-Nasr Salt Company. They have acquired two already excavated sites (Sites 214-215) which contain two wineries and a big house of the late Roman period and numerous mounds of pottery (Figs. 6-7). These sites are due to be dredged and leveled to be used as evaporations basins to produce salt. They have only been partially recorded. There are many other places in the area free of archaeological remains that could be used for such purposes. Other sites, particularly those on the northern shore of the survey area, are being used as rubbish tips totally swamped by rubbish from the city and isolated from the main road by means of barbed wire. These sites are distinctive by their bad smell. U rban D ev el opment Urban development is considered to be one of the most dangerous factors that affect the archaeological sites of the 149

LAKE MAREOTIS: RECONSTRUCTING THE PAST

F ig. 6: ( abov e) Site 214: the partial l y ex cav ated w inery ( L ake Mareotis R esearch P roj ect) .

F ig. 7:

( right) Site 214: part of the fragmented pav ed a a i a i i ak Mareotis R esearch P roj ect) . region. Vast portions of land have been acquired, many divided into smaller units to be sold and used for housing. oth the lo al edo in and ig om anies that wield rofit and power are responsible for destroying and leveling vast areas at a remarkably rapid rate. The Lake Mareotis Research Project has had direct experience of these actions over the last four years of survey. This can be summarised in the following four case studies:

do mentation of the site, we ret rned to find that the three hills were totally destroyed by the locals and the Hellenisti tom flattened ig 12 t is worth mentioning that this site had not been previously registered or recorded by the SCA. El - G amal ( Site 204) The situation at the site of El-Gamal was a little different as the site was quite impressive and important, to the extant that it had already been partially recorded by the SCA, and a site guard had been assigned. The destruction of this site was ndertaken in stages first a so er field was leared to the east of the tell; the following year another one was cleared on the prohibited area of the archaeological site towards the limit of the lake shores, destroying part of the tell and site boundaries, and so the destruction continued (Fig. 13).

Site 09 A Kibotos harbour, the only well preserved example either on the sea, or on the lake. Nevertheless, the governorate has dedicated an area of about 8 hectares around the site to build a private university without leaving any access to the archaeological site (Figs. 8-9). Accordingly, over time, the site will be destroyed. Ez - Z eraa El bahry ( Site 201) When visited in September 2004, there was a hill of about 5 m in height, which contained several mounds of pottery with traces of building plots everywhere. When we returned in May 2007, the 5 m high hill had disappeared and there were only private property signs and modern partitioning walls visible (Figs. 10-11).

Even those sites renowned for their archaeological importance are also under threat. To the southeast of Taposiris Magna, the extensive seawall and jetty associated with the site are slowly being encroached by small concrete building plot walls, purchased for urban development (see Boussac & El Amouri this volume).

N agea Ez - Z ohorat ( Site 20) A similar pattern of destruction was noted at Nagea EzZ ohorat (Site 202) as was witnessed at Ez-Z eraa Elbahry (Site 201). The site comprised three hills containing amongst other remains, a cistern, a quay, and a rock-cut Hellenistic tomb. We revisited the site in August 2007 and just few days after we had commenced the recording and

Further destruction is caused at many sites by the introd tion of agri lt re, arti larl the lti ation of figs and olives. Cattle are also allowed to graze freely around the whole region, including on established archaeological sites. Moreover, seasonal extraction of reeds in marshy lake side areas permits the access of heavy loaders and trucks to a number of the archaeological sites. 150

S. RAMSES & A. OMAR: OUTRAGE AND DESTRUCTION

F ig. 8: K ibotos harbour ex tending from the southern shore of the l ake ( L ake Mareotis R esearch P roj ect) .

F ig. 9: O ne of the univ ersity properties on the l and surrounding the K ibotos ( L ake Mareotis R esearch P roj ect) .

F ig. 10: A rcheol ogical tel l div ided for priv ate possession ( L ake Mareotis R esearch P roj ect) .

F ig. 1 1: Site 201 (

F ig. 12:

F l attened H el l enistic tomb at Site 20 dur

P riv ate property sign and w al l establ ished on L ake Mareotis R esearch P roj ect) .

ing bul l doz er w ork in 207 ( 151

L ake Mareotis R esearch P roj ect) .

LAKE MAREOTIS: RECONSTRUCTING THE PAST

i

s

s

a h

a i sa

i

Conclusion and Recommendation From the above mentioned, the slow destruction of archaeologi al sites in areotis an e s e ified as follows lti ating the o ndaries of ar haeologi al tell sites and in some cases, on the tells themselves. etting ar ed wire aro nd man of the sites to gi e the impression that they are private property, and when the a thorities are notified, the signs are not remo ed sing lldo ers and tra tors to ra idl destro and le el the archaeological hills. elling ar haeologi al sites, es e iall the nregistered sites, to private individuals, by employing unorthodox means that appropriate the names of established people, to ensure that legally the sites cannot be retrieved, thus preventing anyone, even the governmental organizations, from taking any action.

ak

a

is

s a h

Generally, the aim of these suggestions would be to create a project which promotes a particular concept for both the short- and long-term management of the region, which considers its archaeological and historical importance. Such a plan should also aim to promote and preserve Mareotis for the coming generations. This is a great challenge that we all face. Bibliography Abd-Elhady, D., 2008, asa sasya sayaha. Alexandria. A tl as A l moqea A l athariaya 3: El-Beheira Governorate, Egypt. Supreme Council of Antiquities, August 2002. Cairo. A tl as A l moqea A l athariaya 4: The Rest of Lower Egypt Governorates, Egypt. Supreme Council of Antiquities, August 2002. Cairo. Buttler, A.G., F ateh El arab L emesr, M.F. Abo Hadid & T. El-Massrein (transl.), 1989. Cairo. El-Fakharani, F., 1983, Recent Excavation at Marea in Egypt. In G. Grimm, H. Heinen & E. Winter (eds.), A egyptiaca T rev ernsia I I , D as R öm isch- B yz antinische Ä gypten. A kten des international en Symposions 26.30. September 1978 in T rier, 175-186. Mainz. Empereur, J.-Y ., & Picon, M., 1998, Les Ateliers d’ Amphores du Lac Mariout. In J.-Y . Empereur (ed.), Commerce et artisanat dans l ’ A l ex andrie hel l é nistique et romaine. A ctes du col l oque d’ A thè nes, organisé par l e CN R S, l e L aboratoire de cé ramol ogie de L yon et l ’ Ecol e franç aise d’ A thè nes, 1 1- 12 dé cembre 198: 75-88. Athens, Paris. Rodziewicz, M.D., 1998, Fom Alexandria to the West by Land and by Waterways. In J.-Y . Empereur (ed.), Commerce et artisanat dans l ’ A l ex andrie hel l é nistique et romaine. A ctes du col l oque d’ A thè nes, organisé par l e CN R S, l e L aboratoire de cé ramol ogie de L yon et l ’ Ecol e franç aise d’ A thè nes, 1 1- 12 dé cembre 198: 93-103. Athens, Paris. Strabo, G eography, H.L. Jones (transl & ed.), 1996. London, New Y ork.

Accordingly, it has been established that the problems of the region are e tensi e and om le , t on e identified can be resolved by the means of establishing a compromise between protecting Mareotis’ archaeological sites, whilst at the same time meeting the increasing needs of development and modernization. Therefore, the following is suggested to overcome the problems: oordination etween the different regional a thorities including those that operate in the interest of archaeology, investment, irrigation, and the governorate, to establish an integrated long term plan, which recognises a satisfactory om romise that identifies the needs of all arties nif ing the ma ing s stem of ea h organi ation, in order to get acquainted with each others properties, in order to stop the illegal acquisition of land. ort the reme o n il of nti ities, oth to provide the facilities needed to control and protect such a vast area, and to implement archaeological legislation. en o raging lo al interest in the im ortan e of the sites both as a source of income, and as an important aspect of cultural heritage. romoting the whole region to e ome an o en archeological park in order not to be neglected and misused, otherwise it will soon disappear. 152

S. RAMSES & A. OMAR: OUTRAGE AND DESTRUCTION

153

LAKE MAREOTIS: RECONSTRUCTING THE PAST

154

INDEX

Select Index Places/topography Abu Mina, 41, 54, 67-73, 77-78, 81, 115, 127-130, 137 Abuqir Lake, 13, 15-16, 18 Abusir, 15, 35 Abusir Ridge, 36, 43, 142 Alexandria, 1-4, 7-8, 12-13, 15-16, 18-20, 25-28, 32-33, 35-36, 47, 49, 54-55, 57, 66-67, 69, 71-72, 75-78, 8283, 87, 89, 100, 103, 107-108, 110, 113, 115-116, 123125, 127-128, 135, 137, 141-143, 147 Beheira, 37, 107-115, 119, 122, 124-125, 147 Cairo, 13, 15, 67, 107, 123, 125, 128 Canopic Branch/Canopic Nile, 4, 7, 13, 18, 25, 36, 75, 107-108, 110, 116, 119, 122, 124-125, 135, 142-143 Chaereou, 7 Fayum, 44, 110, 127 Huwaryia/Hauwariya, 68-70, 75, 78 Kanubiye Canal, 108, 114 Kom el Giza (see also Beheira), 107-115 Lake Mareotis/Lake Mariout/Maryut Lake, 1-4, 7-8, 1119, 21-23, 25-28, 30-33, 35, 37, 44, 47-48, 57, 66-67, 72-73, 75-77, 83, 87, 95-96, 119, 121, 123-124, 128129, 135-137, 139-143, 147-152 Libya, 1, 3-4, 8, 75-77, 157 Marea/Mareia (see also Philoxenité), 1-4, 7-8, 14-15, 17, 27-28, 30-32, 35, 37, 44, 47-52, 54-55, 57-61, 66-67, 69-73, 75, 77-83, 91, 100, 128-129, 135-139, 141 Mareotic Arm, 32-33, 135-136, 138, 141-143 Mareotis Island, 28-32, 35, 37, 41-42, 44, 135-136, 139142, 147-148 Maryut Depression, 25, 35, 142 Mediterranean Sea, 11-12, 15-17, 19, 25, 27, 127-128 Naukratis Canal, 123, 125 Nile Delta, 36-37, 44, 83, 107, 110, 135, 137 Nile River, 3-4, 7-8, 11-13, 15-16, 18, 25, 27, 35-36, 39, 44, 50, 54, 66, 75, 83, 95-96, 103, 107-108, 110, 116, 125, 127, 135, 142-143 Nile Valley, 3, 27, 35, 41, 44, 54, 103, 127 Nitria/Nitraria, 8, 13-15, 143 Philoxenité/Philoxenite, 27, 35, 47, 54-55, 67, 69, 71-73, 75, 77-78, 83, 135-137, 139 Plinthine, 4, 14, 67, 87-88 Sais, 119, 124 Schedia, 7, 13, 107-110, 114-116, 125, 129, 142 Schedia Canal, 18, 142-143 Taposiris Magna/Taphosiris, 7-8, 27, 67, 71, 87-88, 92, 101, 136-137, 139, 141, 150 Western Desert, 3, 5, 123-124

El-Fakharani, F., 1-2, 27, 31, 37, 47, 49-53, 69-70, 76, 78, 129 El-Falaki, M., 1, 17-18, 37, 47, 57, 67, 75, 137 Herodotus, 3, 7, 54, 73, 75, 137 Justinian, 7-8, 76 Pliny, 4 Procopius, 7-8, 55 Sophronius, 8 St Menas, 51-52, 54, 78 Strabo, 3-4, 7-8, 18, 35-36, 39, 47, 54, 66, 75, 127, 142-143 Theodosius, 8 General amphora/amphorae, 27-29, 32, 37, 43-44, 58, 62, 72-73, 75, 78, 81-83, 87, 90, 100-103, 113, 115-116, 121, 124, 127, 135-136, 141-142 auger cores/auger survey, 30, 37-38 basilica, 51-53, 75-78, 81-83, 129 bath complex/bath facility, 28, 58, 70-72, 75, 110, 116, 128-129 bridge, 42, 89-92, 100, 107-108, 116, 137 camel, 4-5, 55 canal, 3-4, 7-8, 14-15, 18-19, 21, 27, 35-36, 50, 52, 66, 75, 91, 95-96, 107-108, 110, 114, 116, 121, 123-124, 135, 142-143 causeway, 25, 58, 60, 71, 77, 87, 89-91, 93, 95, 100, 135 cistern, 2, 28, 36-37, 42, 53, 58, 69-70, 72, 80, 136, 142, 150 coins, 51-53, 70, 79, 82, 91, 96, 100-101, 103, 107, 110, 113-115, 128 Coptic Encomium of St Menas, 54-55, 69, 72, 77 evaporation, 37, 149 finewares, 28, 121 fish farms, 19-22, 96, 100, 147, 149 fish-tank, 96-100, 103 floods/seasonal flooding, 4, 16-17, 22, 35-37, 39, 43, 54, 87, 90-91, 95, 124, 137, 143 forts, 49, 124 glass, 5, 25, 27, 52, 66, 75, 121, 142 harbours, 4, 7, 25, 27, 31, 38-39, 47-51, 53-55, 57, 66-73, 75-78, 81, 83, 87, 90-91, 93, 95-96, 98, 100-101, 103, 107-108, 110, 116, 124, 136-139, 142-143, 147, 149151 hypogea/hypogeum, 54, 69, 71 jetty/jetties, 28-29, 32, 37, 39, 41-43, 47, 49, 53-53, 58, 60, 67, 71-72, 75, 77-78, 81, 83, 89, 91, 93-98, 100-101, 103, 136, 138-139, 142, 150 karum, 36, 43, 71 Kibotos/Kibotos harbour, 31, 47-48, 71, 75, 91, 139, 150151 kiln, 31-32, 36-37, 43-44, 70-73, 75, 77-78, 82-83, 127, 130, 136 lake port, 67, 70, 72, 75 lake wall, 30, 32, 39, 43, 139-140, 149

People Athanasios, 77 Christodorus, 8 Claudius Ptolemy/Ptolemy the Geographer, 4, 36, 47, 54 Coste, P., 87, 89-91 De Cosson, A., 1, 35, 67, 75, 87, 89

155

LAKE MAREOTIS: RECONSTRUCTING THE PAST lighthouse, 49, 58, 137 mole, 28, 31, 41, 67, 139 natron, 3, 5, 15 papyri/papyrological documents, 4-5, 8, 96, 110, 130-132 papyrus, 7, 54, 75, 142 Pharos, 3, 137 pilgrim, 54-55, 69, 71-73, 77-78, 82, 137 pilgrimage, 66, 68-72, 78 pilgrimage centre, 67, 71-72 pollution, 27-28 port, 14, 25, 27, 42, 47, 49-52, 66-67, 69-73, 75, 78, 83, 96, 123-124, 135 qadus, 28 quarry, 27, 32, 36, 43, 67, 71, 77, 91, 107, 116, 129, 143

quay, 28-29, 32, 39, 41, 43, 47, 49-50, 66-67, 72, 77, 136140, 142, 150 sakkia/sakia/saqiya/sāqiyah, 28, 30, 32, 37, 41, 76, 79-81, 83, 136, 142 sebakh, 107, 110, 123 sewage, 25, 27, 77, 83 tombs, 2, 17, 32, 49, 54, 69, 71, 111, 116, 150-151 villa, 35, 43, 58, 77-78, 111, 113-114, 116, 124 vineyards, 5, 43, 75, 78, 127, 130-132 wine, 4-5, 25, 27, 43, 53-54, 70, 75, 77-78, 83, 101, 103, 121, 127-132, 142, 149 wine production, 30, 32, 70, 114, 127, 130, 132, 135-136, 142, 147 winery, 32, 43-44, 53, 69-70, 123, 127-132, 135, 149-150

156

ISBN 9781407306544 paperback ISBN 9781407336527 e-format DOI https://doi.org/10.30861/9781407306544 A catalogue record for this book is available from the British Library

BAR

PUBLISHING

Contents CONTENTS LIST OF FIGURES, TABLES AND MAPS INTRODUCTION CONTRIBUTORS

i ii vii viii

1. FAWZI EL-FAKHARANI: PIONEER EXCAVATOR AT MAREOTIS. Mona Haggag

1

2. THE MAREOTIC REGION IN ANCIENT SOURCES. Mohamed S. Abd-el-Ghani

3

3. A NOTE ON LAKE MAREOTIS IN BYZANTINE TIMES. Mostafa El Abbadi

7

4. A STUDY OF THE EVOLUTION OF THE MARYUT LAKE THROUGH MAPS. Ismaeel Awad

11

5. LAKE MAREOTIS RESEARCH PROJECT. Lucy Blue

25

6. THE RESULTS OF A PRELIMINARY SURVEY AT MAREOTIS ISLAND. Dylan Hopkinson

35

7. THE CITY OF MAREA/PHILOXENITÉ. REFLECTIONS ON THE ALEXANDRIA

47

UNIVERSITY EXCAVATIONS, 1977-1981. Mona Haggag 8. MAREA PENINSULA: OCCUPATION AND WORKSHOP ACTIVITIES ON THE

57

SHORES OF LAKE MARIOUT IN THE WORK OF THE CENTER D’ÉTUDES ALEXANDRINES (CEAlex, CNRS USR 3134). Valérie Pichot 9. ON INTERPRETATIONS OF ARCHAEOLOGICAL EVIDENCE CONCERNING

67

MAREA AND PHILOXENITE. Mieczyslaw D. Rodziewicz 10. MAREA OR PHILOXENITE? POLISH EXCAVATIONS IN THE MAREOTIC REGION 2000-2007.

75

Krzysztof Babraj & Hanna Szymańska

11. THE LAKE STRUCTURES AT TAPOSIRIS. Marie-Françoise Boussac & Mourad El Amouri

87

12. SCHEDIA, ALEXANDRIA’S HARBOUR ON THE CANOPIC NILE. INTERIM REPORT ON

107

THE GERMAN MISSION AT KOM EL GIZA/BEHEIRA (2003-2008).

Marianne Bergmann, Michael Heinzelmann & Archer Martin

13. RECENT SURVEY WORK IN THE SOUTHERN MAREOTIS AREA. Penelope Wilson

119

14. WINERIES OF THE MAREOTIC REGION. Dorota Dzierzbicka

127

15. WATERFRONT INSTALLATIONS AND MARITIME ACTIVITIES IN THE

135

MAREOTIC REGION. Emad Khalil 16. LAKE MAREOTIS RESEARCH PROJECT. PHASES OF OUTRAGE AND DESTRUCTION.

147

Sameh Ramses & Ahmed Omar

SELECT INDEX

155 i

LAKE MAREOTIS: RECONSTRUCTING THE PAST

List of Figures, Tables and Maps Figures Chapter 4. Ismaeel Awad 1: Maryut Lake region. 2: 1548, Vray portraict de la ville d’Alexandrie en Egypte, Pierre Belon du Mans. 3: 1570, detail from Aegyptus antiqua, including a focused map of the region of Alexandria, Abraham Ortelius. 4: 1588, detail from Africae Tabula VII, Livio Sanuto. 5: 1655, detail from Royaume et désert de Barca et l’Egypte, Nicolas Sanson D’Abbeville. 6: 1717, detail from Carte de la Basse Egypte et du cours du Nil, Paul Lucas. 7: 1722, detail from Carte de l’Egypte Ancienne, Claude Sicard. 8: 1743, detail from Carte d’Egypte, Richard Pococke. 9: 1753, detail from Carte de l’Egypte ancienne et moderne, Robert de Vaugondy. 10: 1764, detail from Carte des Embouchures du Nil, et partie de son cours, Jacques Nicolas Bellin. 11: 1801, Carte d’Alexandrie, French campaign. 12: 1801, Carte d’Alexandrie, French campaign, original scale 1:100,000. 13: 1866, Carte des Environs d’Alexandrie, Mahmoud Bey El-Falaki. 14: 1866, Carte des Environs d’Alexandrie, Mahmoud Bey El-Falaki, original scale 1:200,000. 15: 1911 (1914), Atlas of Egypt, Vol. I, Lower Egypt, Survey of Egypt (G. Daressy), original scale 1:50,000. 16: 1949, map series 1:100.000. Survey of Egypt, original scale 1:100,000. 17: 1970s, map series 1:25.000, Egyptian General Survey Authority, original scale 1:25,000. 18: 1991, map series 1:50.000, Egyptian General Survey Authority. 19: 2001, Landsat satellite image, resolution 15m/pxl. 20: Evolution diagram of Maryut Lake’s surface areas (1801-2001). 21: Comparison between Maryut Lake 1801, French campaign map series 1:100,000 and 0 m level in the Mareotic region. 22: Comparison between Maryut Lake 1801-2001. Chapter 5. Lucy Blue 1: Location of Alexandria and Lake Mareotis. 2: An approximation of the ancient and present limits of Lake Mareotis. 3: Sites located along the northern shoreline, the southern shoreline and on Mareotis Island. 4: Topographic, ceramic and auger survey plan of Site 44 on Mareotis Island. 5: An example of a ‘lake wall’ feature (Site 44). 6: The Kibotos site (Site 09). 7: The ‘complex building’ (Site 13).

11 12 12 13 13 14 14 14 14 15 16 16 17 17 18 18 19 20 20 21 22 23 26 27 29 30 30 31 31

Chapter 6. Dylan Hopkinson 1: The location of Mareotis Island and the sites mentioned in text, modified from De Cosson 1935. 2: Schematic of the main features of Mareotis Island. 3: The littoral features mentioned in this paper and topographic survey of the eastern settlement of the island. 4: The littoral features mentioned in this paper and topographic survey of the western settlement of the island. 5: Longitudinal and transverse schematic sections through the ‘water tower’ on Mareotis Island.

42

Chapter 7. Mona Haggag 1: Kibotos harbour structure on the south shore of Lake Mareotis (Site 9 – Lake Mareotis Research Project). 2: Kibotos harbour structure, mooring ring. 3: Rock-cut Kibotos harbour structure at Ikingi Mariut. 4: Byzantine harbour of Marea. 5: Plan of the eastern harbour of Marea, after El-Fakharani 1977. 6: Decumanus between the eastern and middle ports of Marea. 7: Plan of shops off the main decumanus. 8: Basilica, Marea.

47 48 48 50 51 51 51 52

ii

35 36 38 40

9: Enigma Buliding, Marea. 10: Winery excavated by El-Fakharani 1983.

52 53

Chapter 8. Valérie Pichot 1: Map of Egypt: location of Lake Mariout and the site of Marea. 2: General plan of the site identified as Marea. 3: The peninsula of Marea. 4: A) pier to the north of Marea peninsula, view from the north-west. B) ancient causeway connecting the peninsula to the mainland, view from the west. 5: Ancient causeway under excavation of Sector 1, view from the east. 6: Ancient causeway under excavation of Sector 2, view from the north. 7: Landmark and jetty to the south of the peninsula, view from the south-west. 8: Results of the geophysical survey undertaken on the Marea peninsula. 9: Sector 4 - Hearths in FS4024, under excavation. 10: Overall plan of the western zone of Sector 3. 11: Overall plan of the eastern zone of Sector 3. 12: Eastern zone of Sector 3 under excavation. 13: Sector 3 - One of the two limestone sphinxes discovered in the tower building ST300. 14: Sector 3 - Bronze candelabra after restoration in the CEAlex laboratory. 15: Sector 3 - Hoe chalk burner. Chapter 9. Mieczyslaw D. Rodziewicz 1: Central part of the plan of Mareotis drawn by Müller-Wiener in 1966, with the pilgrimage road from the lake harbour to Abu Mina and settlements, among them settlement “M” in Huwaryia village. 2: Abu Mina with surrounding settlements/karms, and pilgrimage road towards the lake harbour. 3: Byzantine House/Hospitium at the pilgrimage road from the lake harbour to Abu Mina. 4: Byzantine pilgrimage harbour/Philoxenite (Nos. 5-17) and western/coastal rural remains (Nos. 1-4). On the eastern side, structured causeway (No. 18). Chapter 10. Krzysztof Babraj & Hanna Szymańska 1. Remains of structures along the southern shore of Lake Mareotis. 2. Marea. Byzantine baths. 3. Marea. Baths and western courtyard. View from the south-west. 4. Marea. Well operated with sāqiyah. View from the south. 5. Marea. Funerary chapel. View from the west. 6. Marea. Plan of the basilica. 7. Marea. Apse of the basilica. View from the west. ‘A’ and ‘b’ locate the two burials. 8. Marea. Basilica. Amphora kiln. View from the west. Chapter 11. Marie-Françoise Boussac & Mourad El Amouri 1. General map of the site and location of areas under study at Taposiris Magna and Plinthine. 2. Drawing by Pascal Coste, made in 1820. 3. Map of the remains of the ancient city of Taposiris by A. De Cosson, 1935. 4. Location of the different sectors of the site of Taposiris that were studied along the lake shore. 5. View of the Taposiris bridge from north. 6. Plan and elevation of the north pile of the bridge, November 2004. 7. Model and reconstruction of the bridge in Taposiris Magna. 8. Overall plan and detailed outline of the eastern jetty. 9. The southern end of the eastern jetty. 10. Jetty at Gamal. View from north-west. 11. Jetty at Quseir. View from south-east. 12. Plan and elevation of the inner south face of the north flush and buttress. 13. View of the north flush during excavation, from east. 14. Lay out, elevations and sections of north and south walls of the water supply channel through the eastern jetty. 15. Water supply channel through the eastern jetty opening into the hoop-shaped basin. 16. Grooves and notches in the water supply chanel. 17. View of the hoop-shaped basin, the water supply channel and the fish-tank. iii

57 58 59 60 60 60 60 61 62 63 64 65 65 65 66 68 68 70 71

76 78 79 80 80 81 82 82 88 89 89 90 92 92 92 93 93 94 94 94 95 97 98 98 99

LAKE MAREOTIS: RECONSTRUCTING THE PAST 18. 19. 20. 21.

The pavement in the hoop-shaped basin. The opening between the hoop-shaped basin and the fish-tank. Main types of ceramics found in Taposiris Magna (Sector 9) during the Late Roman era. Ceramics from the filling of the hoop-shaped basin.

Chapter 12. Marianne Bergmann, Michael Heinzelmann & Archer Martin 1: Alexandria and its hinterland in 1925, Survey of Egypt, sheet 4. 2: Schedia. Topographical map based on quickbird satellite image (2005) indicating excavated areas (A) and corings (H). 3: Schedia. Map with reconstruction of the ancient topography. 4. Schedia. Plan of the southeastern settlement (Kom el Giza, Kom el Hamam) with excavation areas 1980-92 and 2003-06. 5: Kom el Giza, Area 3. Plan of Hellenistic baths. 6: Kom el Giza, Area 1. Plan of Roman villa, tombs and later basins. 7: Kom el Hamam. Plan with excavations areas 2003-2006. German Mission at Kom el Giza/Beheira. 8: Kom el Hamam. Plan of Area 6 with Roman storage building and later phases. 9: Kom el Hamam. Reconstruction of the Roman storage building. 10: Kom el Hamam, Area 9. Roman enclosure wall. 11: Kom el Hamam. Plan of Area 8 with basins for wine-production.

99 99 101 102 108 109 109 110 111 111 112 112 113 114 115

Chapter 13. Penelope Wilson 1: Ancient sites in the Delta, highlighting the Mareotis region. 2: Map of the sites surveyed, with 0 metre sea-level contour area highlighted in grey. 3: The main mound at Kom el-Mahar. 4: Late Ptolemaic to Early Roman pottery sample from Kom el-Mahar. 5: Red brick structures at Kom Trugi, excavated below foundation level. 6: Section through the cemetery mound at Kom el-Qadi.

119 120 121 122 123 123

Chapter 14. Dorota Dzierzbicka 1: A map of the Mareotic region with marked locations of wineries. 2: A schematic drawing of a Type 4 winery with names of its elements used in Greek papyri from Egypt.

127 130

Chapter 15. Emad Khalil 1: The approximate ancient and present limits of Lake Mareotis. 2: The Lake Mareotis Research Project survey area along the shores of the western Mareotic Arm. 3: Taposiris Magna tower. 4: The middle quay of the harbour of Marea. 5: The quay of Marea harbour was made of large regular limestone blocks without evidence of the use of mortar. 6: The box-shaped harbour on the southern shore of Lake Mareotis. 7: One of the quays extending into the lake at the northern shore of Mareotis western arm. 8: A lake wall extending parallel to the southern shore of the lake. Evidence for red mortar (opus signinum) can still be seen between the blocks. 9: Some of the multi-room square buildings located at Site 118 along the northern shore of the Mareotis Island. The buildings could have been used for storage purposes. Chapter 16. Sameh Ramses & Ahmed Omar 1: General plan from Alexandria to Mareotis. 2: Map of sites investigated during the Lake Mareotis Research Project. 3: Two examples of fish farm deformation along the southern shore of the lake. 4: Lime storing and collecting at Site 44. 5: Lake wall affected at Site 44. 6: Site 214: the partially excavated winery. 7: Site 214: part of the fragmented paved floor of a Roman period waterfront building. 8: Kibotos harbour extending from the southern shore of the lake. 9: One of the university properties on the land surrounding the Kibotos. 10: Archeological tell divided for private possession. 11: Private property sign and wall established on Site 201. 12: Flattened Hellenistic tomb at Site 202 during bulldozer work in 2007. 13: Two soccer fields erected at the tell boundaries around Site 204. iv

135 136 137 138 138 139 140 140 141

147 148 149 149 149 150 150 151 151 151 151 151 152

Tables Chapter 4. Ismaeel Awad 1: Comparison table between different periods of Maryut Lake.

21

Chapter 13. Penelope Wilson 1: Summary table of sites surveyed by the ‘Sais and its Hinterland Project’ in the region of southern Lake Mareotis.

124

Chapter 14. Dorota Dzierzbicka 1: Installations forming part of a wine-making complex mentioned in Greek papyri containing references to immovable property found in vineyards or to the appurtenances of a winery.

131

Maps Chapter 7. Mona Haggag 1: Marea in Lake Mareotis. 2: Marea’s public buildings on the lake shore.

48 50

v

LAKE MAREOTIS: RECONSTRUCTING THE PAST

vi

Introduction The following collection of papers represents the final synthesis of a conference entitled The International Conference on the Archaeology of the Mareotic Region. Lake Mareotis: Reconstructing the Past hosted by the University of Alexandria, Egypt between 5th & 6th April 2008. The objective of the conference was to bring together scholars that work in the Lake Mareotis region of Alexandria to a forum where they could present and discuss their ongoing research and field projects. The two-day conference was a great success as it provided an opportunity to share data and tackle themes and issues of common concern. It was also an opportunity for the Lake Mareotis Research Project, a collaboration between the Centre for Maritime Archaeology, at the University of Southampton, and the Department of Underwater Antiquities (DUA) of the Supreme Council of Antiquities (SCA), Alexandria, in conjunction with the Centre for Maritime Archaeology and Underwater Cultural Heritage, University of Alexandria, to show case the Lake Mareotis survey project and present recent results to colleagues. Lake Mareotis is a shallow body of brackish water currently about 90 km2 in area that extends south of Alexandria for about 12 km and to the west in the form of an arm some 40 km and is 3 km wide. Lake Mareotis was much larger in antiquity having been subject to siltation and reclamation over the last two thousand years. Due to its location and characteristics, Lake Mareotis it is known to have contributed significantly to the ancient economy of Alexandria and to the internal and international transport system of Greco-Roman Egypt. The significance of this region has been acknowledged since at least the 1930’s (De Cosson 1935), and subsequently highlighted by the pioneering work of scholars such as ElFakharani (1983). This volume is published in honour of these early pioneers, in particular El-Fakharani who not only initiated scholarly investigation of a number of sites in the Lake Mareotis region but who also engaged for the first time, future Egyptian scholars by including and encouraging students of Archaeology from the University of Alexandria to participate in excavations at Marea (see Haggag this volume). Thus, the conference and this publication attempts to follow in the footsteps of El-Fakharani by seeking to encourage an inclusive dialogue between all scholars currently active in the region. The following papers represent the most recent attempt to contextualise and interpret Lake Mareotis. Current research, not only that generated as a product of the Lake Mareotis Research Project (Blue & Khalil, forthcoming), but also exciting new discoveries from other sites around the shores of the lake, are beginning to provide a more comprehensive understanding of the breadth and scale of activities conducted around the shores of Lake Mareotis and highlight the substantial contribution the region made to the ancient economy of Alexandria. There is still much to be discovered but with continued dialogue we can begin to tackle unanswered questions and resolve sometimes contradictory, interpretations. This volume therefore presents a collection of thoughts and perspectives that are often variable in their focus and approach. A degree of editorial consistency has been employed but in order to retain the particular and diverse perspectives of the individual authors, specific nomenclature has been maintained. To that end, a variety of spellings for Lake Mareotis (Maryut, Mariout) and sites in the Mareotic region such as Marea, Mareia etc., are supported. It is hoped that this volume will encourage future discussion and publications that continue to highlight the importance of the Lake Mareotis region in antiquity. Lucy Blue Acknowledgements The editor would like to acknowledge both the British Academy and the Leverhulme Trust who provided financial support to undertake the Lake Mareotis Research Project of which this volume is just one product. In addition, the support of Dr Emad Khalil and the University of Alexandria is acknowledged for hosting and organising the conference. I would also like to thank the anonymous reviewer of this volume for their helpful comments. Bibliography El-Fakharani, F.A., 1983, Recent Excavation at Marea in Egypt. In G. Grimm, H. Heinen & E. Winter (eds.), Aegyptiaca Trevernsia II, Das Römisch-Byzantinische Ägypten. Akten des internationalen Symposions 26.-30. September 1978 in Trier, 175-186. Mainz. Blue, L., & Khalil, E. (eds.), forthcoming, A multidisciplinary approach to Alexandria’s economic past: The Mareotis case study. Southampton Monograph Series. Oxford De Cosson, A., 1935, Mareotis: Being an Account of the History and Ancient Monuments of the North-West Desert of Egypt and Lake Mareotis. London vii

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Contributors Marie-Françoise Boussac Professor in Greek History Paris Ouest Nanterre University 92000 Nanterre, France [email protected]

Mostafa El Abbadi Professor Emeritus of Greco-Roman Studies Faculty of Arts University of Alexandria Hussein Sobhi St. El-Shatbi 21526 Alexandria, Egypt [email protected]

Delphine Dixneuf Institut Français d’Archéologie Orientale 37, rue al-Cheikh Aly Youssef B.P. Qasr al-Ayni 11562 11441 Cairo, Egypt [email protected]

Mohamed S. Abd-el-Ghani Professor of Ancient History Dept. of Archaeology and Greco-Roman Stiudies Faculty of Arts University of Alexandria Hussein Sobhi St. El-Shatbi 21526 Alexandria, Egypt

Dorota Dzierzbicka Department of Papyrology Institute of Archaeology University of Warsaw Krakowskie Przedmieście 26/28 00-927 Warsaw 64 Poland [email protected]

Mourad El Amouri Ipso Facto Bureau d’Étude et de Recherche, Archéologie et Océanographie 4 rue de Tilsit 13006 Marseille, France [email protected]

Thomas Faucher ANR-Nomisma Université Paris-Sorbonne-Paris IV 1, rue Chevalier de la Barre 75018 Paris, France [email protected]

Ismaeel Awad Topography Department Centre d’Études Alexandrines 50, rue Soliman Yousri 21131 Alexandria, Egpyt [email protected]

Mona Haggag Bibliotheca Alexandrina P.O. Box 138 El-Shatbi 21526 Alexandria, Egypt [email protected]

Krzysztof Babraj Archaeological Museum in Kraków ul. Senacka 3 31-002 Kraków, Poland [email protected]

Michael Heinzelmann University of Cologne Institute of Archaeology Albertus Magnus Platz D-50923 Cologne [email protected]

Marianne Bergmann Georg-August-University Goettingen Archaeological Institute Nikolausberger Weg 15 D-37073 Goettingen [email protected]

Dylan Hopkinson Centre for Maritime Archaeology University of Southampton Highfield Southampton, England SO17 1BJ [email protected]

Lucy Blue Director, Centre for Maritime Archaeology University of Southampton Highfield Southampton, England SO17 1BJ [email protected]

viii

Sameh Ramses Supervisor of Egyptian Excavations & Surveying Team Department of Underwater Archaeology Supreme Council of Antiquities (SCA) 19 Abdel Moety el Gazaly, Stanley Alexandria, Egypt [email protected]

Emad Khalil Centre for Maritime Archaeology & Underwater Cultural Heritage Faculty of Arts University of Alexandria Hussein Sobhi St. El-Shatbi 21526 Alexandria, Egypt [email protected]

Mieczyslaw D. Rodziewicz Ul. Literacka 25 m.10 01-864 Warsaw, Poland [email protected]

Archer Martin American Academy in Rome Via Angelo Masina, 5 I-00153 Roma [email protected]

Hanna Szymańska Research Centre for Mediterranean Archaeology University of Warsaw 11, Sharia Mahalla, Heliopolis Cairo, Egypt [email protected]

Ahmed Omar Inspector, Department of Underwater Archaeology Supreme Council of Antiquities (SCA) 19 Abdel Moety el Gazaly, Stanley Alexandria, Egypt

Penelope Wilson Department of Archaeology Durham University South Road Durham, England DH1 3LE [email protected]

Valérie Pichot Centre d’Études Alexandrines 50, rue Soliman Yousri 21131 Alexandria, Egpyt [email protected]

ix

LAKE MAREOTIS: RECONSTRUCTING THE PAST

x

Fawzi El-Fakharani: Pioneer Excavator at Mareotis Mona Haggag

This volume is dedicated to the memory of Fawzi El-Fakharani who was, in many ways, a pioneer investigator of the area of Mareotis.

certain or clear that it was a temple dedicated to Hercules (Kanellopoulos 1994: 60-65). In 1970 Fakharani was delegated to Libya, where he established a department of Classical Archaeology. In Libya, Fakharani excavated the area of the Byzantine palace of the City of Tocra, ancient Taucheira. The patriarch, who presented the city at the council of Nicea in 325 CE, must have resided in this palace. Fakharani uncovered parts of the Eastern Church to which the palace was attached. He also unearthed the church s confirmatorium and baptistery, as well as two beautiful mosaic panels (pers. comm.).

Fawzi El-Fakharani was born in Alexandria in 1921. He graduated from the Faculty of Arts, University of Alexandria in 1946 and directly after his graduation he went to the United Kingdom for higher studies. In 1953 he obtained a Masters degree from the University of Liverpool. His undergraduate dissertation was entitled “The origins of stucco reliefs in Classical Art”. Afterwards he moved to University College London where he acquired a doctoral degree in 1957 on “Stucco Reliefs in Roman Art”.

In 1973, back in Alexandria, Fakharani began his studies of the city of Marea, thus realising his life’s dream, which was to discover and document the ruins of this city. After a long and exhaustive period of preparation, he finally obtained the permission to excavate the site identified previously as the city of Marea, to the south of the western extremity of Lake Mareotis, some 45 km distance from Alexandria. With the exception of previous identifications of the site as that of Marea made by Mahmoud Pasha ElFalaki (1872: 85, 86, 91, 96, 97, 101), and the valuable monograph on Mareotis by De Cosson (1935: 131-135), the academic community of archaeologists and researchers had paid no particular interest in the area of Mareotis. Fakharani’s determined enthusiasm and efforts to identify the city’s area proved successful, as the results of his excavations shed light on the various stages of life within the city and on the importance of the area as a whole (El-Fakharani 1983: 175-204). Some of the articles of this volume are devoted to Marea (see Haggag, Pichot, Rodziewicz, Babraj and Szymańska through which the reader can figure out that the discovery of Marea was and is, of great importance to researchers, as well as to the community at large.

Fakharani returned to Egypt and was appointed lecturer in the University of Alexandria in 1958. In 1964 he was promoted to an assistant professor, and won the chair of Professor of Classical Archaeology in 1974. He was appointed Chair of the Department of Classical Civilization between 1976 and 1980, and the director of the Institute for Mediterranean Studies between 1986 and 1989. In 1990 he became Chair of the Department of Archaeology at the University of Tanta, Egypt. In 1966 Prof. Fakharani was delegated to the University of Amman in ordan. There, he began his first endeavours in archaeological fieldwork. His excavation target was to continue uncovering the grand Roman theatre of Amman. He published the results of his investigations in the J ahrbuch Des Deutschen Archäol ogischen Instituts Archäol ogischen Anzeiger, in 1975 (El-Fakharani 1975a: 377-403). In 1967, his interests shifted towards the site of Qal’at AlJabal on which stands the so-called Temple of Hercules (El-Fakharani 1975b: 533-554). There he reinvestigated the remains and uncovered most of the building’s foundations and annexed colonnades. Scholars differed with regard to which deity the temple was dedicated, yet Fakharani’s hypothesis was that the building had characteristic features unknown to temples of the East Roman Provinces in the time of Marcus Aurelius, although common in libraries of the same epoch in the neighbouring countries (El-Fakharani 1975b: 554). Fakharani reached the conclusion that the building was not a temple but the library built by Marcus Aurelius at Philadelphia, the city that he described as Alexandria’s daughter city. Despite his reasonable arguments and study published in the periodical of the University of Rostock, the building was still termed as the “Temple of Hercules”. This remained the case until a study conducted in 1994 of an inscription carved on the architrave of the building showed that it was not at all

With only ten thousand Egyptian pounds as a budget, Fakharani was given the daunting task of locating a suitable area to undertake a bold and pioneering step in Egyptian archaeology, his goal being to accommodate both male and female Egyptian students in the field. This was something that had never been undertaken before and had to be managed with great care and respect for social traditions. Not only did he have to secure separate sleeping and study areas but he also had to supply electricity to the camp, food and sustenance for the students, and transportation to the desert area of Mareotis. All of this he accomplished on a meagre budget. Besides the difficulties of obtaining permission from the university to take the students to the field, he had also to convince their parents, especially those of the female students, that their daughters would be safe and 1

LAKE MAREOTIS: RECONSTRUCTING THE PAST secure. It was not easy for Egyptian families during the 1970s to send their daughters to the desert for field training, but Fakharani managed to allay any of the families fears and worries.

German team. They all emphasised that the remains from the Ptolemaic era lay only in the area of the nursery which then belonged to the Faculty of Agriculture, the University of Alexandria. Again, it took determination and patience to acquire the permission to excavate this site. During this time, Fakharani excavated a few trial trenches in the area around the remains of an alabaster tomb. Although the trenches yielded nothing of conspicuous importance, a Byzantine well just next to the tomb shed new light on the use of this area during this period. As the level of the monolithic pavement of the early Ptolemaic tomb was higher than that of the Byzantine rim of the uncovered well, it became evident that the tomb was not in situ as Adriani had assumed (Adriani 1940: 15-23), and that it had been moved to its present location during the Byzantine or early Islamic periods (author’s interpretation).

Soon after the ancient buildings emerged from the desert sands, articles and interviews in local and international media started to appear. These articles helped win the support of the families of the students who had been allowed to camp in the desert with Fakharani on this pioneering expedition. The student families soon realised from these articles and media reports just how important the work was that their young daughters and sons were accomplishing. Thus, Prof. Fakharani overcame all obstacles and all the students, including myself, who participated in the excavations, were, for the first time in the history of the department, to become the lucky generation of students who had the opportunity to be trained in the field. We are extremely grateful to Prof. Fakharani for that experience. The directors of the Lake Mareotis Research Project, Lucy Blue, Sameh Ramses and Emad Khalil (to whom this volume owes its inspiration), are now reminding us of those wonderful days by involving students in their current fieldwork surveying the shores around Lake Mareotis, the first time this has happened since the pioneering days of Fakharani.

At the same time that Prof. Fakharani received the permission for his excavation of the nursery, Parkinson’s disease had already begun digging its deadly roots into his aging body. Accordingly, he entrusted the project to another great scholar who has also made considerable contributions to the archaeology of Alexandria and its environs, namely Jean-Yves Empereur. In March 2004, Fakharani passed away but he left behind a generation of empowered, highly motivated, committed and devoted students who strive to continue their work to the standards and example set by Fawzi El-Fakharani.

Prof. Fakharani never once lost sight of the idea that Marea was the “Pompey of the East”, to use his own expression (El-Fakharani 1977: 5). It deserves more care and diligence as it represents a unique archaeological site that gives us a complete picture of the everyday life of ordinary people, in contrast to other archaeological projects in Egypt which have a tendency to focus on discovering how our once noble Pharaonic families lived and prepared for the afterlife. Marea, together with the results of survey and excavations from other settlements around the shores of the Lake (see Blue, Hopkinson, Boussac & El Amouri, Bergmann, et al., Wilson, Dzierzbicka, Khalil, and Ramses & Omar this volume), illustrate how people in a postPharaonic era lived, what their beliefs were, what their baths looked like and how they were used, the shape and layout of their shops, cisterns, and how they buried their dead. Marea necessitates more preservation, conservation, restoration and preparation efforts to emphasise the site’s potential, in order to amongst other things, place it firmly on the tourist map of Egypt.

Bibliography Adriani, A., 1940, Fouilles et decouvertes. Alexandrie. I. Tombeau en alabaster du cimetiere latin. Annuaire 3, 1935- 1939: 15-23. Alexandria. De Cosson, A., 1935, Mareotis: Being an Account of the History and Ancient Monuments of the North-West Desert of Egypt and Lake Mareotis. London El-Fakharani, F.A., 1975a, Das Theater von Amman in Jordanien. J ahrbuch des Deutschen Archä ologischen Instituts Archaologischer Anzeiger 3: 377-403. El-Fakharani, F.A., 1975b, The Library of Philadelphia (?) Or The So-Called Temple on the Citadel Hill in Amman. Wissenchaftlische Z eitschrift der U niversität Rostok XXIV.6: 533-554. El-Fakharani, F. A., 1977, A Preliminary Report on the 1st Season of Excavation at Marea. Excavation Report presented to the Antiquities Department, Egypt. El-Fakharani, F.A., 1983, Recent Excavation at Marea in Egypt. In G. Grimm, H. Heinen & E. Winter (eds.), Aegyptiaca Trevernsia II, Das Römisch-Byzantinische Ägypten. Akten des internationalen Symposions 26.-30. September 1978 in Trier, 175-186. Mainz. El-Falaki, Mahmoud Bey, 1872, Memoire sur l’antique Alexandrie, ses faubourgs, ses environs. Copenhagen. Kanellopoulos, C., 1994, The Great Temple of Amman, Vol. I: Architecture. American Center of O riental Research – Amman: i-xii.

inally, in 1999, akharani again took us to the field. This time akharani s field was in Alexandria itself. His investigations were of the Latin cemetery which he believed to have been the site of the royal Ptolemaic burial ground. A geophysical survey was conducted in three successive phases: firstly, by colleagues from Alexandria niversity, secondly, in collaboration with Mrs. Calliope Lemniou Pappacosta, a researcher from Greece in consultation with experts from the University of Patras, and thirdly, by a

2

M. ABD-EL-GHANI: ANCIENT SOURCES

The Mareotic Region in Ancient Sources Mohamed S. Abd-el-Ghani

and on the south Lake Marea, also called Mareotis: “This is filled by many canals from the ile, both from above and on the sides …” (Strabo 17.1.7).

Before addressing the Mareotic region in antiquity it would be appropriate to explore how Egyptian frontiers are discussed in the classical sources. There is a crucial and decisive paragraph in Strabo (17.1.5) to this effect – one that makes particular reference to the western frontiers of Egypt – the primary focus of this paper – and is quoted “verbatim” as follows:

What is particularly interesting to note about the inhabitants of the Mareotic region, and the Western Desert of Egypt in general, is the ethnic distinction that characterises them apart from the rest of the Egyptians, even after their identity as Egyptians was settled from an administrative viewpoint. It is a phenomenon that is observed throughout the ancient texts through to the present day. This contrast between the Mareotes and the Egyptians is explicitly expressed in a document dated from the middle of the 4th century AD about the smuggling of natron.1

“Now the early writers gave the name Egypt to only the part of the country that was inhabited and watered by the Nile, beginning at the region of Syene and extending to the sea but the later writers down to the present time have added on the eastern side approximately all the parts between the Nile and the Arabian Gulf (i.e. the present Red Sea), and on the western side the parts extending as far as the oases, and on the sea-coast the parts extending from the Canobic mouth to Catabathmus (the present el-Salloum) and the domain of the Cyrenaeans. For the kings after Ptolemy became so powerful that they took possession of Cyrenaea itself and even united Cyprus with Egypt. The Romans, who succeeded the Ptolemies, separated their three dominions and have kept Egypt within its formers limits.”

Marea is stated by Herodotus as being located at the key position with regard to the Egyptian frontiers to the west. He relates that king Psammetichus I of the 26th Dynasty posted garrisons at Elephantine on the side of Ethiopia, at Daphnae of Pelusium on the side of Arabia and Assyria, and at Marea on the side of Libya (Herodotus II.30). Marea was also mentioned by other classical authors as the battlefield of some pitched battles in antiquity before the conquest of Egypt by Alexander the Great. In his reference to the second Egyptian revolt against the Persian rule under King Artaxerxes (c. 460 BC), Thucydides pointed out that Inaros son of Psammctichus, “setting out from Maria, the city just north of Pharos, caused the greater part of Egypt to revolt from king Artaxerxe” (Thucydides I.104).

Among the early writers referred to above by Strabo, it was Herodotus (II.17) who adopted the ancient theory that Egypt was the land inhabited by Egyptians and was watered by the Nile. Nevertheless, he considered that the Egyptians were not only the inhabitants of the Nile Valley and the Delta proper, particularly the area that was directly watered by the Nile River, but also those places to which the Nile waters extended. In order to illustrate this, he set out to write a significant account about the inhabitants of the cities of Marea and Apis. His account, in this concern, goes as follows:

Diodorus of Sicily reported that the decisive pitched battle between King Apries, the legitimate king (588-566 BC) of the Saite Dynasty, and Amasis II, one of the king’s generals and claimant to the throne who became a renowned king, took place near the village of Maria (Diodorus Siculus I.68),2 after the former’s defeat by the Greeks of Cyrenê and Barcê .

“The men of the cities of Marea and Apis, in the part of Egypt bordering on Libya, thinking themselves to be not Egyptians but Libyans, and misliking the observance of the religious law which forbade them to eat cow s flesh, sent to Ammon saying that they had no part or lot with Egypt: for they dwelt outside the Delta and did not consent to the ways of its people, and they wished to be allowed to eat of all food. But the god forbade them: all the land, he said, watered be the Nile in its course was Egypt and all who dwelt lower than the city Elephantine and drank of that river’s water were Egyptians.” (Herodotus II.18)

The first specific and detailed account about Lake Marea and the Mareotic region was that of Strabo. After his statement that Lake Mareotis was filled by many canals from

1. P. Lond. II.231, p. 285 - Hunt and Edgar, Select papyri II. 428 = P. Abinn. I.9, ll. 4-5: ὥστε ὅσα νίτρα καταλαμβάνεις εἴτε δια Μαρεωτῶν εἴτε δια Αἰγυπτείων κατερχόμενα. 2. Cf. Herodotus II.169 where he identified the site of that battle to be at the town of Momemphis “ἐς Μώμεμφιν πόλιν” which is roughly identified as Kom Abu Billu by Ball 1942: 18 oyotte & Charvet 1997: 118, n. 249 identified it as aujourd hui K m el-Hisn in their commentary on Strabo 17.1.22

That they drank of that river’s water is well attested in Strabo s account of Alexander s first visit to the place of Alexandria, the advantages of the site, and his determination to found a city bearing his name on the site. Among the advantages of the (new) city’s site was its location between two (seas), on the north the Egyptian sea, as it is called, 3

LAKE MAREOTIS: RECONSTRUCTING THE PAST the Nile, both from above and on the sides, he goes on to complete the picture of the whole region as follows:

sites of the “Libyan coastal nome” from immediately before Catabathmus (es-Salloum) to Glaucus Promontory (El-Imâ yid) (V. 5. 4 - 7), he goes on to report the sites of the “coastal Mareotic Nome” or “Νομοῦ Μαρεώτου παράλιος”, i.e., namely, the village of “Cheimw” (El-Bordâ n) and Plinthine (ruins 6 km NE of Sidi Kireir) (Ptolemaios IV.5.8). Finally, he comes further east to Chersonesos the Smaller (El-Dekheila) and then to Alexandria, the capital of the whole of Egypt, to quote Ptolemy (Ptolemaios IV.5.9).

“...and through these canals the imports are much larger than those from the sea, so that the harbour on the lake was in fact richer than that on the sea, and here the exports from Alexandria also are larger than the imports and anyone might judge, if he were at either Alexandria or Dicaearchia (Puteoli in Italy) and saw the merchant vessels both at their arrival and at their departure, how much heavier or lighter they sailed thither or therefrom. And in addition to the great value of the things brought down from both directions, both into the harbour on the sea and into that on the lake, the salubrity of the air is also worthy of remark. And this likewise results from the fact that the land is washed by water on both sides and because of the timeliness of the Nile’s risings … at Alexandria, at the beginning of the summer, the ile, being full, fills the lake also and leaves no marshy matter to corrupt the rising vapours. At that time, also, the Etesian winds blow from the north and from a vast sea, so that the Alexandrians pass their time most pleasantly in summer.” (Strabo 17.1.7.)

In another passage of Ptolemy (13-32) he gives a list of the villages inland of the Lybian ome from west to east the last village to the east being the one called “Mareotis”. In another passage he provides another list of the “cities and villages of the Mareotic Nome” (the part inland), further to the east on the confines of the previous Lybian ome: the farthest village to the east being that called the “village of ancient Marea” or Παλαιμάρεια κώμη (Ptolemaios IV.5.34). According to Ptolemy’s calculations the village of Mareotis, the farthest point to the east of the Libyan Nome, was situated on longitude 58 (27º .25’) and latitude 28º .20’, however, the village of ancient Marea is located at longitude 60 (29º .25’) and latitude 30º .10’. Thus, the village of “Mareotis” was situated to the south-west of the village of “Ancient Marea”, at a location yet to be identified with a specific ancient site. Ball 1942: 114 could not identify the present location of either site.

From the above paragraph of Strabo the advantages of Lake Mareotis with its vivid harbour, as well as that of the Mareotic region in general, is quite obvious: intensive commercial activities of imports and exports from the interior of Egypt as well as Mediterranean cargoes an excellent and healthy climate which was also a tourist attraction in itself. Another passage Strabo (17.1.14) sheds more light on the size of the lake and the other activities of the inhabitants of its surrounding shores. He describes the situation as follows:

Some of these aspects referred to in the works of the classical authors concerning the Mareotic region are reflected in papyrological documents, scarce though they are. In the light of such documents the region of Mareotis constituted an administrative “nome” during the 1st (OGIS 669 = S.B. V.8444 AD 68 , l. 48 P. Iand. IV.53, col. 3, l. 1 AD 88 , 2nd P. Ryl. II.78, l.8 AD 157 P. slo III.84 AD 138161 ll. 12-14 P. Haun. II.22, l. 3. and 3rd (P. Flor. III.338, l.6 BG I.13, AD 289, l.2. centuries AD.

“Lake Mareia, which extends even as far as this [ the Chersonesus/ Marabit according to oyotte & Charvet 1997 or el-’Dekheila’ according to Ball 1942] , has a breadth of more than one hundred and fifty stadia and a length of less than three hundred. It contains eight islands and all the shores round it are well inhabited and the vintages in this region are so good that the Mareotic wine is racked off with a view to ageing it.”

About a century after Strabo’s visit to Egypt, Pliny the Elder in his Historia Naturalis reported about Lake Mareotis as follows:

The second important aspect highlighted in these few documents is the close relationship between the Mareotic Nome and Alexandria. Some individuals in these documents would come to Alexandria, through the Mareotic ome, for business or official or judicial purposes. This is clear in some private or business letters P. Haun. II.22 P. Flor. III.338). Sometimes they hint at some degree of co-ordination between the high officials of Mareotis and Alexandria when necessity requires. In a poorly preserved document, the essence of which could be understood in its general lines owing to the existence of several key words, two persons domiciled in the Mareotis region are mentioned in a petition to the strategos of Alexandria regarding the theft of two camels belonging to them. According to the publisher s reading and filling of the lacunae, which seems almost plausible, they ask the Alexandrian strategos to write to his Mareotic counterpart to summon some suspected person to present himself before him for investigation (P. Oslo III.84). Why was the Alexandrian strategos asked to interfere instead of resorting directly to the Mareotic strategos? Does this further illustrate a degree of collaboration between the two strategoi?

“Lake Mareotis, which lies on the south side of the city, carries traffic from the interior by means of a canal from the Canopic mouth of the ile it also includes a considerable number of islands being 30 miles across and 250 miles in circumference, according to Claudius Caesar.” (Pliny V.XI.63)

Strabo and Pliny also note the variations in measurement of the lake particularly during the inundation season in the summer when the ile flooded. rom their accounts it is clear that the lake was thus much more extensive in antiquity than today oyotte & Charvet 1997: 104, n. 179 Strabo 17.1.14). Last but not least among the classical works concerning the description of the Mareotic region, is the Geography of Claudius Ptolemaios in the 2nd century AD. In his fourth book (Chapter 5 about Egypt) Ptolemy begins his description of the geography of Egypt from west to east stating the longitude and latitude of each given site. After covering the 4

M. ABD-EL-GHANI: ANCIENT SOURCES Bibliography

The theft of two camels in this document also leads us to comment on the repeated reference to camels in the Mareotic documents. This is not unexpected from an area situated on the edge of the Western Desert where the camel was the chief means of traffic and transport. In one of the documents relating to business affairs, a sum of 900 drachmae is reported as a rent for some camels (P. Haun. II.22, ll. 9-10 and note on l.9.). In another document dated AD 289, two Mareotes sell an Arabian camel belonging to them to one of the chief officers primpilaris of the Prefect of Egypt for 16½ talents of silver (BGU I.13, II. 1-6).

Papyri BGU = Aegyptische U rkunden aus K öniglichen ( later Staatlichen) Museen zu Berlin, G riechische U rkunden (13). H. Satzinger (ed.), 1904. Berlin. P. Abinn. = The Abinnaeus Archive: Papers of a Roman Officer in the reign of Constantius II (1-9). H.I. Bell, V. Martin, E.G. Turner & D. Van Berchem eds. , 1962. Oxford. P. Fay. = Fayum Towns and their Papyri (134). B.P. Grenfell, A.S. Hunt, D.G. Hogarth & .G. Milne eds. , 1900. London. P. Flor. = Papiri greco-egizii, Papiri Fiorentini III. Documenti e testi letterari dell‘ età romana e bizantina (III, 338). G. Vitelli (ed.), 1915. Milan. P. Fouad = Les Papyrus Fouad I (11). A. Bataille (ed.), 1939. Cairo. P. Hamb. = G riechische Papyrusurkunden der Hamburger staats- und U niversität sbibliothek (I, 39). P.M. Meyer (ed.), 1924. Leipzig. P. Haun. = Papyri G raecae Haunienses (II, 53). Letters and mummy labels from Roman Egypt. A. B low- acobsen. transl. & ed. , 1981. Bonn. P. Iand. = Papyri Iandanae IV, 53 . K. Kalbfleisch, E.H. Schaefer, G. Rosenberger, D. Curschmann, L. Eisner, . Sprey, G. Spiess, L. Spohr & . Hummel eds. , 1912. Leipzig. P. Lond. = G reek Papyri in the British Museum (II, 231). F. Kenyon & H. Bell eds. , 1973. Milan. P. Oslo = Papyri O sloenses III, 84 . S. Eitrem & L. Amundsen (eds.), 1925. Oslo. P. Ryl. = Catalogue of the Greek and Latin Papyri in the John Rylands Library. Manchester (II, 98, 227). C.H. Roberts & E.G. Turner eds. , 1952. Manchester. S.B. = Sammelbuch griechischer U rkunden aus Aegypten V, 8444 . . Bilabel & E. Kiessling eds. , 1952. Wiesbaden.

Some of the Mareotic products are also mentioned in the few available documents. In spite of the excellence and distinction of the Mareotic vineyards and wine, there appears to be limited documentary sources that comment on their merits. A single document, a private letter dating to the 2nd century AD, in which the sender tells his friend and addressee that he dispatched him a variety of gifts, among which we find: ... and a number of bunches of grapes, - 6 small clusters: 2 of them white, 2 of the ‘royal’ variety, the Mareotic as it is called, and likewise 2 smoke coloured, - in a small basket through Harklides” (P. Fouad, 77, ll. 15-20, l. 17). In other documents dating to the 3rd century AD, there are occasional references to “strips of linen”, ταινίδιον: some of them white, some “Mareotic” (P. Ryl. II. 227, ll. 25-33). If we notice that the use of ταινίδιον is a diminutive of ταινεία which means “a headband worn especially as a sign of victory and if we bear in mind that a strip of land near Lake Mareotis was named ταινεία, mentioned in Athenaeus (I.33e) and C. Ptolemaios (4.5.14), could one infer some connection between this place and the production of “linen headbands” whence “it might have derived its name? There is another mention of “ a glassware ... of a fine Mareotic quality” from the 4th century AD (P. Fay. 134, ll. 4, 6). Another 4th century document makes a clear reference to the illegal smuggling of natron by Mareotes or other Egyptians across the desert. In this correspondence between two high officials in charge of the natron monopoly (AD 342-51), strict orders are issued to detain the camel drivers and their beasts, whether in Arsinoe or elsewhere, in order to safeguard the natron of the Fiscus (P. Abinn. I.9 = P. Lond. II.231).

O ther Sources Athenaeus, Deipnosophists, C. Burton (transl.), 1927. London. Ball, ., 1942, Egypt in the Classical Geographers. Cairo Diodorus Siculus, Diodorus Siculus. C.H. Oldfather (transl.), 1935. Cambridge, MA. Herodotus, The Histories. A. de Sélincourt (transl.), 1971. Middlesex. Pliny the Elder, Natural Histories. . Bostock & H.T. Riley (eds.), Perseus Digital Library: http://old.perseus.tufts.edu/cgi-bin/ptext?doc= Perseus:text:1999.02.0138:toc accessed anuary, 2010 . Ptolemaios (Claudius Ptolemy), The G eography. E.L. Ste venson transl. & ed. , 1932. ew ork. Strabo, G eography. H.L. ones transl. , 2001. London. Thucydides, The Peloponnesian War. .M. Dent transl. , 1910. ew ork, London. oyotte, ., & Charvet, P., 1997, Strabon, Le voyage en Egypt. Paris.

Finally, there are some references to army troops stationed in the Mareotic region, together with the payments and extra payments to them, as well as the divisions of the army to which they were assigned (P. Iand. IV. 53 (AD 88), col. 3 P. Hamb. I.39 AD 179 63 BB . Although the image of the Mareotic region, especially in the papyri, is somewhat fragmentary, mosaic-like, and far from complete, it does shed, however, some light on the region. It is clear that the material evidence of future archaeological excavations will add a great deal to our knowledge of this region.

5

LAKE MAREOTIS: RECONSTRUCTING THE PAST

6

M. EL ABBADI: LAKE MAREOTIS IN BYZANTINE TIMES

A Note on Lake Mareotis in Byzantine Times Mostafa El Abbadi

This paper will present readings of selected passages in the ancient sources that have a bearing on Lake Mareotis and its region. As will be seen, the name Mareotis was often exchanged for Mareia, the name of its capital city and harbour, which in Pharaonic times served also as a garrison city protecting the north-western approaches to Egypt, as briefly stated in Herodotus II.18 & 30 and Thucydides I.104.1 .

Emperor ustinian built many constructions and in particular the residences of the magistrates and baths.”

Of special interest is his use of the Greek form Taphosiris, and not the Latin corruption Taposiris, which was used by Strabo. Also the construction work of ustinian reveals the continued importance of the town more than five centuries after the time of Strabo, despite the total change of religion from paganism to Christianity.

The first detailed description of the Lake Mareotis region is found in Strabo 17.1.71 in his well known description of Alexandria where reference is also made to the southern region of Lake Mareotis. In a further significant passage, Strabo 17.1.15 comments on an important aspect of the economic value of the lake:

As regards the southern confines of Alexandria, Procopius reports VI.1. 1-5 : The ile River does not flow all the way to Alexandria, but after flowing to the town which is named Chaereou, it then turns to the left, leaving aside the confines of Alexandria. Consequently the men of former times, in order that the city might not be entirely cut off from the river, dug a very deep canal from Chaereou and thus by means of a short branch made the river accessible to it. There also, as it chances, are the mouth of certain streams flowing in from Lake Maria.

The byblus i.e. papyrus plant grows in the Egyptian marshes and lakes, as also the Egyptian cyamus believed to be the lotus plant Germer 1985: 39 ff. from which was made the ciborium a kind of vessel for drinking it has stalks approximately equal in height, about 10 feet. But though the byblus is a bare stalk with a tuft on top, the cyamus produces leaves and flowers in many parts, and also a fruit like our cyamus, differing only in size and taste. Accordingly the cyamus-fields afford a pleasing sight, and also enjoyment to those who wish to hold feasts therein. They hold feasts in cabin-boats, in which they enter the thick of the cyami and the shade of the leaves for the leaves are so very large that they are used both for drinking-cups and for bowls, for these even have a kind of concavity suited to this purpose and in fact Alexandria is full of these in the work-shops, where they are used as vessels and the farms have also this as one source of their revenues, I mean the revenue from the leaves.”

In this canal, it is by no means possible for large vessels to sail, so at Chaereou they transfer the Egyptian grain to boats lemboi which they are wont to call diaremata, and thus convey it to the city, which they are enabled to reach by way of the canal-route and they deposit it in the quarter of the city, which the Alexandrians call Phialè . But since it often came about that the grain was destroyed in that place by the people rising in sedition, the Emperor ustinian surrounded this district with a wall … .”

In this passage, the following points are of special interest: a The freshwater canal from Chaereou received additional supplies from certain streams flowing in from Lake Maria”. It is worthy of note that Procopius states that the freshwater canal to Alexandria branches off at a location he called Chaereou, whereas Strabo uses the name of Schedia. Surprisingly, other sources of the 6th century and later, use the name of Chaereou only. Significantly we can mention the Mosaic Map of Madaba which is also believed to date from the 6th century Avi- onah 1954: 16-18 . Another source is Stephanus Byzantius 677:18 also in the 6th century, who speaks of Chaereou as a city in Egypt. To all appearances, Schedia and Chaereou were two different locations close to one another, Schedia north of the canal and located on the Canopic Branch, whereas Chaereou is situated immediately south of the canal Haas 1997: 25, top of map 2 and seems to have gained in importance at the time of ustinian when Stephanus called it a polis.

Furthermore, in connection with feasting, Strabo makes mention of Taposiris at the western end of Lake Mareotis and adds 17.1.14 that there they also: hold a great public festival …. and near it there is a rocky place on the sea where likewise crowds of people in the prime of life assemble during every season of the year.” With this classical description in mind, we move on to some Byzantine sources in order to identify any similarities or differences in the descriptions attributed to the lake in later periods. An interesting passage is to be found in Procopius work on the Buildings of ustinian in the 6th century VI.1.12 . It deals with the regions to the west and south of Alexandria. He states that to the west: In that territory, is a city one day s journey distant from Alexandria, Taphosiris by name, where they say the god of the Egyptians, Osiris, was buried. In this city the

7

LAKE MAREOTIS: REC

STR CTI G THE PAST

b In that canal the large ile vessels could not sail so their cargo of grain had to be transferred to diaremata light transport boats . This word was known only in Egypt hence Procopius explained it by the Greek word lembos”. The same word with the change of a vowel dieremata” has been found in papyri cf. ertel 1917: 130 . These dieremata were probably also used to navigate across Lake Mareotis.

The above discussed passages provide evidence that with the founding of Alexandria, Lake Mareotis acquired a lively and vital significance that continued into late antiquity. With its variety of flowering plants, the lake provided a pleasurable recreation resort for the Alexandrians who were, according to Strabo, in the habit of holding feasts in cabin boats. Even as far west as Taposiris, Strabo adds, a great public festival was held, obviously in connection with the cult of siris. More than five centuries later, Procopius reports that ustinian showed interest in that town Taphosiris by adding new construction work which is an indication of its continued importance.

c The grain was eventually carried by light transport boats and deposited in the quarter of Alexandria called Phialè . This Phialè is not attested anywhere else and we have the impression that it was close to a landing-place near the canal. However, Procopius speaks of it as already existing and how it used to be plundered in times of trouble hence ustinian surrounded it with a wall for protection.

It is of interest to know that the freshwater canal that links to Alexandria received additional supplies from certain streams flowing in from Lake Maria . This passing mention by Procopius provides additional evidence regarding the intricate hydraulic system of the freshwater canal. Furthermore, this canal provided the final facility for transferring grain from large Nile vessels – which could not sail in the canal - to light transport boats dieremata .

inally, we come to a text by Sophronius early 7th century in his account of the Miracles of Saints Cyrus and J ohannes. The miracle happened to a certain Christodorus, deacon of the church of the two saints at the time of ohn the Almsgiver, then Bishop of Alexandria 610-619 , whom Sophronius describes as the beloved of the poor . In his account, Sophronius has this to say miracle no.8 :

Finally, the lake and the regions surrounding it were of vital economic value to Alexandria, since the waters of the lake teemed with fish and water-fowl and the surrounding fertile lands flourished with large plantations until the early 7th century. Accordingly, Alexandria depended upon this neighbouring hinterland for the regular supply of its daily needs of meat and vegetables.

Lake Mareia: ou Alexandrians beloved of Christ, you all without exception know it well i.e. the miracle , so also do a good number of Egyptians and most of the Libyans and as well as you, who are curious to know about Alexandria. Christodorus once sailed across Mareotis with the intention of inspecting the domains belonging to the church of the saints. A storm suddenly blew violently the time was winter , agitating the waters of the lake and raising great waves. Christodorus was in great danger and thought he was going to perish. Nevertheless, he prayed to the saints who listened to his prayer … . And the wind at once dropped and he was saved.”

Bibliography Avi- ona, M., 1954, The Madaba Mosaic Map. erusalem. Calderini, A., 1935-1987, Dizionario. Cairo, Madrid. C.Theod. = Codex Theodosianus, T. Mommsen & P.M. Meyer eds. , 1904-5, 2 Vols. Berlin. Donner, H., 1992, The Mosaic Map of Madaba. Kampen. http://198.62.75.1/www1/ofm/mad/legends/legends156.html. Accessed ctober 2009. Courtesy Prof. M. Haggag . Gascou, . transl. & ed. , 2006, Sophrone de J érusalem, Miracles de Saints Cyr et J ean. Paris. Germer, R., 1985, Flora des Pharaoischen Aegypten, DAI. Sonderschrift 14. Kairo with thanks to Prof. Mirvat Seif . Haas, C., 1997, Alexandria in Late Antiquity. Baltimore, London. Herodotus, Historiae, C. Hude transl. , 1966. xford. Leontius eapoleos, Life of ohn the Almsgiver. In E. Dawes & .R. Baynes transl. & eds. , Three Byzantine Saints: Contemporary Biographies of St. Daniel the Stylite, St. Theodore of Sykeon, and St. J ohn the Almsgiver 1948. xford. ertel, ., 1917, Liturgie. Leipzig. Palladius, The Lausiac History. C. Butler transl. , 1904. Cambridge. Palladius, Historia Lausiaca. Migne, Patrologia Graecia 34. 1860, reprint 1998. Gottingen.

For our purpose however, the main point of interest in this passage is the mention of the domains of the church . In the Byzantine period, just as important temples of earlier times disposed of great wealth and property, so now the Church of Alexandria possessed considerable property. In 415 Emperor Theodosius, in a law on patronage, guaranteed the property of the Church of Alexandria which had been acquired before 397 C.Theod. I.23.6 . The material wealth of the Church of Alexandria is also asserted by the fact that it owned a merchant fleet in the Mediterranean. Early in the 7th century the same ohn the Almsgiver, used that fleet to import corn from Sicily during a famine at Alexandria Leontius eapoleos, 13 . However, throughout antiquity, the lake s waters teemed with fish and water-fowl P.Tebt. 3.867 Leontius eapoleos, 8 Rufinus, 27.10 . Alexandria depended on its neighbourhood for the supply of its daily needs of meat and vegetables. We may also recall that there were some 2,000 monks near Alexandria at Nitria, south of Lake Mareotis and that however frugal their diet might have been, they were amply sustained by products of the region Palladius VII, 1-2 Sozomen, VI, 29 . 8

M. EL ABBADI: LAKE MAREOTIS IN BYZANTINE TIMES Procopius, Buildings. H.B. Dewing transl. , 1961. London. P.Tebt. = Tebtunis Papyri, vol. III.2, A.S. Hunt & .G. Smyly transl. , 1933. London Rufinus, Historia Monachorum. Migne, Patrologia Latina 21, 1849. Gottingen. Sophrone de rusalem, Miracles de Saints Cyr et J ean, . Gascou transl. , 2006. Paris.

Sozomen, Historia Ecclesiastica. Migne, Patrologia Graecia 67. 1864, reprint 2002. Gottingen. Stephanus Byzantius, O pera, 4 vols., K.W. Dindorf ed. , 1825. Leipzig. Strabo, G eography, H.L. ones transl. , 1917. London. Thucydides, Historiae, H. Stuart ones transl. , 1897. Oxford.

9

LAKE MAREOTIS: REC

STR CTI G THE PAST

10

I. AWAD: MARYUT LAKE THROUGH MAPS

A Study of the Evolution of the Maryut Lake through Maps Ismaeel Awad

This study concerns the evolution of the shape of Maryut Lake and its connections with the Nile River and the Mediterranean Sea from antiquity until the modern day by studying and observing maps of the Maryut Lake region, classified chronologically from the oldest to the most recent.

The aim of this study is to use maps to measure the evolution of the Maryut Lake, to test the value of using the map as a tool in the study, to evaluate the type of results obtained, and lastly, to try to forecast the shape and state of the Maryut Lake in the future. The Maryut Lake region consists geographically of five different parts (Fig. 1). In the north, a sandy strip extends along the Mediterranean coastline. To the south there is a calcareous ridge known as Abusir-Mexx that has an average altitude of 30 m high. The center of the Maryut region is divided into two sub-divisions, the first is the Maryut Basin which is situated towards the east with an average altitude attaining sea level; the second sub-division is the Maryut Valley situated towards the west of the region. And finally, the Maryut Ridge lies towards the south-west and has an average altitude of 50 m high.

In the following pages, two different types of available maps are considered. The first type of map represents edited illustrations of Maryut Lake by different historians and ancient cartographers who built up their illustrations through visual observation or by drawing. The study of these illustrations using the superimposition method, essentially overlapping one map on another, cannot be applied, as any attempt to do so would result in the production of inaccurate results based on the fact that these illustrations display approximate distances and inexact locations. The second type of map utilised topographic maps, which when edited using accurate topographic methods, can produce more accurate results. The application of topographic methods began during the French campaign at the end of the 18th century in which levelling and goniograph tools are used to construct maps in Egypt. Topographic methods have developed over time through to the current use of satellite images in mapping. The data extracted from these maps are accurate enough to be simply compared and overlapped.

About 20 maps and illustrations are used in this study, of which the oldest map was edited at the end of the 16th century and the most recent was extracted from a Landsat satellite image produced in 2001.There are two different ways of studying these different types of maps, as will be revealed during the course of this paper. Fig. 1: Maryut Lake region. Digitised by I. Awad. Archives CEAlex.

11

LAKE MAREOTIS: RECONSTRUCTING THE PAST Illustrations The illustrations utilised in this study date from the end of the 16th century until the end of the 18th century. Some of them were constructed ‘indoors’ without ever stepping foot in Egypt. They were done by using other maps edited by other publishers, used as base maps for adding new information. or this reason it is difficult to use these illustrations to derive any form of accurate reference or in application to an overlapping study.

the evolution of names and locations of settlements such as towns or villages existing all around the Mareotic region. Illustrations edited before the end of the 16th century used to ignore the Maryut Lake as it was of little importance at that time, certainly when compared with the town of Alexandria. Ancient map makers rarely mentioned Maryut Lake, locating it at the margin of the map and illustrating its location in small letters. They portrayed an ignorance of the lake’s shape, such as for example the illustration edited in 1548 by Pierre Belon du Mans (Fig. 2).

However, some useful information could be extracted from examining these illustrations, such as: the evolution of the connections between Maryut Lake and the Nile River, Nile branches and the Mediterranean Sea,

The illustrations used in this study are classified chronologically from the oldest to the most recent: Fig. 2: 1548, Vray portraict de la ville d’Alexandrie en Egypte, Pierre Belon du Mans. J ondet G ., 1921, pl. II. Archives CEAlex.

Fig. 3: 1570, de tail from Aegyptus antiqua, including a focused map of the region of Alexandria, Abraham O rtelius. J ondet G ., 1921. Archives CEAlex.

12

I. AWAD: MARYUT LAKE THROUGH MAPS 1.01: 1570, 1 Abraham O rtelius (Fig. 3)2 It can be observed from the general map that the Maria, et Mareotis lacus qui et Ara potes (Maryut Lake’s) shape took a direction from north-east to south-west incorporating eight islands as shown in the figure.

River), which was also connected to the Nile River at its south-eastern end. The boats depicted in the Moeris Swamp show that there was extensive navigation during this period between Moeris Swamp, Maryut Lake and the Nile River.

Maryut Lake had two connections, the first was the connection with the Myris quod et Moeris stagnum (Moeris Swamp) to the south through the Lycus u ius (Lycus

The second connection related directly to the Nile River in three places:3 the first one to the north related to the Canopica fossa4 (Canopic Branch), the second connected the lake with the Canopic Branch at the north-east near the town of Schedia, and finally, the third connected the lake directly with the Nile River at its eastern side. 1.02: 1588, L ivio Sanuto (Fig. 4)5 In this map, the Bycheira lacus (Maryut Lake) has deformed in its shape forming a triangle enclosing a large number of islands (21 islands). Just one connection linked the lake to the Canopic Branch at its eastern end while no connection was illustrated with the Moeri dis lacus (Moeris Lake). 1.03: 1655, N icolas Sanson D’Abbeville (Fig. 5)6 The map of D’Abbeville has similar limitations to the previous map. The shape of Maryut Lake is similar although more rectangular and adopting more of a northwest to southeast orientation. However, most of the islands that had been previously observed have disappeared. 1.04: 1717, Paul Lucas (Fig. 6)7 No connections are shown with the Lac de la Mareote (Maryut Lake) in this map while some islands reappeared (nine in total). The Moeris Lake has totally disappeared, and at the same time the Lac du Madie (Madieh Lake/Abuqir Lake is illustrated for the first time.

Fig. 4: 1588, detail from Africae Tabula VII, Livio Sanuto. Archives Biblioteca Nazionale Centrale Di Roma.

1.05: 1722, Claude Sicard (Fig. 7)8 In Sicard’s illustration the Mareotis Palus (Maryut Swamp’s) form has changed again to resemble the shape of a bird or a duck. Some connections are apparent between Maryut Lake and the Nile to the east that joined together before flowing into Maryut Swamp. A small lake named Nitraria Palus9 itraria Swamp was identified to

Fig. 5: 1655, detail from Royaume et désert de Barca et l’Egypte, Nicolas Sanson D’Abbeville. Commercial reproduction.

1. There is another illustration of Ortelius dated from 1595, found in a late commercial reproduction from the beginning of the 21th century. 2. Aegyptus antiqua, map of the Antique Egypt. 3. There is a fourth connection shown in the focused map of Alexandria and not in the general map. It is situated between the first and the second connections at the north passing near the village of Chabrie pagus. 4. Mentioned in the focused map of Alexandria. 5. Africae Tabula VII, map of Africa, sheet no. 7. 6. Royaume et désert de Barca et l’Egypte, Kingdom of the desert of Barca and Egypt. 7. Carte de la Basse Egypte et du cours du Nil, map of Lower Egypt and the Nile River. 8. Carte de l’Egypte Ancienne, map of Ancient Egypt. The original map edited by Claude Sicard was not found. This is a synthesis of two copies extracted from the original map done by D’Anville and Danlisle and the place name’s list of Sicard. Sauneron & Martin 1982. 9. There are two lakes named Nitraria in the map: one near the Maryut Lake which is mentioned in the study and the other located in Wadi Natrun between Alexandria and Cairo.

13

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig. 6: 1717, detail from Carte de la Basse Egypte et du cours du Nil, Paul Lucas. Archives IFAO .

Fig. 7: 1722, detail from Carte de l’Egypte Ancienne, Claude Sicard. Sauneron & Martin 1982, Archives IFAO .

Fig. 8: 1743, detail from Carte d’Egypte, Richard Pococke. Archives IFAO .

Fig. 9: 1753, de tail from Carte de l’Egypte ancienne et moderne, Robert de Vaugondy. Archives J .-Y . Empereur.

the south-east of Maryut Lake, connected with it through Acheron Canalis (Acheron Canal). The town of Marea is illustrated, but strangely to the north-west of Maryut Swamp and not on the southern shores of the lake as located nowadays.

that used to connect Moeris Lake with Maryut Lake. The towns of Plinthine.S. and Taposiris.P.11 were indicated but very far from the lake towards the west. 1.07: ca 1753, R obert de Vaugondy (Fig. 9)12 This 18th century map indicates yet a further change to the shape of Birk Mariou (Maryut Lake) now depicted as being aligned from north-east to south-west. There is still

1.06: 1743, R ichard Pococke (Fig. 8)10 The shape of Birk Mariou.s./ Lacus Mareia.f./ Mariotis.S. (Maryut Lake) changed again in this map taking a triangular form. A lake is mentioned at the place of the Nitraria Lake but located further south than before, and the Moeris Lake appeared as Birk El K aroon/ Lacus Moeris feu Moerius (Karoun Lake).

10. Carte d’Egypte, map of Egypt. 11. Known as ancient navigation ports situated actually at the western arm of the Maryut Lake. 12. Carte de l’Egypte ancienne et moderne, map of Ancient and Modern Egypt. The base map is inspired from Claude Sicard’s map.

Bahr Bellomah i.e Mare seu Lacus Sine Aqua (lake or sea without water) was mentioned at the place of the old canal 14

I. AWAD: MARYUT LAKE THROUGH MAPS one easterly connection between Maryut Lake and the Nile River. A lake is mentioned and located at the same place of the Nitraria Lake as in the previous map. 1.08: 1762, Rigobert Bonne13 The Bonne map has many similarities with the previous map, with just one change which is the appearance of the town of Abusir located at the middle of the northern shoreline of the Maryut Lake. 1.09: 1764, J acques Nicolas Bellin (Fig. 10)14 A connection between Lac Mareotis (Maryut Lake) and the Mediterranean Sea is shown for the first time to the west of Alexandria, with the disappearance of Lake Nitraria. Two connections appeared between Maryut Lake and the Nile (both meet at the same point, southeast of the lake). One of them approached from the east and the other from the Bahr-el-Helame ou Fleuve seché (canal or sea without water) from the south-east, and passed through a Lac de Sel (Salty Lake)15. 1.10: 1799, Edme Mentelle & Pierre-G regoire Chanlaire16 In this map of 1799 Sebaga ou Birk Marioul17 (Maryut Lake) reappeared in the form of a triangle with what appears to be a sea connection to the west of Alexandria. The connection with the Nile River appeared through the Canal de Bahiré Bahirey Canal flowing into the Maryut Lake, and Lac de Natron (Natron Lake) appeared at the place of the ex. Nitraria Lake southeast of Maryut Lake. The Bahhr-bé-lâ- mâ ou Fleuve sans eau (canal without water) was also shown approaching from the Birket-il K erun (Karoun Lake) which is located to the south in the Feï um region, extending towards the southwest of Maryut Lake but without any intervening connections.

Fig. 10: 1764, de tail from Carte des Embouchures du Nil, et partie de son cours, J acques Nicolas Bellin. Archives J .-Y . Empereur. geo-referencing the maps in the same coordinate system, digitising each map in AutoCAD program, overlapping each digitised map one on the other, analysing the digitised features in ArcGIS program.

Towns are named around the Maryut Lake such as the town of Mariout (Marea) located on the northern shoreline of the lake and Tour des Arabes (town of Abusir) near to the Mediterranean Sea.

2.01: 1801, French campaign, scale: 1:100.000 (Figs. 11 & 12)18 This is the first accurate topographic map concerning the Mareotic region. It was edited by the engineers of the French campaign who came to Egypt between 1798 and 1801. The map was constructed by accurate topographic instruments including a level and a goniograph.

Topographic Maps In order to extract precise locations and establish an accurate surface area, four steps were undertaken when studying the second type of map:

In this map, a cut is shown in the dike separating the Bohéï réh Marî oût ou lac Mareotis (Maryut Lake) from the Bohéï réh Maadî éh, lac Madié ou d’Aboûqî r (Madieh or Abuqir Lake). The dike was breached by the English army in April 180119 during the war between the French

13. Carte de l’Egypte ancienne et moderne, map of Ancient and Modern Egypt. 14. Carte des Embouchures du Nil, et partie de son cours, map of the Nile river mouths, and a part of its stream. 15. The Salty Lake is at the place of the Nitraria Lake situated in Wadi Natrun. 16. Focused map Carte particuliè re et détaillée du Delta (Detailed map of the Delta) in the Carte physique et politique de l’Egypte (Physical and political map of Egypt). There is another similar map Carte d’Egypte (map of Egypt) edited by Jean-Baptiste Bourguignon D’Anville in 1765. Charles-Roux 1910. 17. Sebaga is near to Sabkha in Arabic language which means marsh, and Birk comes from Birka in Arabic, too, which means pond.

18. Carte d’Alexandrie, sheet no. 37 extracted from the Atlas of the French Expedition Carte topographique de l’Egypte et de plusieurs parties des pays limitrophes. 19. The dike was breached 19 April, 1801, by the English army under the orders of General Sir John Hely-Hutchinson, in order to cut off the freshwater supply brought by the Alexandria canal from the French garrison located in Alexandria, and to cut the road relating them to the main French army situated in Cairo.

15

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig. 11: 1801, Carte d’Alexandrie, French campaign. Archives J .-Y . Empereur. Fig. 12: 1801, Carte d’Alexandrie, French campaign, original scale 1:100,000. Digitised by I. Awad. Archives CEAlex.

and English. As a result, the water in Abuqir Lake (which was situated at a higher altitude than the Maryut Lake and was directly connected with the Mediterranean Sea), was redirected into Maryut Lake, causing flooding in its eastern (125 km² ) and western (90 km² ) extremities.

The depth of Maryut Lake was measured in units of feet using depth soundings all around the lake. We can observe from the depths shown on the map that the deepest zone in the Lake was the north-eastern zone that was an average depth of about 4.5 m. Maryut Lake becomes shallower

Two lake limits are represented on the map, the first referred to the limit of the flood on 9 August, 1801,20 and the second showed the limit of the lake after the flood had returned to its preceding limit on 10 January, 1802.21

20. The flood limits were obtained from M. Tawiland s map scale: 1:128.000), engineer in the English army. Jacotin 1809-1828: 83. 21. Due to the effect of the northerly wind and to the reduction of water in the Nile River branches. 22. Maybe it was the connection mentioned before in the ancient illustrations between the Mediterranean Sea and the Maryut Lake at the west of Alexandria.

Another dried up old cut22 is illustrated linking Maryut Lake with the Mediterranean Sea to the east of the Anse de Marabou (Gulf of Marabou), west of Alexandria. 16

I. AWAD: MARYUT LAKE THROUGH MAPS

Fig. 13: 1866, Carte des Environs d’Alexandrie, Mahmoud Bey El-Falaki. Archives CEAlex. Fig. 14 : 1866, Carte des Environs d’Alexandrie, Mahmoud Bey El-Falaki, original scale 1:200,000. Digitised by I. Awad. Archives CEAlex.

time before the flood which was about 890 km 211.300 feddans). At this point, the lake’s maximum width (from north to south) was about 45 km, and its maximum length (from east to west) reached about 56 km.

from the north-eastern zone towards the south (2 m in average) and the west (1.5 m in average) until the lake attends 0 m at its limits. In addition, two levelling profiles23 were undertaken across Maryut Lake, one at the Maryut Valley24 and the other along the dried old cut mentioned previously.

23. Le Pè re 1809-1828: Pl. 19. 24. Between the ruins of the town of Marea and the Mediterranean Sea passing through the saint tomb of Aboû e l K eï r.

Thanks to digitising, we are able to ascertain with sufficient accuracy the precise surface area of the lake at that 17

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig. Atlas of Egypt, Vol. I, Lower Egypt, Survey of Egypt (G. Daressy), original scale 1:50,000. Digitised by I. Awad. Archives CEAlex. Fig. 16: 1949 , map series 1:100.000. Survey of Egypt, original scale 1:100,000. Digitised by I. Awad. Archives CEAlex.

2.02: 1866, Mahmoud Bey El-Falaki, scale: 1:200.000 (Figs. 13 & 14)25 This map is interesting because El-Falaki (1872) mentioned that Lac Maréotis (Maryut Lake’s) shape at that time (1866) was similar to the lake’s shape when the town of Alexandria was founded in antiquity, according to Strabo’s notes.26 The author also defined the ancient branches of the Nile as the Canopic Branch and Schedia Canal.

Lake was also reduced in area to 450 km² (which is about 50% of its area as indicated in 1801). This reduction in area occurred in just 65 years, and was especially apparent in the eastern part where its maximum length was about 24 km.

25. Carte des Environs d’Alexandrie, map of the surroundings of Alexandria 26. El-Falaki 1872: 99.

From this map, it can be observed that the Lac d’Aboukir (Abuqir Lake) has reduced in area. Proportionally Maryut 18

I. AWAD: MARYUT LAKE THROUGH MAPS

Fig. 17: 1970s, map series 1:25.000, E gyptian G eneral Survey Authority, original scale 1:25,000. D igitised by I. Awad. Archives CEAlex. 2.03: 1911, Survey of Egypt, scale: 1:50.000 (Fig. 15) 27 The area of Maryut Lake continued to reduce particularly on its eastern margins. The Egyptian government began to have some interest in this zone by managing a system of irrigation canals and drains (for example: El-Omum Drain, whose waters flowed from the south-eastern side of the Maryut Lake) in order to cultivate the region around Maryut Lake. The area of the lake was reduced to 340 km² (38% of 1801s area). However, the dimensions of the lake from east to west appear to have extended compared to previously (about 75 km). The ancient cut between the Mediterranean Sea and Maryut Lake to the west of Alexandria is again mentioned.

north-eastern shores, particularly with the construction of Nozha30 and the seaplane base. By now most of the western part of the lake had dried up, perhaps partly due to the fact that this map was surveyed in the dry season.31

2.04: 1930, Survey of Egypt, scale: 1:100.000 28 The area of Maryut Lake continued to reduce in size particularly towards the south-east as the lake became increasingly exploited for agriculture purposes. An artificial salt marsh appeared for the first time in the middle of the lake in order to produce salt.

A great deal of development had also occurred in the lake:

2.06: 1970s, Egyptian G eneral Survey Authority, scale: 1:25.000 (Fig. 17) 32 By the 1970s a big development had occurred at the eastern and southern parts of Maryut Lake with an expansion of irrigation canals and drains. The Nubareya Canal was completed in 1976 channeling waste water runoff from agriculture lands into the lake, and then into the Mediterranean Sea via the Mexx pump station.

In the middle of the lake, the salt marshes became much larger (it has expanded ten fold in just 20 years). Big changes also took place in the north-eastern part of the lake. The seaplane base was transformed into a big fish farm, and Nozha airport became a part of the land in order to separate the new fish farm from Maryut Lake.

2.05: 1949, S urvey of Egypt, scale: 1:100.000 (Fig. 16) 29 The lake was further reduced on its eastern shores. This map also indicates early interest in expansion along its

Maryut Lake’s surface area has reduced further to just 210 km² (51.000 feddans) (which is about 1/4 of its area in 1801).

27. The maps of this series have been collected from: Survey of Egypt 1914. Eight maps have been used in digitising: sheet no. 009 Ramleh, sheet no. 010 Abuqir, sheet no. 020 Mallaha, sheet no. 021 Alexandrie, sheet no. 022 K afr El Dawar, sheet no. 046 Hammam, sheet no. 047 Bahig and sheet no. 048 K ing Mariout. 28. German edition, four maps have been used in digitising: 88/42 Hammam, 88/48 G hayata, 92/48 Alexandria and 92/54 Damanhur. 29. Arabic edition, four maps have been used in digitising: 0° 31’/29° 40’ Abu El-Matamir, 31° /29° K rair, 31° /29° 40’ Alexandria and 31° /30° 20’ Damanhur.

. 30. Known at that time as Fouad the First airport: 31. The western zone s water level is directly influenced by dry and wet seasons. 32. Fourteen maps have been used in digitising: 90/450 Hammam, 90/465 Mahatet El-G harbaneyat, 91/450 K om Makhoura, 91/465 Burg El-Arab, 92/465 Burg El-Arab, 91/480 Mahatet Bahig, 91/495 Aeroport of Amreya, 92/480 K reir, 92/495 Amreya, 93/480 El-Deir, 93/495 Mexx, 93/510 K om Echo, 94/510 Ramleh and 94/525 Munshaat Buline.

19

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig.18 : 1991, map series 1:50.000, E gyptian G eneral Survey Authority. Digitised by I. Awad. Archives CEAlex. Fig. 19: 2001, Landsat satellite image, resolution 15m / pxl. Digitised by I. Awad. Archives CEAlex.

2.07: 1991, Egyptian G eneral Survey Authority, scale: 1:50.0 00 (Fig. 18) 33 A huge development occurred in the southern and eastern parts of the Mareotic region in the early 1990s which enabled the entire lake to be part of the same irrigation system, and large areas of the middle zone of the lake were transformed into fish farms.

2.08: 2001, LANDSAT satellite image, resolution: 15m / pixel (Fig. 19) 34 Satellite images were used to study more recent changes in the lake morphology due to the unavailability of more

These were the last map series produced concerning the Mareotic region, and edited by the Egyptian General Survey Authority.

33. Six maps have been used in digitising: NH35-L5d Hammam, NH35L6c Burj al-’Arab, NH35-L6d K ing Mariout, NH35-P3a Sidi K irayr, NH35-P3b Alexandria and NH36-M1a K afr El-Dawar. 34. Downloaded from the University of Maryland site (www.umd.edu).

20

I. AWAD: MARYUT LAKE THROUGH MAPS Table 1: Comparison table between different periods of Maryut Lake.

Fig. 20: Evolution diagram of Maryut Lake’s surface areas .

contemporary maps since the last map series was produced in the 1990s.

as the construction of fish farms and artificial salt marshes, all of which require stable quantities of water.

A topographic map was extracted from a LANDSAT satellite image of a medium resolution of 15 m/pixel, which is enough to digitise the limits of the lake, marshes, fish farms, wide canals and drains, but not enough to define the other small features such as buildings, narrow canals, secondary drains or roads.

n the other hand, natural influences to the shape of Maryut Lake have a limited impact nowadays (such as rain and ground water). However, any loss in the lake’s surface area in the future depends on the rate and types of uses of the lake and the implemented management plan. Forecasting the Shape of Maryut Lake Digitising the topographic maps does provide an opportunity to forecast the future shape of the Maryut Lake region, if left to evolve naturally without any external interference. A forecasting study is undertaken by extracting and digitising different contour levels of the Maryut Lake region. The maps utilised in this digitising exercise were the map series of scale 1:25,000 edited in 1940s, and the same map series edited in 1970s; these maps series were chosen because they are the most accurately contoured maps that covered the whole Mareotic region. The accuracy of the results of such an exercise is after all dependent upon the precision of the maps used, and on the data extracted.

The map showed that the area of Maryut Lake’s is still about 210 km² (51,000 feddans). Its maximum dimensions attended 66 km in length and 11 km in width. From the table (Table 1) and the diagram (Fig. 20) showing the evolution in surface areas of Maryut Lake from 1801 until 2001, we can notice that the lake has lost 75% of its surface area in just 200 years. The greatest loss occurred between the period from 1801 to 1866, in which the lake lost half of its area due to drying and land fill. Maryut Lake became more stable with regard to its rate of surface area lose from the 1970s until now. During this 30 year period it lost 0.1% of its surface area. The main contributor to the reduced rate of water loss was irrigation and the increase in activities around the shores of the lake, such

The 0 m contour line was the first digitised, which refers to the limits of the Maryut Basin, which is the same as sea level (0 m). By comparing this contour map with Maryut 21

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig. Comparison bet een Maryut Lake French campaign map series in the Mareotic region. Digitised by I. Awad. Archives CEAlex.

top and m le el bottom

Results (Fig. 22) The surface area of Maryut Lake reduced by less than a quarter (24% ) over a period of 200 years (1801-2001). This does not bode well for the future of Maryut Lake, particularly when you consider the intensive construction projects in and around the lake especially at its eastern end, its northern and at its north-western parts having been continuously divided up, dried out and filled to construct tourist villages and settlements.

Lake s limits in 1801 without flood , it can be seen ig. 21) that they are approximately identical. That means that the lake’s limits in 1801 had reached the 0 m level. Using this kind of study, the Maryut Lake’s limits can be compared to any topographic map with the different digitised contour levels such as the levels of 0 m, -1 m, -2 m and -3 m, in order to study the lake water’s level in any period. We chose these precise levels because the deepest altitude of the Mareotic region is about -4 m which is limited to a small area in the north-eastern part of the lake and the highest level that the lake’s water reaches did not exceed 1 m.

The use of Maryut Lake itself such as fish farms or salt marshes has increased from 1 km² (0.3% of Maryut Lake’s area) in 1930 into 39.6 km² (18.6% of Maryut’s Lake area) 22

I. AWAD: MARYUT LAKE THROUGH MAPS

Fig. 22: Comparison between Maryut Lake 1801-2001. Digitised by I. Awad. Archives CEAlex. Charles-Roux, F., 1910, Les origines de l’Expédition d’Egypte. Paris. Carte d’Alexandrie, 1801, French campaign. Carte topographique de l’Egypte et de plusieurs parties des pays limitrophes [ MAP] . El-Falaki, Mahmoud Bey, 1872, Memoire sur l’antique Alexandrie, ses faubourgs, ses environs. Copenhagen. Jacotin, M., 1809-1828, Mémoire sur la construction de la carte de l’Egypte. Description de l’Egypte, Etat Moderne, Tome 2, 2è me Partie. Paris. Jondet, G., 1921, Atlas historique de la ville et des ports d’Alexandrie. Cairo. Le Pè re, G., 1809-1828, Planche 19, Vallée du Nil et Lac Maréotis. Description de l’Egypte, Antiquités, A. Vol. 5. Paris. Livio Sanuto, 1588, Africae Tabula VII. Biblioteca Nazionale Centrale di Roma [ MAP] . Lucas, P., 1664-1737 (1724), Voyage du Sieur Paul Lucas, fait en MDCCX IV & par ordre de Louis X IV, dans la Turquie, l’Asie, Sourie, Palestine, Haute et Basse Egypte. Paris. Pococke, R., 1745, Observations on the Islands of the Archipelago, Asia Minor, Thrace, Greece and some other parts of Europe. A Description of the East and some other countries. Vol. II. Part II. London. Sauneron, S., & Martin, M., 1982, O euvres III - Parallè le géographique de l’ancienne Egypte et de l’Egypte Moderne. Cairo. Survey of Egypt, 1914, Atlas of Egypt, Vol. I, Lower Egypt, Cairo. de Vaugondy, R., 1753, Carte de l’Egypte ancienne et moderne [ MAP] .

in 2001. That means that the lake’s use has increased about 40 times in just 70 years for salt production. C onclusion The map can be a very useful tool in a study of temporal evolution of any geographic feature, such as the case of Maryut Lake. When assessing the results of such studies, the following conditions must be considered to increase the reliability of the product: Data must be accurate to all the region of study the more accurate the data the better the results). Data maps and satellite images must be accessible. This is a particular problem when investigating the Maryut Lake region where some maps of some of the series were not found due to negligence of archiving or due to security reasons.35 Data must be updated regularly to make the study of the region reliable and more accurate. In the case of the present study, the last series of maps produced concerning the whole region was in 1991. This means that for the last 17 years, no other map series concerning the Mareotic region have been produced, despite continued development and huge observable changes having occurred in the Mareotic region since this time. Bibliography D’Abbeville, N.S., 1655, Royaume et désert de Barca et l’Egypte [ MAP] . Bellin, J.N., 1764, Carte des Embouchures du Nil, et partie de son cours [ MAP] .

35. Several military zones exist around the Mareotic region causing the exclusion of some maps which include these.

23

LAKE MAREOTIS: RECONSTRUCTING THE PAST

24

L. BLUE: LAKE MAREOTIS RESEARCH PROJECT

Lake Mareotis Research Project Lucy Blue

Alexandria was by any standards one of the great cities of the Mediterranean. Since its foundation in 331 BC and for almost a millennium to follow, it was the political, economic and cultural capital of Egypt, and one of the most significant emporia and complex ports in the Hellenistic and Roman worlds. As such, Greco-Roman Alexandria has been the subject of much historical and archaeological research (Fraser 1972; Haas 1997; Empereur 1998; Goddio, et al. 1998) that has revealed the wealth of the city, much of which was generated by trade through its important and monumental harbour. However, the complexity of the Alexandria harbours system, which included not only harbours on the sea, but also on Lake Mareotis, has never been thoroughly understood. The significance of Lake Mareotis in the history of the Greco-Roman port-city of Alexandria, is undeniable (Fig. 1). Settlements and industrial units located along the shores of Lake Mareotis, located to the south and west of the city, are known to have produced amongst other things, glass, textiles, pottery and wine (Empereur & Picon 1998), all of which were transported to the city. Yet the dynamics of this productive hinterland have not been fully appreciated in terms of the important contribution the region made to the economy of Alexandria and thus to Egypt as a whole.

The principal objective of the Lake Mareotis Research Project was to undertake a detailed systematic survey of the western extension of the lake in order to record, quantify and assess the archaeological resources of the area in a comprehensive manner. As a result, we hoped to determine a better understanding of the nature and extent of economic and maritime activities in the Mareotic region and to explore the relationship between the component parts of this complex system and how they developed and changed over time, in order to determine a more comprehensive understanding of the economic functions of ancient Alexandria.

In this context, the Lake Mareotis Research Project set out to investigate and determine the role of the lake, the function of the sites around its shores and the part they played in supporting the metropolis of Alexandria in antiquity.

In ancient times, Lake Mareotis was connected to and fed by the Canopic Branch of the River Nile (Strabo 17.1.7). By the 12th century the lake had silted up but prior to its silting it was much larger and deeper (Warne & Stanley 1993: 53-8; Said 2002). The main body of Lake Mareotis currently covers an area of about 90 km2 and is defined to the north by a chain of lithified carbonate ridges extending parallel to the northern shoreline of the lake and the Mediterranean coast (Fig. 2). It is nearly rectangular in shape and extends south of Alexandria for about 12 km, merging along its southern and eastern shores into the Western Deltaic region. It would once have extended considerably further south, quite possibly a navigable distance of about 50 km. It is separated from the lake’s main body by causeways and shallows and its relative size varies according to the season.

Envi ronmental and Political C ontext of Lake Mareotis Lake Mareotis represents one of the most distinctive geomorphological features along the north-west coast of Egypt (Warne & Stanley 1993). It is unique compared to other lakes on the north coast of Egypt, being the only one with no direct connection to the Mediterranean Sea, but originally with direct access to the River Nile and hence to the whole of Egypt. Therefore, it is believed that the location and characteristics of Lake Mareotis gave the city of Alexandria one of its major advantages (Strabo 17.1.7).

C ontext of the Lake Mareotis Research Project The Lake Mareotis Research Project is a collaboration between the Centre for Maritime Archaeology, at the University of Southampton, and Department of Underwater Antiquities (DUA) of the Supreme Council of Antiquities SCA , Alexandria. The project first started in 2004 when a British Academy Small Grant was awarded to support a pilot survey season in the western arm of the lake, to the west of Alexandria. The survey covered an area of approximately 40 km east-west by 3 km wide, along the shores of the western extension of Lake Mareotis, from Sidi Kerir to El-Hammam. During this first season over 60 sites where identified, many equating to new discoveries (Blue & Ramses 2005). The potential wealth of the sites in the area was thus realised and subsequently further funding was sought and awarded from both the British Academy and the Leverhulme Trust to conduct a more detailed survey on both shores of Lake Mareotis and on an island that is located in the eastern region of the survey area.

Thus at present, Lake Mareotis is a body of shallow brackish water that occupies about 13% of an extensive subsea-level depression that is sometimes referred to as the Maryut Depression. The remaining 87% of the depression has been drained and is now used primarily for agriculture. The main body of the lake is heavily polluted by untreated sewage and industrial waste. It has also been divided into many sub-basins which are used for industrial and agricul25

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig. 1: Location of Alexandria and Lake Mareotis ( Lake Mareotis Research Proj ect) .

26

L. BLUE: LAKE MAREOTIS RESEARCH PROJECT tural purposes. Moreover, the shores of the lake are subject to continuous irrigation, drainage and reclamation, which means that it is constantly changing and unstable (Warne & Stanley 1993: 29-30 rihy, et al. 1996: 282 . The average depth of the remaining lake is less than 1 m, although it is assumed to have been greater in antiquity to allow boats to sail safely.

and Boussac & El Amouri (this volume) on the Hellenistic port of Taposiris Magna. In addition, rescue work has been undertaken by the Supreme Council of Antiquities (SCA) at a number of smaller sites around the shores of the lake largely in advance of development. The majority of known sites have not been systematically studied however, and the pilot survey conducted in 2004 revealed that only a fraction of the available evidence is currently recognised. In addition, as indicated previously, the area is massively under threat from development and pollution. Large sections have been quarried, particularly of the northern limestone ridge of the study area, while extensive sections of the remaining lake are being exploited for aquaculture and its shores cultivated for agriculture. Much of the land has been sold for development, and along the edges of the lake harmful polluting industries such as salt factories, geochemical and petrochemical processing plants, as well as numerous sewage plants, all present a real threat to this lacustrine environment – the landscape is constantly changing and with it the archaeological record is being destroyed (see Ramses & Omar this volume). For all these reasons it was determined that a comprehensive survey of the area was essential in order to record what remains before the information is lost forever.

Due to its location and characteristics, Lake Mareotis played a significant role in the internal and international transport system of Greco-Roman Egypt (Rodziewicz 1998), particularly after it had been connected to the sea by a navigation canal at Alexandria (Strabo 17.1.7: 31). It became an important conduit of communication between the River Nile and the Mediterranean Sea, and river boats would have transported trade goods from the Nile Valley to the harbours of Mareotis and onwards to Alexandria and beyond. As indicated, Mareotis also supported economic activities around its shores such as the cultivation of vines and grain, and the manufacture of ceramics, wine and glass (Empereur & Picon 1998). However, despite the important role of Lake Mareotis, previous research conducted in the region has been largely limited to specific areas and topics, such as the work of El-Fakharani (1983) amongst others, on the Byzantine port of Marea (Philoxenite) (see now Haggag, Babraj & Szymańska, Rodziewicz and Pichot this volume), Empereur & Picon (1998) on the amphorae and wine production in Mareotis (also see Dzierzbicka this volume),

Objective s of the Lake Mareotis Research Project The objectives of the collaborative Lake Mareotis Research project were to:

Fig. 2: An approximation of the ancient and present limits of Lake Mareotis ( E. K halil) . 27

LAKE MAREOTIS: RECONSTRUCTING THE PAST Systematically survey the western arm of Lake Mareotis (shores 40 x 3 km wide, and Mareotis Island,1 3.7 km in length and up to 680 m wide , bearing in mind the complex palaeogeographical history of the lake. Record, quantify and assess the archaeological resource. Determine the extent, chronology and, where possible, function, of each site. Determine the nature and the extent of the economic and maritime activities of the region. Determine the significance of the Mareotic region in relation to Alexandria, particularly during the Hellenistic, Roman and Byzantine periods when the lake is believed to have been most active. Identify the degree of threat to the archaeological resource in relation to each site before it is lost to development and pollution.

A luxgate Radiometer Magnetometer geophysical survey was conducted to establish the layout of structures beneath the surface at specific sites.2 Ceramic survey was undertaken to determine the chronology, nature and function of each site. Sedimentological survey to determine changing lake levels and identify geomorphological changes was conducted at selected sites. Limited site-clearing to articulate walls and identify more detailed building phases, was undertaken. A photographic archive was compiled of each site and features within the site. Analysis of the five survey seasons conducted in the western arm of Lake Mareotis between 2004 and 2008 is almost complete, and brief preliminary observations are presented in this paper. The final results will be published in a single monograph detailing the work undertaken and an interpretation of the final results Blue & Khalil forthcoming . In total, over 70 sites were documented, ranging in date from the Hellenistic period to the 7th century AD (Fig. 3). A clear distinction was noted between the nature of the sites on the northern shores compared to those in the south and on Mareotis Island. The north shore sites were clearly defined, but quite heavily eroded, tell sites. The sites to the south appear to represent more dense urban concentrations with associated agricultural and industrial areas, particularly towards the east where the island ridge may have once been an extension of the southern shore and the ridge associated with the site of Marea. Many of the sites, particularly on the southern shore, were partially buried under aeolian and alluvial sediments deposited by the eroding hillside and wind-blown material from the north.

Methodology As a result of the pilot survey conducted in 2004 a total of over 60 sites were identified, as previously mentioned. Some of these sites were already documented in the archaeological record, some having been partially excavated by the Western Nile Delta section of the SCA and foreign missions, or noted but not recorded in detail in previous publications. However, at least half of the sites identified are believed to be new discoveries. The sites were prioritised on the basis of their maritime significance, particularly where maritime installations such as moles, jetties, quays, mooring rings etc., had been identified. GPS coordinates were taken to fix the position of each site located, each site was photographed and basic data pertaining to its nature were recorded. Since 2004 four further seasons of survey have been conducted between autumn 2006 and summer 2008, with the objective of recording in detail the maritime installations and to place the sites in their environmental and economic context. In order to achieve these objectives, the following methodology was implemented: A desk-based assessment of the region was undertaken with the aid of high resolution Quickbird satellite images, to establish the nature and extent of the sites already identified. This information along with all the data acquired as part of the survey project, is archived in a GIS-based database. Each site was surveyed using a GPS-based Real Time Kinematic (RTK) satellite navigation system to produce detailed topographic maps of each site. Specific buildings and/or features were planned using a Leica TCR705 Total Station, and the data was downloaded in the field into AutoCAD via TheoLt software.

N orth Shore Sites (Fig. 3) A total of ten sites were identified on the northern shore. They date from the Hellenistic to Late Roman periods with a concentration of activity in the Roman period. They were primarily tell sites and their size ranged from c. 7000 to 180,000 m2. The ceramic assemblages at the sites included a mixture of amphorae, finewares, tablewares and cooking wares. Many of the finewares and amphorae were imports. Many of the sites on the northern shore contain wells and cisterns as well as evidence for mortar-lined basins, probably from bath complexes, which would suggest a domestic context. Limited evidence of irrigation exists in the form of water wheels or sakkias, and qadus pots associated with water wheels or sakkias were found at some Early and Late Roman sites. The evidence suggests that these sites were largely domestic in nature and each site represents a fairly contained and distinct unit with a clear association with

2. Preliminary investigations in the 2007 season were not as insightful as we had hoped, due to the fact that the fluxgate radiometer M36 we had been using had limitations. Thus, the 2008 geophysical survey was conducted with a Bartington Instruments Grad 601-2 dual sensor fluxgate gradiometer. The dual sensors survey at twice the rate of a single sensored instrument and with more refined results.

1. The Lake Mareotis Research Project refers to this ridge, largely surrounded by water located in the south-eastern region of the survey area some distance from the southern shore of the lake, as Mareotis Island. However, its island status in antiquity has yet to be determined and it is also acknowledged that it is not, and was not in antiquity, the only island in the lake.

28

L. BLUE: LAKE MAREOTIS RESEARCH PROJECT

Fig. 3: Sites located along the northern shoreline, the southern shoreline and on Mareotis Island ( Lake Mareotis Research Proj ect) . sites extend across the lake towards each other. Thorough investigation of the ceramic samples collected from these sites will shed more light on their possible nature and relationship.

the waterfront. A number of the settlements featured jetties extending into the lake on their southern shores. The survey also revealed possible spatial and functional relationships between a number of sites on the northern shore and specific sites on the island and along the southern shore. For example, it was noticed that at least one of the largest sites on the northern shore of the lake (AlGamal, Site 204) is located immediately opposite a settlement located at the western end of the island (Site 23; see Hopkinson this volume). Quays and jetties from these

South Shore Sites (Fig. 3) A total of 26 sites were identified on the southern shore. The sites ranged in date from the Hellenistic to the Late Roman periods, but generally exhibited limited activity in the Early Roman period. The earlier period displays some imports, largely amphorae from the Aegean region, but 29

LAKE MAREOTIS: RECONSTRUCTING THE PAST

is predominantly composed of locally produced material, while the Late Roman material reveals a greater amount of imported pottery from all over the eastern Mediterranean.

Fig. 4: Topographic, ceramic and auger survey plan of Site 4 on M areotis Island ( image Lake Mareotis Research Proj ect) .

The sites towards the western extremity of the southern shores formed distinct and in some cases (Site 44 and Site 109) quite substantial units, and are spaced well apart. The archaeology at these sites was largely concentrated at the foot of the limestone ridge (Fig. 4). The sites extended north towards the shore of the lake where many exhibited a series of linear features that extend parallel to the shoreline down to the water’s edge and were open to the E-NE. They appeared to support low-lying areas of marshy ground between the main occupation area and the shoreline ridge, or in many cases a boundary wall to the north. This type of linear feature, known as a ‘lake wall,’ essentially contained a body of water and could have been used for agricultural purposes, perhaps for feeding animals or harvesting reeds, as is still practiced in the region today. This was noticed at a number of sites and also towards the western end of the island (see below) (Fig. 5).

Fig. 5: An example of a ‘ lake wall’ feature ( Site 4) ( photo Lake Mareotis Research Proj ect) .

To the east, the sites were located on a limestone ridge that is effectively an extension of the ridge upon which both Mareotis Island and the site of Marea are situated. The nature of these sites appears to reflect similar dates and activities to that of the eastern end of the island. A number of sites have been identified that contained wine production facilities and/or sakkias designed to facilitate the shifting of water for example Site 100, and Sites 01 and 106 . Generally, however, these sites are less substantial than 30

L. BLUE: LAKE MAREOTIS RESEARCH PROJECT of a few ceramics previously identified by El- akharani El- akharani 1983: 176-178 . Therefore, these buildings might contribute towards resolving the mystery of the so far elusive Pharaonic harbour of Marea raser 1972: 146 El- akharani 1991: 28, 1983: 176 .

those on the island or by comparison with Marea, and they tend to cluster opposite the western end of the island, or along the ridge that extends between Marea and the island. Further investigation is needed to identify the relationship between the eastern end of the island, the south-easterly sites and the western extent of Marea.

Mareotis Island (Fig. 3) Mareotis Island is a limestone ridge joined to the southern shore at its western end. It is possibly an adjunct of the southern shore that has extended out into the lake from the Marea Ridge. It is aligned roughly northeast – southwest and is 3.7 km in length and 680 m at its widest point. Some 40 sites were initially identified primarily concentrated at the eastern and western ends of the island, with a few sites in its central section that were largely industrial-scale Early Roman kiln sites (see Hopkinson this volume).

Two sites stand out as exceptions to the others noted along the southern shores; the Kibotos (the square-shaped harbour) (Site 09), and the ‘complex building’ (Site 13) (Figs. 6 & 7 . The Kibotos Site 09 is essentially a harbour area constructed of large limestone blocks (0.8 x 1 m), defined by a series of moles that enclose an area some 60 m long to S and 36 m E to W wide see Khalil this volume). A second site, known as the ‘complex building’ (Site 13), is located a short distance to the southwest of the Kibotos and is composed of a continuous wall constructed of huge blocks (1.3 x 1.1 m), with a smooth outer surface that slopes up and inwards from the foot of the building. It has been proposed that both these buildings date to the Pharaonic period (Haggag 1984: 277-280). No earlier Pharaonic material has yet been identified at any of the sites recorded in the survey area, with the exception

The majority of the material is Hellenistic in date, particularly towards the eastern end of the island, with a flurry of activity again in the Late Roman period, particularly towards the western end of the island. The ceramic assemblage indicates some imported material especially for the earlier Hellenistic period, but in general it is locally

Fig. 6: The K ibotos site ( Site 09) ( photo Lake Mareotis Research Proj ect) .

Fig. 7: The ‘ complex building’ ( Site 13) ( photo Lake Mareotis Research Proj ect) .

31

LAKE MAREOTIS: RECONSTRUCTING THE PAST produced, particularly the production of amphorae. The nature of the structures, with the exception of the centrally located industrial-scale amphorae production kiln sites, is essentially urban in character. The scale of the buildings, particularly those located on the ridge towards the east, is impressive, and many are associated with numerous waterfront structures located along the northern shore on the edge of the coastal plain. At the western end of the island, the settlement extends from the ridge to the north over a flat plain. Whereas the ridge supports further urban structures, the plain to the north appears to be more agriculturally orientated, with features including possible ‘lake walls’ and sakkias similar to those identified along the south-western shores of the lake (see above).

the lake’s Mareotic Arm (see Khalil this volume). The frequency and abundance of these installations along the shores of the Mareotic Arm reflect the extent of maritime activity that was taking place in this region. Besides the jetties and quays which are positioned perpendicular to the shore, another type of maritime installation has also been identified, positioned parallel to the shore, and described as a lake wall. The so-called lake walls define the shores and protect them from the effects of silting and sedimentation. As many as five lake walls were discovered along the southern shore of the lake and the northern shore of Mareotis Island, areas which are more subject to silting as a result of the prevailing northwest wind. The lake walls could have been utilised as docking facilities for merchant vessels, but were more probably associated with agricultural activities, perhaps the cultivation of reeds as is still witnessed in the region today.

There is also evidence for quarrying on the island, as well as rock-cut tombs. Towards the eastern end of the island, a possible town wall has been identified together with a series of jetties and platforms associated with the waterfront, possible warehouses and a small inlet embayment. All these are discussed in more detail in Hopkinson (this volume).

The Lake Mareotis Research Project has also recorded numerous archaeological sites of industrial and agricultural nature such as amphorae production sites, wineries and the remains of several water wheels (sakkia) which were used for lifting water for irrigation and for use in baths and other domestic purposes. With one exception, all the amphorae and wine production sites are located on the southern shore of the lake or on Mareotis Island. In antiquity, the southern shores of the lake were characterised by a fertile plain which supported and enabled agricultural activities to flourish. In addition, the abundance of calcareous clay, particularly suitable for amphorae production, resulted in a thriving large-scale amphorae industry. The study of the ceramic assemblages collected from the surveyed sites indicate that the amphorae production centres were active from the Ptolemaic to the Late Roman period.

What is worthy of note is the fact that the majority of sites on the island are concentrated along the ridge and the northern shores, and that several were clearly built in a very specific relationship with sites on the lake s north shore (see above). The northern concentration of sites has also raised questions about the island’s status. Does the fact that so few sites have been found along the island’s southern shores suggest that it may have been linked to the southern shore of the lake in antiquity? Recent geomorphological work suggests that in fact the area to the south of the island has always been waterlogged to some degree, but that the waters were shallow and access by boat was problematic, and as such activities in this area were restricted.

Since no lake-side archaeological sites were recorded on the southern shore of the lake opposite Mareotis Island, questions have been raised about the nature of the island and whether or not it was indeed an island in antiquity. However, sedimentological analysis carried out during the final survey season revealed that Mareotis Island was probably an extension of a ridge that extended some 10 km to the west from the site of Marea. In addition, it was ascertained that the area between the island and the southern shore of the lake was either waterlogged or marshland in antiquity (Flaux forthcoming).

Preliminary C onclusions The Lake Mareotis Research Project has identified and systematically surveyed over 70 sites dating from the Hellenistic period through to the 7th century, in what still remains of the western arm of the once much larger Lake Mareotis. The sites vary extensively in nature, size and function, but demonstrate the importance this area had in relation to Alexandria and the mechanisms by which Mareotic products were produced and arrived at the city. Local pottery production is indicated by the numerous kiln sites discovered in the region, particularly on the island. There is also considerable evidence for imported pottery from as early as the late 4th century BC and as late as the 7th century AD. The nature and size of the sites identified varied from rural small holdings to large urban settlements, with associated production sites, water storage facilities and agricultural and industrial complexes.

Previous research conducted in the Mareotic region has been largely limited to specific sites however, the Lake Mareotis Research Project has systematically investigated all the archaeological remains along the shores of the lake’s western arm. The project also assessed the results of previous archaeological work undertaken in the Western Deltaic region. Based on this, it is reasonable to suggest that the Mareotic Arm was in fact the most active area in Lake Mareotis in antiquity, and that most of the Mareotic products arriving in Alexandria were in fact coming from the western Mareotic Arm, rather than from settlements associated with the lake’s main body (see Khalil this vol-

Different types of waterfront installations were recorded, including more than ten anchorage facilities, which form the majority of maritime installation along the shores of 32

L. BLUE: LAKE MAREOTIS RESEARCH PROJECT que d’Athè nes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole franç aise d’Athè nes, 11-12 décembre 1988. Athens, Paris. Empereur. J.-Y. (ed.), 2002, Alexandrina 2. Cairo. Empereur, J.-Y., & Picon, M., 1998, Les ateliers d’amphores du Lac Mariout. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athè nes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole franç aise d’Athè nes, 11-12 décembre 1988: 75-88. Athens, Paris. Flaux, C., forthcoming, Geomorphological survey of the south-west shores of Lake Mareotis. In L. Blue & E. Khalil (eds.), A Multidisciplinary Approach to Alexandria’s Economic Past: the Mareotis Case Study. Oxford. Fraser, P.M., 1972, Ptolemaic Alexandria, Vol. I. Oxford. rihy, .E., 1996, Some Proposals for Coastal Management of the Nile Delta Coast. O cean & Coastal Management 30.1: 43-59. Goddio, F., Bernard, A., Bernard, E., Darwish, I., Kiss, Z., & Yoyotte, J., 1998, Alexandria: The Submerged Royal Q uarters. Oxford. Haas, C., 1997, Alexandria in Late Antiquity: Topography and ocial Con ict. Baltimore. Haggag, M., 1984, Ports in Ancient Egypt till the Arab Conquest. Unpub. MA diss., Alexandria University. Hassan, F.A., 1997, The Dynamics of a Riverine Civilization: A Geoarchaeological Perspective on the Nile Valley, Egypt. World Archaeology 29.1: 51-74. Rodziewicz, M.D., 1998, Fom Alexandria to the West by Land and by Waterways. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athè nes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole franç aise d’Athè nes, 11-12 décembre 1988: 93-103. Athens, Paris. Said, R., 2002, Did Nile Flooding Sink Two Ancient Cities? Nature 415: 37-38 Strabo, G eography, H.L. Jones (transl.), 1949. London. Warne, A.G., & Stanley, D.J., 1993, Late Quaternary Evolution of the Northwest Nile Delta and Adjacent Coast in the Alexandria Region, Egypt. J ournal of Coastal Research 9.1: 26-64.

ume). Finally, the Lake Mareotis Research Project has outlined the true scale of activities in the Mareotic Arm with respect to industrial, agricultural and urban sites, being much more extensive than previously suggested, thus supporting the significant importance of the area and highlighting the role it played in the economy of Hellenistic and Roman Alexandria. Acknowledgments The author would like to thank the Supreme Council of Antiquities for permission to work in the Mareotis region, particularly colleagues from the Underwater Department who were instrumental to the success of this project, in particular the support and help of the project co-director Dr Sameh Ramses and project manager Dr Emad Khalil. The project would not have been possible without the financial backing of both the British Academy and the Leverhulme Trust particularly for funding Dr Khalil’s two-year post doctoral fellowship. Finally, thanks is due to all students of Southampton and Alexandria Universities and Egyptian and UK colleagues who worked on the project without whom the survey would not have happened. Bibliography El-Fakharani, F.A., 1983, Recent Excavation at Marea in Egypt. In G. Grimm, H. Heinen & E. Winter (eds.), Aegyptiaca Trevernsia II, Das Römisch-Byzantinische Ägypten. Akten des internationalen Symposions 26.-30. September 1978 in Trier: 175-186. Mainz. El-Fakharani, F., 1991, The Kibotos of Alexandria. Studi Miscellanei 28: 21-28. Blue, L., & Ramses, S., 2005, Lake Mareotis Research Project – Preliminary Report from Pilot Survey, August 2004. Newsletter of the Archaeological Society of Alexandria 15 ( February) : 5-16. Blue, L., & Khalil, E. (eds.), forthcoming, A multidisciplinary approach to Alexandria’s economic past: The Mareotis case study. Oxford. De Cosson, A., 1935, Mareotis: Being a Short Account of the History and Ancient Monuments of the NorthWestern Desert of Egypt and Lake Mareotis. London. Empereur, J.-Y. (ed.), 1998, Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du collo-

33

LAKE MAREOTIS: RECONSTRUCTING THE PAST

34

D. HOPKINSON: MAREOTIS ISLAND SURVEY

The Results of a Preliminary Survey at Mareotis Island Dylan Hopkinson

Introduction

northern and southern limits of the western arm of Lake Mareotis and form a natural flooded depression known as the Mallahet Maryut Depression (Fig. 1). The island measures 3.7 km long by an average of 400 m (maximum 680 m) wide and is connected to the mainland by an area of low-lying marshy ground to the southwest at the foot of the Gebel Mariout Ridge (Fig. 2). The lands around Lake Mareotis formed an important agricultural production centre in antiquity known to have been richly populated with farming villages, market-towns and villa estates (Strabo G eography 17.1), and would have been dependent on seasonal rainfall and ile floodwaters for their irrigation and fresh drinking water.

Location The site known as Mareotis Island lies 65 km southwest of Alexandria within the western arm of Lake Mareotis, closely associated with its southern shore. It lies 4 km to the west of the site of Marea/Philoxenité, and directly north of the modern town of Bahig (Fig. 1). The island was subject to systematic survey in 2007 and 2008 as part of the broader Lake Mareotis Research Project (see Blue this volume). A large number of archaeological features are clearly visible on the ground which date to the Greco-Roman periods largely indicated by the substantial quantities of ceramics present in surface scatters. This paper presents the overall character of the archaeological features observed, and some preliminary interpretation with specific reference to the littoral zone on the northern coastal plains where ‘maritime’ and lakeside activities were largely focused.

Environmental Background In antiquity Lake Mareotis provided a transport connection between Alexandria and its agricultural and industrial hinterland in the Mareotic region; and via the Nile waterway to the Egyptian interior along the Nile Valley. Strabo states that the lake is filled by many canals from the ile, both from above and on the sides” (G eography 17.1.7). It was these canals that provided the fresh water which came

The island is formed from the remnants of a small limestone ridge which lies between the two much larger Abusir and Gebel Mariout Ridges. These two ridges define the

Fig. he location of Mareotis Island and the sites mentioned in te t modified from e Cosson

35

.

LAKE MAREOTIS: RECONSTRUCTING THE PAST Fig. 2: Schematic of the main features of Mareotis Island ( Hopkinson) .

from the Canopic Branch of the Nile apparently keeping the lake water sweet, and filling the cisterns of Alexandria. The direct link to the ile during the seasonal floods is also likely to have had a significant impact on lake levels. The exact location and number of these connections is not firmly understood, but it is likely that the ancient courses would have mirrored those of the more recent canals (Warne & Stanley 1993: 15-16).

of sites located on the southern shore of the lake, where Roman foundations are observed to have been buried under 3 m of deposits (Khalil 2005: 39). Holocene sedimentation rates between 0.02 and 0.07 cm per year are indicated by Smithsonian boreholes either side of the island (Warne & Stanley 1993: 50). This sedimentation is likely to have obscured archaeological features over time, and to have raised the level of the lake bed while consequently reducing the depth of the lake waters (Flaux 2008: 9-10, III.2).

There is sufficient environmental evidence to show that the climate of the region has not changed significantly in the past two thousand years (Ball 1942: 20; Brooks 1949: 333). Evidence for this includes an analysis of a weather diary kept in Alexandria by Ptolemy in the 2nd century, which concludes that “the number of rainy days per annum was much the same as it is nowadays, though the rainfall was more evenly spread out over the year” (Murray 1951). So we can conclude that the climate was semi arid as it is today. Today rainfall in the Western Coastal Desert region is greatest along the margin by the sea and is typified by its great variability from location to location and from year to year, which suggests that in antiquity even this fertile and productive region would have had a precarious existence were it not for the fresh waters of the lake and the reliable annual ile floodwaters.

Modifications to the Lake since the G reco-Roman Period The water regime in the area has been greatly modified since antiquity, most notably by the silting of the connections between the lake and the Nile; when the Canopic Branch of the ile finally silted up around the 12th century the lake also began to dry up and shrink in size. Strabo also tells us that the extent of the lake must have been much greater in antiquity than it is today, and extended around 40 km to El Bordan to the southeast and to Kom Al Trouga 70 km to the southwest of Alexandria (De Cosson 1935: 26) (see Fig. 1). Warne & Stanley also support these dimensions which amount to around 700 km2 of land lying below the 0 m sea-level datum (Warne & Stanley 1993: 53).

Light north-westerly winds are generally observed in the region and these transport large quantities of clayey loess sediments from the degrading limestone ridges and silts and sand from the coastal plain. These silts are deposited in colluvial sheets at the base of the ridges and on north facing shores. This phenomenon has been noted previously (Warne & Stanley 1993: 53), and at a number

It is very likely that even without a direct connection to the lake the effects of the ile flood would still have been noticeable in the water levels due to groundwater seepage (A.G.,Brown, Univ. of Southampton, pers. com.). When the Aswan Dam was constructed in 1964 there was therefore a significant change to the water level in the ile 36

D. HOPKINSON: MAREOTIS ISLAND SURVEY Delta aquifers and the lake. As a result the hydrodynamic environment today is very different from antiquity and we have no direct way of understanding the range of water levels that would have been experienced during the floods. As Ramses & Omar (in this volume) indicate, modern infrastructure, fisheries and agricultural drainage ditches have further modified the lake, splitting it into several artificial basins and remodelling large stretches of coastline. These basins have water regimes that are artificially managed for modern needs, and large amounts of water are pumped into the sea to keep the waters artificially low for land reclamation projects.

this sakia to the cisterns of the town. To the south of this are the remains of two very interesting pottery kilns with a large heap of broken pottery thrown there as it was ‘scrapped’ from them. North of the sakia there is a long jetty running into the Lake from which ferry-boats once plied.” (De Cosson 1935: 130)

The reference here to waterwheels (‘sakia or ‘saqiya’), cisterns, ceramic mounds from pottery production and lakeside features characterise the island in general terms, but the description bears little direct relation to the position of the remains observed today. Consequently the archaeology of Mareotis Island represents a significant untapped resource that can enhance our understanding of the region.

Today the lake is fed only by water seepage from the sea, precipitation, and by irrigation ditches that channel agricultural waste-water from the Beheirah Governorate. As a result of all of these changes, it has been calculated that the current lake level is only 13% of the ancient lake surface area (Warne & Stanley 1993). Observations made during the 1970s show that the western arm of the lake was totally dry at least from Marea/Philoxinite and to the west. However, a recent geoarchaeological survey as a component of the Lake Mareotis Research Project indicates that the depositional environment close to the modern shoreline had always been lacustrine except during the recent centuries (Flaux forthcoming).

Fieldwork In August 2007 under the auspices of the Lake Mareotis Research Project, a six-week survey was undertaken to plot the visible archaeological remains on the island using RTK GPS and Total Station surveys. In order to provide an understanding of the dating of features a controlled surface ceramic collection strategy was adopted. An RTK GPS topographic model was also completed with 5 m transects across the island. In addition to this, a geoarchaeological auger survey was undertaken in 2008 to help contextualise the results and to begin to understand how site formation processes can inform our understanding of the site. Of particular interest here was the relationship between the ancient water levels and the archaeological features, and the impact of the changing lake levels on the development of the sites. A more detailed account of the Lake Mareotis Research Project and its methodologies is presented in this volume (see Blue this volume, see also Khalil and Ramses & Omar this volume)

Deltaic subsidence and tectonic shifts may also have changed the topography of the region as it is known to have done within the Nile Delta region further to the east; however it has not proved possible to use existing geoarchaeological datasets to understand the significance of these events. Consequently as a result of the various geographical factors described, we must accept that the current lake level cannot be used as a reference datum from which to reconstruct ancient lake levels and to assess the impact of the annual floods and the summer evaporation on the shoreline. However, understanding these relationships is a fundamental factor to our interpretation of the archaeological structures lying near the shore.

The majority of archaeological structures are located on high ground, occupying the central limestone ridge that forms a series of hills that extends as a discontinuous spine along the length of the island. Two main settlements were identified concentrated at the eastern and western extremes of the island, with further activities evident on the north facing coastal plains below the central ridge. The central section of the island is some 2 km long and is much less densely occupied than the western and eastern extremities. The activities here represent industrial amphorae production and possible agricultural activities.

Previous Research There has been a significant increase in our understanding of the archaeology of the Mareotic region in recent times brought about by the work of El-Fakharani (1974, 1983, 1984), Rodziewicz (1983, 1990, 1998a, 1998b, 2002), and Empereur (1986, 1998; Empereur & Picon 1986, 1998, 1992), however previous research on the island itself appears to have been negligible. There are numerous cartographic representations of the region, two of the most informative being the one on Sheet 37 of the Atlas G éographique in the escription de l gypte from 1809, and on Carte des en irons d le andrie produced in 1866 by Mahmoud Bey El-Falaki; however, the detail pertaining to the island in both cases is limited and difficult to reliably geo-reference. A further cryptic description of the archaeology on the island is given by De Cosson (1935):

C hronology The ceramics recovered from the island indicate that the settlements was well established some time during the Hellenistic period and that occupation potentially continued through to the 7th century AD. The western settlement appears to have been continually occupied as indicated by ceramic forms recovered, whereas the eastern settlement seems to have suffered a decline in activity in the Early Roman period which reached its nadir some time in the 4th to 5th centuries with a late recovery in the 6th century. The results of the ceramic survey are not commented on in detail here, but will be published shortly (Tomber & Thomas forthcoming).

“Many buildings can be traced, and at the eastern end the circular stone platform and oblong well of an ancient sakia will be found. Stone channels are traceable leading from

37

LAKE MAREOTIS: RECONSTRUCTING THE PAST Archaeology of the Eastern Settlement The eastern settlement (Fig. 3) is concentrated on the high ground of the limestone ridge which measures approximately 650 m by 350 m; the majority of this area is covered with building remains of various construction styles and materials. The layout and construction styles of the visible building remains indicate some degree of social organisation; such as possible administrative and commercial areas, some are grouped around what appear to be large, open areas at a number of locations on top of the ridge. A long boundary wall is located winding along the top of the ridge in the north-western limit of the settlement; this wall measures 240 x 1.20 m wide, and may have served a defensive function protecting the settlement from the west. To the west of the settlement there is a small low-lying depression that could well have formed a natural inlet prior to silting; around its northern and southern margins the remains of buildings that may have been possible wharf structures have been identified. This inlet appears to have served as an inland harbour which would have provided sheltered mooring. Auger cores taken during the 2008 season confirm that the sediments in this depression were indeed deposited in lacustrine conditions in antiquity (Flaux forthcoming).

are difficult to interpret in detail because there are several phases of activity. The relative dating of these features has not yet been possible. These structures appear to fall into several categories of linear features, walls and lakeside buildings. Linear Features ( G roup A) The first category of structures is a series of five linear features located on the low-lying northern coastal plain that extends between the present northern shoreline of the island and the foot of the limestone ridge (Fig. 3, Group A). These features range between 65 to 120 m in length, and appear as low, roughly built earthen ridges up to 4 m wide which are spaced roughly between 30 to 100 m apart. At the waters edge, additional features can be seen that extend along the current shoreline. As yet, these water front features have an undefined chronological relationship to the linear features (Group A) and represent primary and secondary activities over a period of time. The landscape between the base of the limestone ridge and the lakeside features appears to be largely devoid of additional structures and the linear features are relatively isolated with only a few abutting or adjacent features. It has already been established that little is known about the absolute water-levels of the lake in antiquity, and unfortunately the 2008 geoarchaeological survey did not target the area around the linear features therefore our interpretations

To the north of the eastern settlement, along the northern coastal plain of the island, between the lake and the foot of the ridge, are a series of structures and buildings which

Fig. he littoral features mentioned in this paper and topographic sur ey of the eastern settlement of the island ( Hopkinson) .

38

D. HOPKINSON: MAREOTIS ISLAND SURVEY of the exact nature and function of these features remains speculative, however two possibilities present themselves. Firstly, they may represent crude earthen jetties projecting from the ridge into the lake during periods of higher water-levels. The other possibility is that the coastal plain was dry in antiquity and that the earthen ridges may have divided the coastal land into units, potentially for cultivation as is suggested by the interpretation of similar features identified in the western settlement. We know from ancient sources that the Nile had a dramatic effect on the Delta region; Strabo states “at the rising of the Nile the whole country is under water and becomes a lake, except the settlements; and these are situated on natural hills or on artificial mounds, and contain cities of considerable size and villages, which, when viewed from afar, resemble islands” (G eography 17.1.4). It is therefore possible that the annual variation in the level of the lake level was large enough for both scenarios to be valid.

‘quay’ wall to demarcate the edge of the lake either as protection from inundation or to form a working platform, perhaps for the mooring and loading of boats. In some instances the construction of these sections of wall seem to be related to the linear earthen ridges of Group A, and it would appear that in some cases they were built to connect or at least respect, the linear features lakeside termini. The elevations recorded from various sections of these walls indicate that the surviving top course of the walls are not all at the same level but in fact exhibit a difference in height of over 1.83 m. This may indicate that either the top course of some of the sections is missing, or that these walls were built at different times and at different mean lake-levels rather than as a single planned ‘quay’ building effort. Coastal uildings Associated with the first group of linear features Group A) are a large number of walls indicating the presence of a number of buildings that had been constructed along the shoreline in this area (see Fig. 3). Since the relative dating of structures has yet to be determined it is not possible to discuss the phasing of the structures. However, what is apparent is that if the linear features were jetties and if they were built at the same period as the coastal buildings, then these buildings would have been susceptible to serious flooding. It is vital that the relationships and phasing of these features is investigated as it will have important implications with respect to interpretation of the landscape and the features located within it.

Linear Walls ( G roup B) A second group of linear walls were observed to the west of Group A along the north-western shores of the lake in the eastern settlement area. This group is located to the west of the wall that winds along the top of the western side of the limestone ridge, and down onto the coastline (see Fig. 3). These linear features were observed as low rubble walls 0.50 m thick, their northern lakeside ends are close to the current waterline and appeared to extend inland for distances of between 25 to 60 m. None of these walls reached the foot of the limestone ridge and in most cases extended less than half way across the plain. Three of these walls were spaced approximately 35 m apart with a further group 30 m to the east that was composed of three roughly parallel walls in very close proximity. The coastal ends of the first three walls had associated low rubble linear structures or platforms which extended for a short distance along the coastline.

Amongst these coastal buildings are a group of structures that deserve particular attention. Three isolated groups of buildings of similar construction were identified at three different areas of the plain (see Fig. 3; Group D). Two of these are located on the north coast and a third on the eastern extreme of the island with a south-easterly coastal aspect. The buildings appear to be arranged in symmetrical blocks facing each other and around 13 m apart, each measuring approximately 10 m wide and between 30 and 55 m in length, and were aligned perpendicular to the coast. The building pairs consist of single rows of cellular rooms measuring roughly 10 m by 5 m. Their location and alignment relative to the coast suggests that their function was somehow related to the lake and that they could have been warehouses, shops or boathouses.

The alignment of the main walls of Group B is interesting since they were constructed on the same orientation as the linear features discussed above (Group A). However, the Group B linear features are not perpendicular to the shoreline since this section of the coast turns to the south and is west facing. As a result these features are on an oblique alignment relative to the current shoreline. The stone construction, relative close spacing, alignment to the coast, and their location to the west of the long winding ‘boundary’ wall, suggest a distinct function as compared to the Group A linear features. One possibility is that these features may have been concerned with land reclamation or stabilisation, serving perhaps as a series of ‘groins’.

Summary Overall there appears to be a great deal of established activity in the eastern settlement. Buildings on the ridge settlement hint at developed social organisation and those on the northern coastline show a strong association with the lake. Features interpreted as an inland harbour, lakeside walls, and possible jetties, all suggest an established interaction with the lake. There are a far greater number of buildings on the northern coastline as compared to the south coast in this part of the island, which may suggest that the southern coast was possibly dry or marshy and therefore inaccessible to boats in antiquity.

shlar Lake alls Group C Large sections of dressed ashlar blocks are observed forming stretches of wall along the northern edge of the plain (see Fig 3). These walls closely respect the natural topography of the current shoreline and appear to have originally been laid as stretchers extending parallel to the shore. This might suggest that the walls were built as a sort of 39

LAKE MAREOTIS: RECONSTRUCTING THE PAST Archaeology of the Western Settlement The western settlement occupies a similar location to the eastern settlement, situated on the high ground of the limestone ridge which measures 820 by 200 m (Fig. 4). The area exhibits the remains of various buildings, with a mixed function suggested by the construction materials and surface finds. verall the preservation of these structures is not good; a large area in the extreme west has been heavily damaged by earthmoving machines, and in other instances only patchy areas of construction and dislocated wall lines survive. As a result, it is not possible to identify any zoning of functionality or social organisation, although a few significant groups of structures were observed. As in the eastern settlement the northern coastal plain appears to be the focus of activities. Once again it is the lakeside features that are the focus of interest in this paper although some of the structures on the ridge will be discussed.

Ashlar Shoreline Wall Alignment ( Feature E) The area along the waterline is densely overgrown with water reeds and has silted up making the water level very shallow and the ground marshy. Amongst the reeds can be seen the upper surface of sections of a long straight wall (see Fig 4). All the shoreline features in the western settlement appear to respect this northern boundary. The wall is built from limestone ashlars laid as stretchers with a flush north-facing aspect, suggesting that it was laid in a specific relation to the lake. The various sections of this feature appear to suggest a continuous wall with overall dimensions 0.40 m by as much as 245 m in length. No excavations of this feature were undertaken so it is not possible to indicate the number of courses or the height of the wall. The location and construction of this wall suggest that it has a significant relationship to the shoreline if water levels similar to those seen today are assumed. It appears to be a lakeside wall protecting the land behind or creating a working platform from which to conduct activities related to the lake. Some of the other features close to this wall appear to butt against this wall suggesting that it may also form part of a larger pattern of structures.

The features of the coastal plain are sparsely distributed and appear to fall into three clear categories: a section of ashlar wall aligned along the shoreline; long, thin linear walls forming low ridges perpendicular to the coastline; and localised masonry features typified by curvelinear walls and associated ashlar lined pits, there is an overall lack of secondary activities such as the coastal buildings on the eastern coastal plain, which makes interpretation somewhat easier.

Linear Walls ( G roup F) The second group of features located on the coastal plain of the western settlement, is a series of roughly paral-

Fig. 4: The littoral features mentioned in this paper and topographic survey of the western settlement of the island ( Hopkinson) .

40

D. HOPKINSON: MAREOTIS ISLAND SURVEY Schi ler identifies 14 different types of saqiya from Roman, Islamic, and modern contexts (Schiøl er 1973: 13), including wheels with qudus pots attached directly to the rim, and wheels which powered a ‘pot-garland’ or bucket chain. In the case of the features associated with the western settlement it seems that the curved walls represent part of the animals’ circular track, while the ashlar lined pit is likely to be associated with either the wheel or the gear mechanism.

lel linear features aligned perpendicularly to the current coastline; they are constructed from faced ashlar blocks that in some cases have degraded to stone rubble and in some areas only survive as low earthen ridges. Seven potential features were observed in this group (see Fig. 4), all of which have some connection with the current coastline and appear to respect the alignment of the possible lakeside wall. These linear features extend back from the coast, south towards the main ridge; their lengths vary between 25 and 100 m but on average only reach halfway across the plain and appear to show no direct relationship with the ridge itself. Two of these walls are connected by a linking wall at their southern ends.

Many sherds of qudus pots were observed along the northern flank of the limestone ridge adjacent to the plain but were not recorded anywhere else on the island; the location directly adjacent to the water is also appropriate for this interpretation of water management, and places the northern coastal plain in an agricultural setting.

These walls have much in common with similar features at the eastern end of the island, they have the same orientation relative to the coastline and occupy the ground between the coast and the ridge, with their coastal terminus closely associated with sections of ashlar wall aligned along the coast. This suggests that the walls could have potentially functioned as jetties when lake waters were high or as boundary markers potentially prescribing plots of land along the coast that butt against the ashlar wall.

In the light of this interpretation it is possible to reinterpret the long linear walls between the ridge and the quay wall; these may potentially be seen as boundaries demarking areas of farmed land irrigated by saqiyas. Oleson concludes that rudimentary saqiyas from modern contexts could usefully irrigate similar portions of land to those possibly represented by the area of land contained within the linear walls (Oleson 1984: 369).

Cur elinear alls and shlar Lined Pits Features G The land between the linear walls had a sparse distribution of ashlar constructed features that form the third group of structures. They are represented by sections of curved wall in two locations where additional associated wall lines were observed. These curved walls are incomplete but appear to represent rounded features with projected diameters of about 10 m (see Fig. 4).

Linear Mole Extending into the Lake towards a Small Island At the north-eastern extent of the western settlement in the marshy ground of the shoreline plain is a site that is unique on the island see ig. 2 . This is significant because it is the only ancient feature on Mareotis Island that currently extends into the lake, and as such it represents a useful indicator of lake levels in antiquity. The feature is a 250 m long linear mole construction that extends out into the lake to join with a small island that measures roughly 30 by 15 m. This small island supports a rectangular building measuring 20 x 10 m that was built from substantial masonry blocks. Three jetties extended from the north of this building into the lake measuring between 7 and 15 m in length. The feature extends towards a promontory settlement on the opposite side of the lake where there is a similarly built reciprocal linear construction that extends for a short distance into the lake on the same alignment towards the small island. These features imply a functional relationship with the lake perhaps for controlling the movement of transport along the lake as well as between the two settlements.

The interior structure of one of these rounded features is preserved and takes the form of ashlar built components lining a rectangular slot which measured 4.80 by 0.90 m. This feature, although unexcavated and partially observed, bears a significant similarity to ancient water-lifting devices known as saqiya which have been observed archaeologically from broadly contemporary Greco-Roman sites in the region such as Abu Mina located some 15 km south of the island (Schiøl er 1973: 129-136; Oleson 2000: 263; see Fig. 1), and Tuna Al-Gabal near Hermopolis in the Nile Valley (Schiøl er 1973: 141-148). There are further saqiya features in the Mareotic region which have not been published in detail; examples in the immediate vicinity of the island are described as having a “circular stone platform and oblong well” (De Cosson 1935: 130).

Tower-like Feature ( Feature H) A final feature in the western settlement is worthy of mention, this is a small artificial mound of collapsed building material some 4 m in height which lies just northeast of the main ridge on which the settlement is located. After cleaning the flat top of the mound, masonry elements were observed that formed a structure similar to the ashlar lined pits mentioned above. Rather than having a single pit slot this feature had a pit that was divided into two separate apertures with overall dimensions very close to the saqiya discussed above.

Waterwheels were in common use in Egypt from Hellenistic through to medieval times, and are simple devices using a gear mechanism to convert the horizontal movement of an animal walking round a circular track into vertical movement for lifting water. The water was held in pots called ‘ udus which are connected to the wheel. Saqiya are used predominantly in agricultural contexts (Oleson 1984: 126) but have also been recorded in the bath-house buildings, as at Abu Mina (see Fig. 1). Waterwheels are also known to occur in many configurations 41

LAKE MAREOTIS: RECONSTRUCTING THE PAST Fig. Longitudinal and transverse schematic sections through the ater to er on Mareotis Island ( Hopkinson, after McCann et al. .

top of the tower and a further 1.5 m to the vat on top of ridge, an overall height in the order of some 5.5 m, the additional height of about 1.5 m being accounted for by the assumed superstructure which held the qudus pot-garland above the surviving tower.

It is possible that this feature is the top of a tower used to lift water from a well known to exist close to the bottom of the mound using a pot-garland passing through each aperture. An elevated ‘bridge’ constructed from earth and masonry blocks connects the tower to the main ridge at a point where a plaster lined vat is located as part of a larger building complex that could potentially represent a public bathhouse or cistern and water management feature (Fig. 5). An example of such a tower from the late 1st century AD is known from the port of Cosa in Italy (Oleson 2000: 258). The so called ‘spring house’ at Cosa initially lifted water to a height of 3 m but it burnt down and was replaced with a structure that lifted water over a height of 10 m to an elevated conduit. If this hypothesis is correct the structure on the mound would have lifted water to about 4 m to the

Summary The western settlement is generally poorly preserved, however there is an apparent lack of secondary activities on the northern plain which makes it easier to interpret the features that survive. The presence of water lifting devices and possible plots of agricultural land suggest that the majority of the coastal plain was above the mean waterlevel in antiquity. As such an alternative interpretation of the linear walls as possible jetties is not so well supported, 42

D. HOPKINSON: MAREOTIS ISLAND SURVEY although they could have served this function during the flood season. The absence of further buildings on the plain may suggest that this community did not have a strong association with the lake for their subsistence; however the presence of the promontory towards the small island may suggest that at some point the settlement had a significant role in administering activities on the lake.

plications for the interpretation of the other ‘coastal’ walls on the island. To the south of the central ridge of the island, there are two adjacent ‘plots’ of land which are enclosed by low walls of faced stone and rubble (see Fig. 2). These features resemble ‘karum or vineyard plots; features that are usually found inland in a 15 km wide band south of the Gebel Mariout Ridge, as indicated on Survey of Egypt maps from 1927. K arum are features that are poorly researched but the author believes they were used to raise the depth of topsoil on agricultural plots with the advantage of increasing the space for healthy root growth and for storing rainwater in the soil reservoir (Hopkinson 2007). There seems then to have been some small scale industry and subsistence agriculture in the middle of the island that is not directly related to either of the main settlements.

Archaeology of the C entral Island The land between the eastern and western settlements on the island is roughly 2 km long and appears to have been sparsely utilised in antiquity for industrial and agricultural activities. There are four sites of large scale pottery manufacture (see Fig. 2; Kiln “A” & Kiln “B”), each with extensive associated storage rooms, and large mounds of ceramic sherds wasted during the firing process. It seems that these potteries, like many identified near the coast on the southern mainland, were exclusively producing amphorae forms known to have been used for making and packaging the large quantities of wine for which the region is renowned (Empereur & Piccon 1986). Two of the kilns originate in the Hellenistic period (see Fig. 2; Kiln “A”), one close to both the eastern and western settlements. These sites continued in use through to the mid-1st century AD, when amphorae production switched to two adjacent sites (see Fig. 2; Kiln “B”). Production of amphorae at the new sites continued from the late 1st to the 5th centuries AD.

Water Leve ls and C onnection to the Mainland One important consideration for the interpretation of the features on the island is the relationship between the island and the mainland to the south, and the local hydrological environment throughout its occupation. The island is currently largely surrounded by water but it should not be assumed that this was the situation in antiquity because modern irrigation and dyke formation have modified the movement of water. A number of the features that have been identified on the island may suggest a greater connectivity to the mainland to the south. All of the potential maritime focused features such as the possible jetties, quays and lake walls are located on the north facing shores of the island, and the absence of these features on the southern shores suggests the absence of water in this area that could be usefully exploited. Moreover the features that are observed along the current southern shore appear to be associated with dry land activities. The ‘karum plots are agricultural features thought to be concerned with collecting and maximizing the use of precipitation and as such suggest that there is no alternative readily available water resource in close proximity. The second feature that appears to preclude the existence of a significant southerly body of water in antiquity is the large kiln structure observed in the west of the island (see Fig. 4; ‘kiln site’). These kilns are known to have significant subterranean fire chambers up to 4 m deep where the fire was lit, with a perforated firing plate above to allow the hot air to circulate (El-Fakharani 1983 . However, the firing plate of this kiln is currently only a short height above the lake waters to the south allowing little room for the firing chamber and indicating that it could only have been used if the land to the south between the island and the mainland had been dry. The inference that there was connectivity between the island and the southern shores of the lake, is supported by what we know about the organisation of wineries of the region, consisting of three essential elements: the kiln, the winery, and the estate villa (Empereur & Picon 1992). At least three winery structures are known in the immediate vicinity of the island on the southern mainland, and no structures suitable for the production of wine have been identified on the island itself. It is possible to suggest therefore that

Further evidence of light industry on the island is indicated by two small quarrying sites on the limestone ridge (see Fig. 2), although their capacity is very small compared to the quantity of stone used in the settlements and it is more likely that the majority of the building stone was transported across the lake from the large quarry sites known to exist in antiquity on the Abusir Ridge to the north (Oliver & De Cosson 1938: 169-170). In the middle of the island, associated with the kiln sites, there are a group of structures that occupy the top of the ridge that could have been workshops or accommodation for the workmen involved in these industries. A number of features from these structures extend down towards the northern coastline, although on the plain below there is only one visible structure. This is a 75 m long section of ashlars laid as stretchers aligned with the current coastline (see Fig. 2). The wall lies 70 m inland from the current coastline, which is surprising since the flush face of the wall at first suggests a coastal function similar to the sections of lake wall or quay found associated with the eastern and western settlements. However, if this was its function then the land directly north of the wall equates to a large area of siltation and aeolian soil deposition. Small scale excavations either side of the wall in 2008, however, indicated that there was only one course of stones surviving and that the soils surrounding the feature appeared to display similar sediment regimes and were not deposited underwater. It therefore seems that this section of wall had no direct association with the lake, and its purpose is unclear. This finding is also significant since the construction of a flush face apparently respecting the coastline has im43

LAKE MAREOTIS: RECONSTRUCTING THE PAST the kilns on the island were producing amphorae for the wineries on the southern mainland. However, the recent geomorphological enquiry undertaken as part of the Lake Mareotis Research Project has confirmed that the body of water that currently extends from the southern shores of the island was in fact waterlogged in antiquity (Flaux forthcoming). The island ridge was an extension of the Marea Ridge and thus enclosed a body of shallow water to the south. How continually waterlogged this area was in antiquity has yet to be determined but the geomorphological survey has indicated that there was continuous lake sedimentation between the island and the southern shores since antiquity. Thus, it is assumed that access from this area into the main body of the lake was restricted by the ridge, thus deterring settlement on the southern shores of the island, as witnessed by the lack of archaeological sites in this region. It is clear however that the situation is very complicated and that our understanding of the archaeology of the island is incomplete.

El-Fakharani, F.A., 1983, Recent Excavation at Marea in Egypt. In G. Grimm, H. Heinen & E. Winter (eds.), egyptiaca re ernsia II as R misch y an tinische Ägypten. Akten des internationalen Symposions . . eptember in rier, 175-186. Mainz. El-Fakharani, F., 1984, The Kibotos of Alexandria. Studi Miscellanei 28: 23-28. El-Falaki, Mahmoud Bey, 1872, Memoire sur l anti ue le andrie ses faubourgs ses en irons. Copenhagen. Empereur, .- ., 1986, n atelier de dressel 2-4 en gypt au IIIe siécle’ de notre ére’. In J.-Y. Empereur (ed.), Commerce et artisanat dans l le andrie hell n isti ue et romaine. ctes du collo ue d th nes organis par le C R le Laboratoire de c ram ologie de Lyon et l Ecole fran aise d th nes d cembre 599-608. Athens, Paris. Empereur, J.-Y., 1998, Alexandria Rediscovered. New York. Empereur, J.-Y., & Picon, M., 1986, A La Recherche des Fours d’Amphores. In J-Y. Empereur & Y. Garlan (eds.), Recherches Sur Les Amphores G recques. Actes du collo ue international organis par le Centre national de la recherche scientifi ue l ni ersit de Rennes II et l cole fran aise d th nes th nes eptembre 103-126. Athens, Paris. Empereur, J.-Y., & Picon, M., 1998, Les atelier d’amphores du Lac Mariout. In J.-Y. Empereur (ed.), Com merce et artisanat dans l le andrie hell nisti ue et romaine. ctes du collo ue d th nes organis par le C R le Laboratoire de c ramologie de Lyon et l Ecole fran aise d th nes d cembre 75-88. Athens, Paris. Empereur, J.-Y., & Picon, M., 1992, La Reconnaissance des Productions des Ateliers Cérmiques: L’ Exemple de La Maréotide. E trait des Cahiers de la C rami ue gyptienn 3: 145-152. Flaux, C., forthcoming, Geomorphological survey of the south-west shores of Lake Mareotis. In L. Blue & E. Khalil (eds.), A Multidisciplinary Approach to Alexandria s Economic Past the Mareotis Case tudy. Oxford. Hopkinson, D.J., 2007, An Assessment of G reco-Roman Lakeside Agriculture and Trade in the Mareotic Region of Egypt. Unpub. MA diss., University of Southampton. Khalil, E.K.H., 2005, Egypt and the Roman maritime trade: a focus on Alexandria. Unpub. PhD thesis, University of Southampton. McCann, A.M., Bourgeois, J., Gazda, E.K., Oleson, J.P., & Will, E.L., 1987, The Roman Port and Fishery of Cosa a Center of ncient rade. Princeton. Murray, G.W., 1951, The Egyptian Climate: An Historical Outline. The G eographical J ournal 117.4: 422-434. Oleson, J.P., 1984, G reek and Roman Mechanical WaterLifting e ices he istory of a echnology. London. Oleson, J.P., 2000, Water-Lifting. In Ö . Wikander (ed.), andbook of ancient ater technology, 217-302. Leiden

C onclusion The archaeology of Mareotis Island consists of two settlements which appear to be concerned with agricultural, industrial and water management activities as well as lacustrine focused features which may indicate that fishing activities or lake transportation services were conducted from the island. The island represents a largely unexplored archaeological asset that has the potential to give insights into daily life in a non elite community; when looked at in conjunction with other known sites in the area it may provide an opportunity to begin to address settlement distributions and social organisation and make comparison between the lake region and those communities on the Delta, and perhaps archaeological findings from the ayum or Nile Valley, where the most studied examples of agricultural settlements are known. Such comparisons with sites from the Egyptian interior may prove to be less than representative of ordinary Egyptian practices where the majority of the population was located on the Nile Delta (Bagnall 2007: 227). There is a great potential for future research on Mareotis Island and the broader Mareotis region and these provisional findings only begin to scratch the surface of what the island and Mareotic region in general has to offer. Bibliography Bagnall, R.S., 2001, Archaeological Work on Hellenistic and Roman Egypt, 1995-2000. American J ournal of Archaeology 105. 2: 227-243. Ball, J., 1942, Egypt in the Classical Geographers. Cairo. Brooks, C.E.P., 1949, Climate through the ages. New York. De Cosson, A., 1935, Mareotis: Being a Short Account of the History and Ancient Monuments of the Northestern esert of Egypt and of Lake Mareotis. London. El-Fakharani, F., 1974, The “Lighthouse” of Abusir in Egypt. ar ard tudies in Classical Philology 78: 257272. 44

D. HOPKINSON: MAREOTIS ISLAND SURVEY le Laboratoire de c ramologie de Lyon et l Ecole fran aise d th nes d cembre 27-36. Athens, Paris. Rodziewicz, M.D., 2002, Mareotic Harbours. In C. Déco bert (ed.), Alexandrie Médiévale 2. Etudes Alexandrines 8: 1-22. Schiøl er, T., 1973, Roman and Islamic Water-lifting Wheels. Denmark. Strabo, G eography, H.L. Jones (transl.), 1949. London. Survey of Egypt (27/135) 1:100,000 Map; Sheet 88/48 “El Ghayata”. Survey of Egypt/Ministry of Finances, Cairo. Tomber, R., & Thomas, R.I., forthcoming, Pottery from the Lake Mareotis Research Project. In L. Blue & E. Khalil (eds.), A Multidisciplinary Approach to Alexandria s Economic Past the Mareotis Case tudy. Oxford. Warne, A.G., & Stanley, D.J., 1993, Late Quaternary Evolution of the Northwest Nile Delta and Adjacent Coast in the Alexandria Region, Egypt. Journal of Coastal Research 9.1: 26-64.

Oliver, F.W., & De Cosson, A., 1938, Note on the Taenia Ridge – with especial reference to quarries, sites and an ancient road between Alexandria and Abu Sir. ulletin de la oci t Royale d rch ologie d le andrie 32: 163-175. Rodziewicz, M.D., 1983, Alexandria and District of Mareotis. Graeco rabica 2: 199-216. Rodziewicz, M.D., 1990, Taenia and Mareotis: Archaeological Research West of Alexandria. Annual of the Egyptian Society of G reek and Roman Studies 1: 62-78. Rodziewicz, M.D., 1998a, From Alexandria to the West by Land and by Waterways. In J.-Y. Empereur (ed.), Commerce et artisanat dans l le andrie hell nisti ue et romaine. ctes du collo ue d th nes organis par le C R le Laboratoire de c ramologie de Lyon et l Ecole fran aise d th nes d cem bre 93-103. Athens, Paris. Rodziewicz, M.D., 1998b, Classification of Wineries from Mareotis. In J.-Y. Empereur (ed.), Commerce et artisanat dans l le andrie hell nisti ue et romaine. ctes du collo ue d th nes organis par le C R

45

LAKE MAREOTIS: RECONSTRUCTING THE PAST

46

M. HAGGAG: EXCAVATIONS AT MAREA/PHILOXENITÉ

The City of Marea/Philoxenit . Reflections on the Alexandria University Excavations, 1977-1981 Mona Haggag

In 1977 Alexandria University patronized an expedition under the direction of the late Prof. Fawzi El-Fakharani to search for the ancient city of Marea, mentioned by different classical authors as the capital of the Mareotic Nome during the late Pharaonic period (Thucydides I.68; Herodotus II.30; III.12; 15; IV.161; Diodorus I.68.).

Sadek concluded that all the remains identified were not earlier than the 4th-5th centuries AD (Sadek 1978: 67-80). All trial trenches proved to have remains of stone buildings dating to the Byzantine period. Pot sherds were entirely of Byzantine manufacture (Shahin 1983). The same is true for the architectural material, construction methods and techniques. These results raised a major question about the location of the pre-Byzantine town mentioned in classical literature (Thucydides I.68; Herodotus II.30; III.12; 15; IV. 161; Diodorus I.68). . The Pre-Byzantine Remains (Map 1) By exploring the area some 5 km west of the Byzantine remains, earlier remains of Marea proved to exist beyond the main perimeter of the Byzantine site. A group of harbour installations were found on an island to the west.1

Mahmoud Bey El-Falaki located the site of Marea some 45 km west of Alexandria on the southern shores of the western extremity of Lake Mareotis near El-Hawaria (ElFalaki 1872: 85, 86, 91, 96, 97, 101; see other Haggag chapter in this volume . El- alaki based his identification of the site according to its location on the map of Ptolemy the Geographer IV,5. 16.17 . This location is confirmed by Ball (1942: 25-30), Breccia (1922: 337), De Cosson (1935: 131-135), Rowe (1954: 128-145) and Fraser (1972: 143-146) due to four visible jetties which still project into the lake today. The jetties are associated with a broad ridge of hard limestone extending from east to west across the town parallel to the southern coast of the lake and bordered, from both north and south, by calcarenite land.

The ‘ Pr’ -Shaped Harbour (Fig. 1, Map 1) The main harbour of the island is a three sided structure projecting into the lake basin about 30.5 m wide and 57 m long. It is built of undressed blocks of limestone of Cyclopean size. A number of the upper blocks in this structure have drilled cylindrical holes which were either bollard holes or mooring rings (Fig. 2). According to the traditions of ancient Egyptian harbour architecture, ports of rectangular plan in the shape of the sign “pr” were designed for harbours facing the open seas. Earlier examples existed in the seaport of Alexandria (Jondet 1916: 13-20; El-Fakharani 1991: 25), the most conspicuous of which is the one inside the Western Harbour mentioned by Strabo as the Kibotos (G eography, Xvii, I, 10), another one is associated with the Island of Antirrhodes which lies inside the Eastern Harbour of Alexandria discovered by Goddio (Goddio, et al. 1998: 21-29, pl. I). This feature consists of two rows

The Surve y As part of the 1977 campaign and in collaboration with the University of Guelph, Canada, Mohamed Sadek conducted a proto-magnetometer survey at the site, concentrating on the shoreline area from which the four jetties projected. Data was collected in lines parallel to the lakeshore from west to east in a grid of 800 m long and 100 m wide at intervals of 4 m. Soundings were taken in the area west of the first quay, and some trial trenches were investigated in three locations on the waterfront near the jetties. The trenches went deep into the bedrock which was found at intervals of 4.6-6.00 m below the present ground surface. Fig. 1: K ibotos harbour structure on the south shore of Lake Mareotis ( Site 9 – L ake Mareotis Research Proj ect) ( photo M. Haggag) .

1. It is worthy of notice here that Gauthier, referring to some hieroglyphic text, stated that the ancient city of Marea was located on an island in the lake (Gauthier 1925-1931: vol. III, 53-54).

47

LAKE MAREOTIS: RECONSTRUCTING THE PAST Map 1: Marea in Lake Mareotis ( V. Atef) .

Fig. 2: K ibotos harbour structure, mooring ring ( photo M. Haggag) .

Fig. 3: Rock-cut K ibotos harbour structure at Ikingi Mariut ( photo M. Haggag) .

48

M. HAGGAG: EXCAVATIONS AT MAREA/PHILOXENITÉ of stakes or piles approximately parallel, running southwest to north-east, along the axis of the main branch of the island, that forms a rectangular shaped-harbour against the island’s shoreline. A third one, hewn in the rock at Ikingi Mariut, about 20 km west of Alexandria (Fig. 3), was identified by El- akharani2 (2002: 203-208).

to a dromos. Another similar dromos on the opposite side leads to a funerary chamber. The ceiling is carved to imitate wooden beams in the manner common to Pharaonic cemeteries. False doors were also carved in the walls with a shaft in front of each one leading to the burial chamber below. All tombs have been looted. The only remains were a few local pointed bottom jars which are likely to be Late Pharaonic in date (El-Fakharani 1983: 176-178).

The Fort (Map 1) To the south-west of the pr-shaped port remains of a substantial structure built on a monolithic platform about 21 m wide and 24 m long was uncovered. The structure has an L-shaped plan. The walls seem to have consisted of rough core masonry with outer casing walls. The blocks of the outer casing are carefully cut and fitted together in oblique joints without using any mortar except for a foliage layer of non reddish mortar used in some of the courses to facilitate the sliding of such gigantic sized blocks. These external walls are slanting inward at an angle of about 40 degrees, and are based on larger foundation blocks with squared edges. The building is approached from the south by means of a ramp of dressed and carefully fitted stone blocks (Haggag 1984: 277-280). The use of rough masonry for the core while the huge blocks of the facing are very carefully polished, the oblique joints of the blocks, the use of such a thin layer of mortar, the inward slanting of the outer walls as well as the general plan of the building, are clear indications to a Late Pharaonic date. The building is interpreted as the fort of Psammaticus III (525 BC), from which his son Inaros sailed with his army, aided by some Athenian troops (Thucydides I.104), against the Persians in 465 BC, in an unsuccessful attempt to liberate Egypt.

The Byzantine Harbours (Fig. 4, Maps 1 & 2) The focal point of the town to the east of the Pharaonic/ Ptolemaic remains is the Byzantine harbour area which extends 2 km along the lake shore with four well preserved jetties projecting into the lake at an average height of nearly 2 m above the lakebed. The area comprises three harbours marked by the jetties and a natural promontory namely the west port, the middle and the east port. Along the shore there is a stone built seawall occasionally interrupted by either steps or stone ramps leading directly down to the water. The wall extends some 500 m with an average width of 1 m. At the outer extremity of the western quay of the middle port, there are a number of circular foundations about 4 m in diameter. These features are assumed to have been the base of a light beacon or lighthouse that would have aided navigation at night. Some mooring features can be detected on the upper course of this jetty in the shape of cylindrical bollard holes (Petruso & Gabel 1983: 63). The East Port (Fig. 5, Map 2) Although the east port was completely buried under the sands of the lake which by the time of excavation had dried out, digging revealed that it differs from the other ports of Marea in both shape and function. It is located between the promontory and a north-easterly island which has its own landing place formed by a short artificial dyke to the north-east and a natural promontory to the northwest. This northern island is connected to another smaller one to its south-west. A third quay connects this island to the promontory forming a wide entrance to the eastern port, at least in its first stage of construction. In the second stage, the shape, size and function of the port seem to have been modified. Almost 100 m of the quay that extended from the eastern small island to the promontory were removed and an extension to this quay constructed to the south. It was located parallel and adjacent to the eastern shore of the promontory and was about 200 m in length. Thirty metres before reaching a third island to the south, the quay ends forming an opening between its southern extremity and this small island. This opening provides an exit or entrance to the east port by way of the southern body of the lake, which extended considerably inland at this point. By this modification, the very long quays of the east port became totally isolated from both the mainland and the promontory. Accordingly, the link between the anchored ships with their cargoes and the town was severed. This fact is very important as far as the function of the port is concerned.

The Ptolemaic Q uay (Map 1) To the east of the island, a stone built jetty about 104 m long extends into the lake. Near its southern end, the jetty inclines to the east at an angle of about 175 degrees to extend a further 20 m in length. It is constructed of local regularly sized rectangular blocks, arranged in alternate courses of headers and stretchers (Haggag 1984: 280-283). The fine workmanship, medium size nature of the blocks as well as the ashlar masonry construction style with its vertical joints, are indications of a Ptolemaic date for the quay. The C emetery (Map 1) To the southwest of the island on the limestone ridge, the team came upon a group of burials that have nothing in common with Byzantine tomb types. The cemetery consists of some anthropoid pit tombs. A square shaft tomb is carved into the limestone ridge to a depth of 6 m. Loculi with gabled ceiling are cut either side of the shaft. A nearby chamber tomb is composed of a sloping passage hewn into the rock that ends with six steps leading to an open court. An opening cut in one side of the court leads

2. For detailed discussion of this pr-shaped ports of Pharaonic date as well as examples, see Haggag 1984: 263 ff. El- akharani 1991: 25, fig. 1,2.

49

LAKE MAREOTIS: RECONSTRUCTING THE PAST Map 2: Marea’s public buildings on the lake shore ( V. Atef) .

Fig. 4: Byzantine harbour of Marea ( photo M. Haggag) .

depth of 0.25 m below the pavement, a square drainage canal, coated with a thick layer of plaster and covered by a row of huge rectangular blocks, extends across the decumanus. This drains into a similar canal that extends under an unpaved side street and from south to north crossing the decumanus at a right angle. Thus, the drainage waters pour directly into the lake (El-Fakharani 1983: 178-179).

El-Fakharani’s interpretation of the function of the east port 1983: 181 is that it worked as a one way traffic harbour for transit navigation: so that seafaring ships carrying goods from different parts of the Mediterranean could enter the east port via its northern opening, their cargoes were to be unloaded on the long quays in order to be reloaded on board river boats to transport goods to the Nile and vice versa.

The Shops (Map 2) The arcade of the decumanus lines a group of 12 constructions (Fig. 7), with different ground plans. The buildings are located side by side with doors that opened out onto the sheltered arcade. They seem to have been built to function as shops with residential quarters for their owners. Remains of white plaster coating are still visible on some of the walls. An upper storey for each of the shops is confirmed by the existence of stairs at the rear of some of the shops and also in some of the back rooms (El-Fakharani 1983: 178-179; Soleiman 2004: 145-164).

The Public Buildings (Map 2) Many, if not most, of the public buildings of Byzantine Marea were mainly erected along the shoreline. Parallel to the lake shore in front of the harbour installations runs the main arcaded street of the city. This decumanus (Fig. 6) extends eastward until it reaches the natural promontory which forms the eastern part of the middle port. It is 10 m wide marked on its southern side by an arcade of which some of the bases of columns can still be seen in situ. The street is paved by big rectangular limestone blocks. At a 50

M. HAGGAG: EXCAVATIONS AT MAREA/PHILOXENITÉ The Basilica (Fig. 8, Map 2) To the west of the decumanus, at the intersection of the western side street, a huge complex of structures came to light during excavations (Sadek 1992: 549-554). The structures consist of two conjoined apsidal basilicas furnished with what seems to be public bathing facilities for sanitary purposes, similar to those discovered and interpreted by Grossman at St Menas (Grossmann 1986: 12-13; Haggag 1984: 284-289). The Bakery At the promontory which forms the eastern extremity of the middle port, a huge building with an area of about 700 m 2 was uncovered. A large granite mill inside one of the building’s multiple rooms indicates the relation of this building with grain grinding processes. Inside one of the rooms, a marble tile with a carved Greek cross was uncovered. In a compartment annexed to one of the rooms in this building, a clay jar containing a hoard of coins was also found. Courses of Byzantine burnt bricks are inserted in the stone masonry of the walls. The building is surrounded by an external wall supported with buttresses typical of Byzantine methods of construction (Moussa 2002: 479480). Fig. 5: ( above) Plan of the eastern harbour of Marea, after El-Fakharani 1977 ( thanks to El-Fakharani) .

Fig. 6: Decumanus between the eastern and middle ports of Marea ( photo M. Haggag) .

Fig. 7: Plan of shops off the main decumanus ( thanks to El-Fakharani) . 51

LAKE MAREOTIS: RECONSTRUCTING THE PAST Fig. 8: Basilica, Marea ( photo M. Haggag) .

Fig. 9: Enigma Buliding, Marea ( photo M. Haggag) .

The Enigma Building (Fig. 9, Map 2) Next to the western edge of the promontory at its meeting point with the lakeshore along the middle port, an enigmatic building has been uncovered. The building stands completely in the water. It is composed of podia and walls leaving canals 0.75 m broad between them. Some of the canals run in parallel lines with the coast, while others run at right angles with it and the other canals. The entire building is sliding towards the water leaving the outer parts of the podia under water. The canals were studded with small artefacts. The finds, which were mostly intact, included nearly one thousand pieces of small bronze coins, about 200 jugs of red or yellowish clay, St Menas ampulae, other ampulae with crosses, small statuettes of horsemen, fishing hooks and broken glass bottles. This unusual plan made scholars differ in their interpretations of the building’s function. Some suggestions included its possible function as a dry-dock owing to its distinct slope into the lake water Sadek 1978: 70 , but the huge amount of finds inside the canals stands against this hypothesis. Another explanation is that it served as a fishery ikkos Lianos pers. comm.). While the discoverer tended to interpret it as a building for votive offerings in honour of both St Menas

and St Marcus, due to the large quantity of coins, ampulae and other artefacts that were found inside (El-Fakharani 1977: 16; Soleiman 2004: 121-138). Boston Field School In 1979, the Boston University’s Summer Term and Study Abroad Program supported an archaeological field school at the site of Marea. Their work lasted for three consecutive seasons, one month each year.3 In addition to the fruitful student training, the mission made new and important discoveries, the most important of which was the Diolkos of Marea. Along the shore between the middle and western jetties, the team of Boston University came upon remnants of a dockyard (Petruso & Gabel 1982: 12). It consists of two sloping walls, each is over 7 m wide and 20 m long,

3. The team was directed jointly by Profs. El-Fakharani & Gable in collaboration with Petruso and the architect, Boyd. The author of this article had the honour of participating in this field school first as an undergraduate trainée and then as an assistant archaeologist.

52

M. HAGGAG: EXCAVATIONS AT MAREA/PHILOXENITÉ Fig. 10: Winery excavated by El-Fakharani 1983 ( photo M. Haggag) .

pletion of the excavation of the building, it proved to be a huge double peristyle that occupied an area of more than 1,500 m2. Both El-Fakharani (1983: 184-186) and Rodziewicz (1988a: 175-178, 1988b: 267-276) asserted that this discovery represents the first and largest peristyle built for residential purposes to be discovered in Egypt hitherto.

emerging from massive stone foundations going down to the bedrock, to a depth of about 2 m. The runners slip into a central aisle that lies between the jetties. This aisle has some sort of drainage facilities in the form of holes cut into the floor in order to drain water and keep the floor dry. The entire building slides down into the lake basin at a gradient of approximately 1:16 which allows for the use of manpower and log rollers to hall ships out of the water (Petruso & Gabel 1983: 76).

In the middle of the town, some Byzantine dwellings were uncovered. Each is composed of two rooms. Byzantine local pottery ware were scattered inside. Muslim burials were dug inside the room. Green glazed pottery sherds of the Islamic period were found as well as one piece of a Fatimid coin.

The C ity’ s Residential Area Far from the harbour area to the south of the limestone ridge and south of the modern highway leading to Borg ElArab, in the area now called “Hawariya”. During a second season conducted by the Alexandria University expedition team in 1977, various buildings were uncovered in this region which seem to relate to residential and daily life activities. The Peristyle C omplex (Map 1) Adjacent to the modern highway (Map 1), some 2.5 km to the south of the harbour, a peristyle court which leads to a variety of rooms, was uncovered. One of the rooms has a staircase leading to an upper floor of the building. It seems that the building went through two stages in its construction the first stage is indicated by a layer of white plaster coating on the façade. The second stage is marked by the addition of a massive thick walled façade with two windows which have tunnels underneath leading to watercourses that are connected to cisterns. A modern ditch for irrigation water that was dug through the rooms halted a complete investigation of the remains. In 1980, Rodziewicz4 replaced that ditch with a pipe in order to continue uncovering the rest of the building. On com-

Wineries (Map 1) About 200 m to the southwest of the peristyle building, in the middle of the town, a big elaborate winery (Fig. 10) was uncovered (El-Fakharani 1983: 182-184). Another identical in layout but smaller than the first, was also discovered 2 km to the southwest (Arafa 1985: 78-80). Both are of the type known in other places in the Mareotic region.5 They represent the archaeological testimony for the excellence of the Mareotic wine praised by the classical authors (see Dzierzbicka this volume). Each of the two buildings comprises two rooms separated by a low screen wall for squeezing grapes. ne is smaller than the other. Both are entirely coated by four layers of red plaster to prevent any possible seepage of the juice. In the center of the smaller room is a raised round base covered with plaster possibly to support a movable squeezer or a hand press. By means of a lion headed spout, the juice from both rooms pours into a large square basin which is dug in the ground, with another smaller and deeper basin dug in its floor for collect-

4. Rodziewicz was then a consultant for the Egyptian Antiquity Service (EAO) and was in charge of the restoration of the area.

5. A third smaller one that lies to the south of the big winery has been reported by Rodziewicz, see Rodziewicz 1998a: 27-36.

53

LAKE MAREOTIS: RECONSTRUCTING THE PAST ing residues. The inner surface of the big basin’s walls is scratched to allow the plaster coating to hold firm. The inner sides of the basin are surrounded by a shelf approached from all sides by a flight of steps. ther steps lead to the basin s floor. It is thought that this shelf is designed to support wooden beams that held a piece of cloth for filtering the juice poured from the lion shaped spout. In the middle of the north side above the spout, there is a small podium with two funnels pierced each by a hole opening at the basin. El-Fakharani’s hypothesis is that these funnels acted as measures for adding certain amounts of some aromatic flavours that produced the famous taste of Marea s wine6 (1983: 182-184).

town Ball 1942: 114, 117, fig. 17, pl. II, III . More than two and a half centuries after Ptolemy, we hear nothing about Marea. From the 5th century onwards, Byzantine Marea appears. During this time, the city has to play another vital role in Egypt’s history. The town in which the Martyr Abu Mina was buried lies about 15 km to the south of Marea. The sanctuary of St Menas was famous for its miraculous healing capacities. The importance of the area increased gradually until it reached its climax at about the end of the 5th and beginning of the 6th centuries (Kaufmann 1910: 103). Flux of pilgrims and patients from all over the Roman provinces headed to the place seeking cures for their ailments Grossman 1986: 12-13 . lasks ampulae filled with the holy waters of Abu Mina were specially made for pilgrims to take with them on their way back home as a blessing.8

The Byzantine Tomb (Map 1) To the north of the railway line, and at the southern boundaries of the inhabited area of Byzantine Marea, lies a rock cut chamber tomb (El-Fakharani 1977: 19-22). The entrance to the tomb is an arched doorway with basalt jams and lintel. It opens onto a passage with a staircase leading to a square mourning chamber. The ceiling of the passage is barrel vaulted with a helicon vault covering the part which turns to the east towards the arched entrance of the mourning chamber. The latter is surrounded by a stone bench and covered with a cross vault with pendentives. Three burial chambers with cross vaulted ceilings are cut into the walls of the tomb, forming a trefoil plan for the Hypogeum.7

The most suitable and more convenient way for those who are coming from both the Mediterranean and the Nile is to reach the lake harbour of Alexandria and navigate the lake to the nearest harbour before taking the land route to St Menas. In this respect, Marea is the nearest point. At that time, Marea’s earlier harbours were isolated on an island and pilgrims had to be ferried to the southern coast of the lake in order to take the caravan route to the sanctuary. To solve this problem, the old town of Marea was shifted 2.5 km to the east of the earlier harbour area, where the line of the ridge inclines far from the lake shore leaving a larger area including a promontory, a bay and several islands in front of the shoreline, that combined offer a naturally sheltered harbour. This area seemed more suitable to quickly establish the new harbour facilities and provide accommodation for pilgrims (Rodziewicz 1998b: 101-103). Other factors may have contributed to the shifting of the site, the most important of which is the changing lake levels, a fact indicated by the different levels of rising water during the flood seasons over many centuries.

The N ame of the C ity More than two centuries before the establishment of Alexandria, Marea, according to Herodotus (II.30, IV-161; III.12, 15), Thucydides (I.104) and Diodorus (I.68) functioned as a strategic staging post on Egypt’s northern frontiers. After the establishment of Alexandria, the former military role of Marea appears to have come to an end and the city takes on a more civic and commercial role (Rodziewicz 1998b: 101-103). Excavation works yielded neither weapons nor fortifications of neither Ptolemaic or Roman date. At the same time, classical authors, such as Athenaeus (I.33), Vergil (G eorgics, II.91-92), Horace (O des, I.37) and Columella (De Rustica, III.24), give us glimpses of the reputation of Mareotic wines. Marea became an agricultural centre as well as an intermediate station for goods moving between Alexandria and the Nile Valley (Rodziewicz 1983: 199-208). After the time of Strabo (XVII.I.14) who spoke of the city’s wealth in papyrus, bean and vine plantation, the city seems to have shrunk to a small village as stated by Athenaeus (I.33) and as could be inferred from Ptolemy the Geographer who mentioned it as “Palae-Marea” by which name he indicates the existence of the earlier Pharaonic

This leads us to the question that Rodziewicz 2003: 27-39 first raised: Was this new site of Byzantine Marea the city of Philoxenité?9 Philoxenité was built by the Praetorian Praefect Philoxenus upon orders of the Emperor Anastasius (AD 491-518) in order to serve the needs of pilgrims on their way to the shrine of St Menas. Rodziewicz counted on the Coptic Encomium in praise of St Menas, first published by Drescher (1946: 35-72; 126-149), to reach the conclusion of interpreting the later site of Byzantine Marea as the city of Philoxenité. According to the text of the Encomium, Drescher placed Philoxenité on the shores of the main basin of the lake, somewhere close to Kom Truga 35 km to the east of the Holy Shrine. Rodziewicz believes

6. For more detailed interpretations, see Rodziewicz 1998a: 27-36. 7. Various rooms on top of the Hypogeum and its vicinity were uncovered by a team of the Supreme Council of Antiquities SCA later in 1987/88. They represent a complex of funerary chapels for the tomb (Soleiman 2004: 283-286).

8. Just as Muslims nowadays do with the waters of the holy well of Zamzam at Mecca. 9. Before tackling this problem, it is worth noting that Rodziewicz in addition to his responsibility for the restoration works carried on at the site of Marea, made additionally impressive discoveries at the site.

54

M. HAGGAG: EXCAVATIONS AT MAREA/PHILOXENITÉ Bibliography

that the site of the Byzantine harbours at Marea accurately fit the location of the landing place of pilgrims seeking accommodation, food and beverage, before taking the short caravan route to the shrine.10 In this respect, it is important to point out that the Encomium, which is the sole source we know of to date about the city of Philoxenité, includes the following passage:

Ancient Sources Athenaeus, Deipnosophists, C. Burton (transl.), 1927. London. Augusti Meinekii, 1849, Stephani Byantii, Ehnicorum quae supersunt, I. Berlin. Columella, O n Agriculture, H. Boyed (transl.), 1931. London. Diodorus Siculus, Library of History, C.B. Welles (transl.), 1963. London. Herodotus, Histories, H. Carry (transl.), 1912. London. Horace, The O des, C.H. Bennet (transl.), 1925. London. Ptolemy, G eographica, E.L. Stevenson (transl.), 1932. New York. Strabo, G eography, H.L. Jones (transl.), 1949. London. Thucydides, History of the Peloponnesian War, C.F. Smith (transl.), 1969. London. Virgil, G eorgics, J. Jackson (transl.), 1908. Oxford.

...When it was finished he gave it the name Philoxenité… He ordered to build water stations, where he placed water jars, every ten miles along the route between the hospices and the shrine in order to serve the needs of these multitudes” (Drescher 1949: 15-16).

Yet, the distance between the Byzantine harbours of Marea to the sanctuary do not exceed 15 km, that is less than 10 miles, a distance that can not take more than a few hours if riding on camels and half a day if walking on foot. Then, why did the Praefect order the building of water stations every ten miles along this route?11 Rodziewicz’ hypothesis is mainly based upon believing that there is nothing evident at the site to be dated earlier than the Byzantine era. This assumption has been proved to be incorrect through what was discovered on the island and the ridge by the Alexandria University expedition. More recently our Polish and French colleagues have also uncovered some pre-Byzantine remains (Szymanska & Babraj 2001: 37-42, 2002: 47-58 see Babraj & Szymańska this volume Pichot 2004). Moreover, if the newly established Byzantine centre had the name of Philoxenité, why did a Christian author, who lived and wrote after the time of Anastasius and his Prefect, Stephanus of Byzantium (Augusti Meinekii 1849: 432), retain the old name of Marea in his writings?12

Secondary Sources Arafa, D., 1985, Wine Production in Egypt from Alexander to the Arab Conquest: As Compared to the G raecoRoman World. Unpub. MA diss., Alexandria University. Ball, J., 1942, Egypt in Classical G eographers. Cairo. Botti, G., 1902, Studi sul III Nomo dell’egitto inferiore E piu specialmente sulla regione Mareotica. Bulletin de la Sociéte Archéologique d’Alexandrie 4: 41-84. Breccia, E., 1922, Alexandrea ad Aegyptum. Bergamo. De Cosson, A., 1935, Mareotis: Being a Short Account of the History and Ancient Monuments of the NorthWestern Desert of Egypt and of Lake Mareotis. London. Drescher, J., 1946, Apa Mena: A Selection of Coptic Texts Relating to Saint Menas. Cairo. Drescher, J., 1949, Topographical Notes for Alexandria and Distric. Bulletin de la Sociéte Archéologique d’Alexandrie 38: 13-20. El-Fakharani, F.A., 1977, A Preliminary Report on the First Season of Excavation at Marea. Excavation Report presented to the Antiquities Department, Egypt. Cairo. El-Fakharani, F.A., 1983, Recent Excavation at Marea in Egypt. In G. Grimm, H. Heinen & E. Winter (eds.), Aegyptiaca Trevernsia II, Das Römisch-Byzantinische Ägypten. Akten des internationalen Symposions 26.-30. September 1978 in Trier, 175-186. Mainz. El-Fakharani, F.A., 1991, The Kibotus of Alexandria. Studi Miscellanei 28: 21-28. El-Fakharani, F.A., 2002, The Pharaonic Port on the Mediterranean: Its Shape, Development and Importance. In Z.A. Hawass & L.P. Brock (eds.), Egyptology at the a n of the enty first Century Proceedings of the 8th International Congress of Egyptologists, Vol. 2: 203-208. Cairo. El-Falaki, Mahmoud Bey, 1872, Memoire sur l’antique Alexandrie, ses faubourgs, ses environs. Copenhagen. Empereur, J-Y., 1998, Alexandria Rediscovered. New York.

Finally, one would say that it is not unusual for a certain town in antiquity to have more than one name over the different periods of its history. Rhacotis/Alexandria, Thonis/ Heracleion, are very distinctive examples in this respect. It seems reasonable that the old town of Marea shifted eastward to build new facilities suitable for the new role the city had to play.13 Whether the new site took a new name or not is a question that cannot be answered yet with any certainty. In this respect, further readings of the original Encomium are required as well as any other material evidence that may appear through investigations currently being undertaken by different missions at the site.

10. Haas and Empereur share Rodziewicz’s opinion, see Haas 1997: 349; Empereur 1998: 229-239. 11. Whether this is due to miscalculations of the wording of the text, as Rodziewicz mentions, it is something that cannot be judged unless further readings of the original Encomium are undertaken, see Rodziewicz 2003: 27-39. 12. It is not evident that the various citations of Marea in the ancient literature refer to exactly the same spot as in the case of Procopius: Iacobus Haury 1913: vol. III 2, 171. We find some confusion in the use of “Marea” and “Mareotis” used to describe both the town and the lake or sometimes the whole lake district (Petruso & Gabel 1980: 1-27). 13. In 1902 Botti stated that there were two towns named Marea, the ancient and the new (Botti 1902: 73-75).

55

LAKE MAREOTIS: RECONSTRUCTING THE PAST Fraser, P.M., 1972, Ptolemaic Alexandria, Vol. I. Oxford. Gauthier, H., 1925-1931, Dictionnaire des nomes géographiques contenus dans les textes Hiéroglyphiques, Vol. III. Le Caire. Goddio, F., Bernard, A., Bernard, E., Darwish, I, Kiss, Z., & Yoyotte, J., 1998, Alexandria: The Submerged Royal Q uarters. Oxford. Grossmann, P., 1986, Abu Mina: A G uide to the Ancient Pilgrimage Center. Cairo. Haas, C., 1997, Alexandria in Late Antiquity: Topography and ocial Con ict. Baltimore. Haggag, M., 1984, Ports in Ancient Egypt till the Arab Conquest. Unpub. MA diss., Alexandria University. Haggag, M., 2002, Two Religious Buildings at Byzantine Marea. In Z.A. Hawass & L.P. Brock (eds.), Egyptology at the a n of the enty first Century Pro ceedings of the 8th International Congress of Egyptologists, Vol. 2: 284-289. Cairo. Jondet, M.G., 1916, Les Ports Submergés de l’ancienne Ile de Pharos. Le Caire. Kaufmann, C.M., 1910, Die Menasstadt und der national heiligtum der altchristlichen Ägypter, I. Leipzig. Moussa, .S., 2002, uelques aspects de la vie quotidiè nne représentés á Marea Byzantine. In Z.A. Hawass & L.P. Brock (eds.), Egyptology at the Dawn of the enty first Century Proceedings of the 8th International Congress of Egyptologists, Vol. 2: 478-486. Cairo. Petruso, K.M., & Gabel, C., 1980, Marea: An Environmental and Cultural Study at Lake Maryut, Lower Egypt: A Research Prospectus. Boston U niversity African Studies Center Working Papers 25: 1-27. Petruso, K.M., & Gabel, C., 1982, Marea: A Byzantine Port in Northern Egypt. Boston U niversity African Studies Centre, Working Papers 62: 1-23. Petruso, K.M., & Gabel, C., 1983, Marea: A Byzantine Port on Egypt’s Northern Frontiers. Archaeology (Sept./Oct.): 62-77. Pichot, V., 2004, La fouille de l’Ile de Marea: Le site, Pros pections et campagnes de fouille: http://www.cealex. org/sitecealex/activities/terrestre/marea/genef.htm, CEAlex (accessed March 2008). Rodziewicz, M.D., 1983, Alexandria and District of Mareotis. G raeco-Arabica 2: 199-208.

Rodziewicz, M.D., 1988a, Remarks to the Peristyle House in Alexandria and Mareotis. Praktika, International Congress of Classical Archaeology in Athens, September 4t h -10th, 1983: 175-178. Athens. Rodziewicz, M.D., 1988b, Remarks on the Domestic and Monastic Architecture in Alexandria and Surroundings. In E.C.M. vanden Brink (ed.), The Archaeology of the Nile Delta: 267-276. Amsterdam. Rodziewicz, M.D., 1998a, Classification of Wineries from Mareotis. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athè nes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole franç aise d’Athè nes,11-12 décembre 1988: 27-36. Athens, Paris. Rodziewicz, M.D., 1998b, From Alexandria to the West by Land and by Waterways. In J.-Y. Empereur (ed.), Commérce et artisant dans l’Alexandrie héllénistique et romaine. Actes du colloque d’Athè nes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole franç aise d’Athè nes,11-12 décembre 1988: 95-102. Athens, Paris. Rodziewicz, M.D., 2003, Philoxenité - Pilgrimage Harbour of Abu Mina. Bulletin de la Sociéte Archéologique d’Alexandrie 47: 27-47 Rowe, A., 1954, A Contribution to the Archaeology of the Western Desert: II. Bulletin of the J ohn Rylands Library 36: 128-145. Sadek, M., 1978, The Ancient Port of Marea. Cahiers des É tudes Anciennes VIII: 67-80. Sadek, M., 1992, The Baths at the Ancient Harbour of Marea. The Sesto Congresso Internazionale di Egittologia, Vol. I: 549-554. Turin. Shahin, B., 1983, Local Pottery in Byzantine Egypt: A tudy of the finds at the City of Marea. Unpub. MA diss., Alexandria University. Soleiman, N.M.S., 2004, Marea: An Archaeological Study and the Manner of its Tourist Investment. Unpub. PhD diss., Alexandria University. Szymanska, H., & Babraj, K., 2001, Marea: First Interim Report, 2000. Polish Archaeology in the Mediterranean XII: 37-42. Szymanska, H., & Babraj, K., 2002, Marea: Second Interim Report, 2001. Polish Archaeology in the Mediterranean XIII: 47-58.

56

V. PICHOT: MAREA PENINSULA

Marea Peninsula: ccupation and Workshop Activities on the Shores of Lake Mariout in the Work of the Center d tudes Alexandrines CEAlex, C RS SR 3134 Val rie Pichot General Introduction to the Site of Marea The site, identified as Marea on the plate of Alexandria in the Atlas géographique d’Egypte1 (Description de l’Egypt 1809: Pl. 10 and then by Mahmoud Bey El- alaki in 1866 El- alaki 1872: 96 , is one of the rare examples of a harbour town on the shores of Lake Mariout that has essentially remained untouched since antiquity and accessible for archaeological studies. Situated some 40 km to the south-west of Alexandria, it stretches more than 25 ha along the southern shore of Lake Mariout at a point where

the width of the lake from north to south is less than one kilometre ig. 1 . Since 1977 the site of Marea has been the subject of archaeological excavations and those discoveries made prior to 2003 all suggested a somewhat late occupation of the site. The excavated remains, dating from the 5th to 7th centuries AD, indicate a harbour town of considerable capacity with large storage facilities as well as public buildings of high quality ig. 2 El- akharani 1983

Fig. 1: Map of Egypt: location of Lake Mariout and the site of Marea ( CEAlex Archives) .

1. Hydrographic map of Lower Egypt drawn in 1801-1802, Description de l’Egypte, Etat moderne, vol. I, Paris, 1809, plate 10.

57

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig. General plan of the site identified as Marea ( CEAlex Archives) .

Petruso & Gabel 1983 Sadek 1992 De Cosson 1935: 131 Rodziewicz 1983, 1998 Szymanska & Babraj 2008 .

will be more precise once we have the results of the ceramological and numismatic studies presently under way.

Since 2003, the Centre d Etudes Alexandrines CEAlex has been working on the peninsula situated some 100 m to the northeast of the ancient town of Marea: topographical surveys, prospecting on foot, geophysical examinations and archaeological excavations have all begun to reveal the general organisation of the peninsula and the existence of a large workshop quarter that extended over much of its surface.

An imposing building constructed of large blocks and hydraulic mortar stood on this causeway about 100 m before its junction with the peninsula. It is usually interpreted as a lighthouse or landmark. A jetty stretching 26 m into the lake is situated to the south-east of this igs. 3 & 7 . The ancient causeway is far from straight. riented south/north for less than 300 m, it takes a turn to the east to follow, firstly, a south-west/north-east direction then west/east to the point of the lighthouse or landmark. It then returns to the south-west/north-east and reaches a workshop quarter. The rather particular shape of the causeway and the jetty to the south-east of the landmark create a mooring basin that is relatively well protected from the prevailing winds.

General Organisation of the Peninsula ig. 3 riginally an island, access to the peninsula was possible either from the lake onto a pier to the north that allowed for boats to moor, or by land across a causeway that was some 5 m wide that connected with the mainland igs. 4a & 4b .

n the north of the peninsula there was a building over 100 m long also built of large blocks and hydraulic mortar ig. 3 . Its northern limit ends in a system of terraces and stairways cut into the rock and leading to a pier. Its general layout around a large rectangular courtyard might suggest a public building for commercial purposes concerned with trade and storage, though it might also be a rich villa. The presence of cistern s under this building is attested by a descending passage situated in the centre of the west wing. Structures in brick and hydraulic mortar visible on the surface of the east wing could be connected with the cisterns, or could be evidence of bath facilities. Surface surveys and drawings have not yet made it possible to understand the exact role and organisation of this building or the function of its associated hydraulic system.

In 2003 the first two excavation sectors were established upon this causeway. Sector 1 igs. 3 & 5 was opened on a very badly preserved part of the track. This excavation allowed us to reveal the presence of foundations of a structure probably associated with controlling movement on and off the causeway. Sector 2 igs. 3 & 6 was established at the point where the track disappears toward the southern part of the concession. The causeway effectively stops at this point and ends in a sort of platform made of numerous amphora fragments and mortar. The archaeological material found here was relatively homogenous and dates the construction of the causeway to between the 1st and 3rd centuries AD. The chronology of its construction 58

V. PICHOT: MAREA PENINSULA

Fig. 3: The peninsula of Marea ( CEAlex Archives) . 59

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig. 4: A) ( above) pier to the north of Marea peninsula, view from the north-west. B) ( right) ancient causeway connecting the peninsula to the mainland, view from the west. Photo V. Pichot ( CEAlex Archives) . Fig. 5: ( left) Ancient causeway under excavation of Sector 1, view from the east. Photo V. Pichot ( CEAlex Archives) .

Fig. 6: Ancient causeway under excavation of Sector 2, view from the north. Photo V. Pichot ( CEAlex Archives) .

Fig. 7: Landmark and j etty to the south of the peninsula, view from the south-west. Photo V. Merle ( CEAlex Archives) .

60

V. PICHOT: MAREA PENINSULA Occupation and Workshop Activi ties CEAlex is currently focusing upon the central part of the peninsula where a quarter developed and evolved during the Hellenistic and Roman periods. This was principally a workshop quarter the presence of which is attested from the 3rd century BC.

a later period, occupation re-adopts the north-north-west/ south-south-east orientation present during the Hellenistic period on the elevated part of the peninsula. Survey of the peninsula led to the discovery of numerous archaeological vestiges connected to light industry. Many zones with concentrations of slag and furnace walls, point out the presence of what were workshops, the greater part of which appear to have been connected to metallurgic activities. The numerous plots showing clear anomalies detected by the magnetic survey certainly correspond to areas of intense metallurgic activity. ther anomalies may possibly represent the presence of potteries and more certainly chalk furnaces from the later era 5th-6th centuries AD , as has been shown by the excavations.

Geophysical magnetic surveys undertaken on the central part of the peninsula have revealed the existence of three major orientations, subsequently confirmed by excavation, that correspond more or less to the layout of the area at different periods ig. 8 . In the Hellenistic period 3rd-1st centuries BC there are two orientations: one south-west/ north-east for the low-lying part of the peninsula and the other north-north-west/south-south-east for the elevated part. At the end of the Hellenistic period and the beginning of the Roman era, the major orientation used within the low-lying areas follows a north-north-east/south-southwest axis, while the north-north-west/south-south-east orientation appears to continue in the elevated areas. At

Metallurgic activities The excavations undertaken in Sector 4 see ig. 3 have revealed extensive furnace activity: extensive not so much in terms of production, which is still difficult to estimate,

Fig. 8: Results of the geophysical survey undertaken on the Marea peninsula. ( T. Herbich & CEAlex Archives) . 61

LAKE MAREOTIS: RECONSTRUCTING THE PAST but rather in the number of hearths discovered. In fact, in this sector where explorations both geophysical and visual suggested an empty terrain, except for a ditch situated to the western edge, some 40 hearths have been unearthed most of which are lying a mere 0.15 m beneath the surface and all are concentrated in the west and north-western part of the sector. f different sizes the size depending to a large part on the extent of destruction they are either circular in plan or occasionally rectangular. While certain examples still hold fragments of objects and/or slag, others are practically destroyed. They are not all contemporary, but they may have functioned in groups of two, three or four, several groups being in operation at the same time. The best conserved hearths were cleared when excavating rubbish ditch S4024 situated on the western edge of the sector. Its fill composed fragments of amphora, ceramics, bones, some fragments of faience objects, charcoal, some fragments of iron and seashell. The majority of the material unearthed was characteristic of the 1st centuries BC and AD, however, certain elements can be dated to an earlier Hellenistic period, e.g. a worked lamp of the 3rd century BC and amphora and ceramic fragments of the 3rd and 2nd centuries BC.

Fig. 9: Sector 4 - Hearths in FS4024, unde Photo V. Pichot ( CEAlex Archives) .

r excavation.

tic period occupation itself potentially related to workshop activities, can be broadly dated to the first centuries of the Roman era. Sector 3, presently under excavation see ig. 3 , was opened in 2003 on the western part of the island and was extended in 2007 into a zone of noticeable magnetic anomalies towards the east. Work on the western part was temporarily suspended to allow, firstly, for the completion of the study of Sector 4 and then the excavation of the eastern part of Sector 3. In 2003 two ensembles of buildings separated by an alley were unearthed in the western part of the sector ig. 10 . The ensemble oriented north-north-east/ south-south-west was disturbed in its north-eastern part by large constructions orientated on a north-north-west/ south-south-east axis. Each building is composed of small workshop rooms opening onto the alley, while at the back there was a large room used for storage or habitation. ne of the activities identified in this workshop quarter is polymetallurgical Cu and e . This is attested by the remains of hearths, flooring and occupation layers characteristic to this type of activity, but also by the discovery of refuse material connected to metallurgical activity slags, drips, hammerscales, unfinished objects, etc. .

This ditch S4024 was cut by several groups of hearths ig. 9 . Among these structures, two features are quite well preserved R4034 and R4035 . Although made up of hearths with different plans one is composed of rectangular hearths, the other of circular they are constructed in a similar fashion: a hearth of 0.30 m diameter to which is attached a much smaller but deeper hearth, the exact function of which remains to be determined. These features were constructed of clay in which fragments of ceramics and amphorae were placed to act as support. The structures are well constructed and the interior walls are perfectly smooth. Their fill is of charcoal, baked clay, small fragments of iron and small furnace slag. The debris discovered suggest that metal working activity was taking place, associated with the fabrication and/or repairing of small iron objects with the use, in certain cases, of copper alloys and lead. Associated with the hearths, traces of post-holes reveal the presence of light structures that could have been more or less permanent. Stake holes near the hearths, however, indicate the probable use of fireguards, windbreaks and/or walls marking out certain specific zones, e.g. for storage.

our workspaces are presently being excavated. The outline of walls marking out storage and circulation zones, have been recognised in three of these rooms. In Space 10 a series of re-laid beaten earth floors have been revealed beneath the last occupation level. According to the initial studies of the archaeological material, it seems that levels of workshop activity can be dated to the 1st century BC to the 1st century AD. A limited trench has been opened in the low-lying zone to the south of Sector 3. The water table was quickly reached at a little more than 1 m below the surface. Three successive floors and associated metallurgic activity layers lots of charcoal, numerous small metal drips dating to the 2nd to 1st centuries BC were revealed, as well as a wall whose orientation revealed the existence of an ensemble laid out on a north-west/south-west axis.

The remains of certain furnaces consisted of no more than the very bottom traces of stakes and posts, a few strips of beaten earth floors and a part of the first foundation course of a wall MR4040 oriented west/east set upon the bedrock. The vestiges of this sector are very worn down, hence the difficulty in envisaging the spatio-temporal organisation of these workshops. Without going into considerations of contemporanity of function and the length of occupation, and having yet to complete the study of the archaeological material, the important metal working activity in this zone, implanted upon the remains of Hellenis62

V. PICHOT: MAREA PENINSULA

Fig. 10: O verall plan of the western zone of Sector 3 ( CEAlex Archives) . 63

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig. 11: O verall plan of the eastern zone of Sector 3 ( CEAlex Archives) . Area Devoted to Cult Activity This occupation is contemporary with that of a construction ST300 partially unearthed in 2007 in the eastern part of Sector 3 igs. 11 & 12 . A paved way some 1.15 m wide situated to the north, leads to the entry of the building. The building2 is rectangular in shape approximately 10.35 m by 11.75 m, oriented north-north-west/southsouth-east along its longitudinal axis. The peripheral walls of an average thickness of 1.4 m and with concave external surfaces, enclose a space divided into several cells separated by walls that are 0.8 and 0.9 m thick.

The building consists of five different spaces. rom the outside one enters into a large rectangular room of 3.5 m by 7.55 m, divided into two by a pavement of slabs that cuts through the middle. This pavement is approximately 0.85 m wide, thus narrower than the opening of the doorway 1 m approx. , and is paved with irregular limestone slabs. It connects the entrance with the only door within the building that is pierced through the interior transversal cross-wall. Its slight irregularity and the fact that is deliberately off the axis of the room, leads one to consider the nature of the flooring that once lay on either side. There are few clues left to satisfy this query. Three fragments of slabs attached to the pavement at three different points along its line might indicate that the missing flooring was made up of the same materiel as the pavement itself, at least on the edges. The walls of this room were covered with a plaster painted to represent alabaster. The remains of another painted coating made to represent red, white and black marble facing might have belonged to a later stage or more probably to the decoration of a room on the upper floor. In front of the interior door, a rectangular cavity 0.33 m by 0.18 m had been dug into one of the slabs of the pavement. It is centred on the median axis of the doorway and in the bottom one can see in the middle, a small depression, oval in plan and conical in section, which is filled with a plug of unbaked clay. This rectangular cavity

In the north-eastern part of the building, the double-faced masonry is composed of medium sized cut blocks of limestone with an inner filling. This construction method differs clearly from the opus incertum of the other walls of the construction. This could represent an act of restoration during a possible second phase of the building s use. The upper part of the walls were built of mud-brick that were found fallen and disintegrated in the interior of the building.

2. The preliminary study of this structure was undertaken by I. Hairy, architect-archaeologist, C RS, SR3134, Centre d Etudes Alexandrines.

64

V. PICHOT: MAREA PENINSULA Fig. 12: ( left) Eastern zone of Sector 3 under excavation. Photo V. Pichot ( CEAlex Archives) . Fig. 13: ( below) Sector 3 - O ne of the two limestone sphinxes discovered in the tower building ST300. Photo V. Pichot ( CEAlex Archives) . Fig. 14: ( bottom) Sector 3 Bronze candelabra after restoration in the CEAlex laboratory. Photo A. Hussein ( CEAlex Archives) . probably held a support, a base bearing either an element of decoration, e.g. statuary, or a functional object. The interior doorway was framed by two pilasters, of which only the bases remain, that bore two little, limestone sphinxes ig. 13 that were found dismantled in the nextdoor room. This interior doorway opens onto a room some 2.40/2.35 m by 4.45 m. At the end of this space, to the south, two small assemblies of blocks, the sides of which have been covered by cleanly cut stone facing, stood against the walls in the corners. These two assemblies form the base of two staircases that led to the upper floors. The two lateral spaces situated to each side of this distribution space are completely closed and inaccessible at this level. The eastern space was filled with clay and other rubble, while within the filling of the other space there is a block of masonry in opus incertum that most probably supported the weight of a now-disappeared superstructure. When considering the different aspects of this construction, its interior layout, as well as the archaeological material discovered here, a hypothesis may be that the function of this building could relate to the representations of the tower-house on ilotic mosaics, or that it was religious. This building was part of a much bigger ensemble, the extension of which to the north-west was partially uncovered by excavation in 2008 and was still relatively intact beneath the construction levels of building ST301. Two spaces clearly similar in size and limited to the east by wall MR3114, were discovered. Their western limit has not yet been determined. The most southerly space is difficult to interpret. It was practically destroyed in its entirety by the implantation of wall MR3082 that lies in part upon wall MR3121 that separates the two spaces. The excavation of the northern space, delimited to the north by MR3116, has revealed the remains of flooring of compact clay in which were set several ceramic storage jars. Although we do not yet know for sure the duration of its use, the evidence of the archaeological material, including a Hellenistic lamp with seven wick-holes, a faience pendant and a bronze candelabra ig. 14 , would suggest that this ensemble was 65

LAKE MAREOTIS: RECONSTRUCTING THE PAST settlement and organisation of a sizeable town on the lakeshore. p until the present, our knowledge of the occupation of the site of Marea was limited to the Byzantine period 5th to 7th centuries AD . The work of the CEAlex on the peninsula has proven a settlement existed here well before the 5th century AD. The excavations undertaken in Sectors 3 and 4 have revealed a large occupation dating to the end of the Hellenistic and the beginning of the Roman era, as well as numerous signs of a Hellenistic occupation prior to the 2nd century BC. The opportunity at Marea to excavate and study a workshop quarter of this size, in an environment so well defined by the fact that it stands on a peninsula, is really quite exceptional. In addition to gathering information on the production itself and on the production lines, it is also possible to study the spatio-temporal organisation of the site in the very heart of the peninsula and to resituate it within its immediate environment habitation, religious area, etc. . Bibliography De Cosson, A., 1935, Mareotis: Being a Short Account of the History and Ancient Monuments of the North-western Desert of Egypt and of Lake Mareotis. London. Description de l’Egypte, Etat moderne, Vol. 1, 1809. Paris. El- akharani, .A., 1983, Recent Excavation at Marea in Egypt. In G. Grimm, H. Heinen & E. Winter eds. , Aegyptiaca Trevernsia II, Das Römisch-Byzantinische Ägypten. Akten des internationalen Symposions 26.-30. September 1978 in Trier, 175-186. Mainz. El- alaki, Mahmoud Bey, 1872, Memoire sur l’antique Alexandrie, ses faubourgs, ses environs. Copenhagen. Petruso, K., & Gabel, C., 1983, Marea: A Byzantine Port on Egypt s orthwestern rontier. Archaeology 36.5: 62-63, 76-77. Rodziewicz, M.D., 1983, Alexandria and district of Mareotis. G raeco-Arabica 2: 199-216. Rodziewicz, M.D., 1998, rom Alexandria to the West by Land and by Waterways. In .- . Empereur ed. , Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athè nes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole franç aise d’Athè nes, 11-12 décembre 1988: 93-103. Athens, Paris. Sadek, M., 1992, The baths at the ancient harbour of Marea. Sesto Congresso internazionale di Egittologia, Vol. I: 549-554. Strabo, G eography, A. Meineke transl. , 1877. Lipsiae. Szymańska, H., & Babraj, K. eds. , 2008, Byzantine Marea. Excavations in 2000-2003 and 2006, Marea, Vol. 1. Krak w.

Fig. 15: Sector 3 - Hoe chalk burner. Photo V. Pichot ( CEAlex Archives) . in use from the end of the 3rd century BC. It was partially destroyed by the installation of one or several occupations to the north during the Roman period that were in part connected to workshop activity,3 and thereafter by chalk burners that were active in a later period ig. 15 , most probably in the 5th and 6th centuries AD, and that largely destroyed all construction in this zone. C onclusion The Mareotis area, today semi-desert, was once a fertile region with widespread agricultural activity. In the Graeco-Roman era, Lake Mariout was a veritable inland sea with intense traffic. As a zone of movement and exchange, it was connected to the ile by canals and with the sea at several points through the course of its history. A passage also allowed traffic access to the canal of Alexandria. The Ptolemaic capital had a lakeside port with quays and warehouses that Strabo 17.1.7 considered richer than the maritime harbours of the city. Marea is one of the best examples of Mariotic lakeside towns whose development was tied to exchange between Alexandria and its hinterland: exports of agricultural production and local industry glass, metal, etc. , imports of raw materials and pilgrimages during late antiquity. Its state of preservation allows us to study the problems of

3. The mission of 2008 has brought to light numerous structures connected to workshop activities of the Roman era to the west and north of the tower building. They are presently under study and will be the object of a supplementary mission in 2009.

66

M. RODZIEWICZ: MAREA AND PHILOXENITE’S ARCHAEOLOGICAL EVIDENCE

On Interpretations of Archaeological Evidence Concerning Marea and Philoxenite Mieczyslaw D. Rodziewicz

The largest and best preserved ancient lake harbour southwest of Alexandria was located on the map of the Mariout region by Mahmoud El-Falaki in 1872 with a description and interpretation of the site as the centre of the ancient Mareotic district being identified as the site of Marea ElFalaki 1872: 96). For a long time this interpretation was considered as correct and unshakable. It was expected that with continuous advances and development of field archaeology in the region, this interpretation would be substantiated by more archaeological evidence. However, in the first half of the past century the site was not investigated by the most active archaeologists in Alexandria such as Breccia and Adriani. Instead, two other sites in the Mareotic region captured the attention of these great archaeologists of the last century. The first such place was Abu Mina discovered in the desert south of Alexandria and excavated from 1905 by German archaeologist C.M. Kaufmann 1908 . The second important site was the large city of Taposiris Magna, west of Alexandria, that spread around the Ptolemaic temple of Osiris, with the neighbouring site of Plinthine, where Breccia 1922: 353 and Adriani 1940 concentrated their field research in the first three decades of the last century. There was also a short German campaign near Amreyia, situated in the desert south of Alexandria, where very informative architectural remains of early Christian date have been uncovered Eilmann, et al. 1930 .

ten by Breccia 1922: 335 and orster 1922 , the port of Marea is hardly mentioned compared with such important places west and south of Alexandria as Taposiris Magna and Abu Mina. The best equipped excavations in the Southern Desert were made by expeditions organized by the German Archaeological Institute in Cairo in the 1930s. They were carried out near the modern market place and settlement of Amreyia Eilmann, et al. 1930 . Specific rural types of early Christian structures characteristic of the region, although rather modest and of an introductory nature, were recorded and published very quickly. It is a great loss that the excavators did not come back to the region after the second world war to continue the research. In the second half of the past century, excavations at Abu Mina, undertaken by the German Archaeological Institute at Cairo, were resumed under the direction of architects, Müller-Wiener and later Grossmann. They extended the area of research around Abu Mina, the largest pilgrimage centre of early Christian Egypt, and surveyed the neighbouring territory of the Lake Mariout harbour. M ller-Wiener 1967 and R der 1967 , devoted their time to investigating not only the ancient settlements in the region, but also the sources of local building material, including ancient quarries located along the southern shore of the lake, particularly one located near the modern village of Bahig M ller-Wiener 1967: 104-117 R der 1967: 118-131 . Their interests also extended to an analysis of the building materials visible on the surface of the land along the southern shores of Lake Mariout and the surrounding area, with the clear intention of identifying a link between the chain of settlements in the desert and the pilgrimage centre at Abu Mina. The article of M ller-Wiener 1967 , entitled Siedlungsformen in der Mareotis”, was based on rational observation and the description of visible remnants of ancient structures on the surface igs. 1 & 2 . In the harbour of so-called Marea, prior to any excavations, he identified on the surface, a double bath, structured frames of the quays, with accompanying buildings, and the most important construction, the monumental three apsidal church M ller-Wiener 1967: 106).1 All his identifications were correct, and they have not altered as a result of later excavations. More im-

The most complete description of the whole area was offered by Anthony De Cosson, the former director of the railway west of Alexandria, in the 1930s. His book on Mareotis was based on then available literature on the subject and his extensive knowledge of the land from Alexandria westwards far beyond Marsa Matruh ancient Parethonium De Cosson 1935: 131 . In his description of Marea he follows the description and name of the extensive lake port remains situated south-west of Alexandria given by Mahmoud El-Falaki. In both, El-Falaki’s and De Cosson´ s descriptions, the site was presented as 1.5 km long, but not very broad, and limited to the chain of visible ruins of fallen walls along the lake shore. Nevertheless, in their opinions the place was the most important centre of the historical Mareotic region, as mentioned in ancient literary sources. Other modern publications concerning ancient Mareotis have dealt very little with the largest lake harbour in the region. et, in the first half of the last century the site was recorded on several maps of the Survey of Egypt series, but was rarely presented in published photographs. The best known are pictures showing the western jetty of the harbour surrounded not by water, but instead by marshes. In the most popular books of that time writ-

1. M ller-Wiener 1967: 106, n. 16: In Marea sind obwohl das aufgehende Maurerwerk weitgehend fortgeschleppt zu sein scheint noch zahlreiche Bauten erkrnnbar grosses Doppelbad, Hafenbefestigungen mit Kaianlagen und drei langen Molen, Kirche mit drei-apsidialem Grundriss usw.usw. eine n here ntersuchung des rtes w re zumal angesichts seiner Bedeutung in der r hzeit usserst w nschenswert.

67

LAKE MAREOTIS: REC

STR CTI G THE PAST

Fig. 1: Central part of the plan of Mareotis drawn by Mül lerWiener in 1966, with the pilgrimage road from the lake harbour to Abu Mina and settlements, among them settlement “ M” in Huwaryia village ( after Mül lerWiener 1967: Fig.1) .

Fig. 2: Abu Mina with surrounding settlements/ karms, and pilgrimage road towards the lake harbour ( after Mül ler-Wiener 1967: Fig. 4) .

68

M. RODZIEWICZ: MAREA AND PHILOXENITE’S ARCHAEOLOGICAL EVIDENCE portantly, he identified a separate settlement M on the plan reproduced on his ig. 1 see ig. 1 , at a distance of 1.5 km south from the ruins of the lake harbour known as Marea M ller-Wiener 1967: 110 . He presented its dimensions 600 x 800 m and described briefly the remains of a brick cistern located at the edge of the modern road, the same that later was identified as a water reservoir serving the double peristyle building, published by its first excavator El- akharani, as the Byzantine House El- akharani 1983: 175 . Thus, M ller-Wiener clearly separated all these inland situated structures from the harbour to the north, and very clearly stated that settlement M was not a part of a larger town plan comparable with ancient Alexandria, as was to be suggested a decade later, and supported by his successor heading the excavations in Abu Mina Grossmann 2003: 13 .

saved most of the monuments excavated by the Alexandria niversity Expedition, headed by Prof. El- akharani. The winery in Huwaryia village, excavated in 1977, received limited reconstruction mainly for protection but also didactic reasons), and was covered with a roof. The excavation of the so called Byzantine House, which most probably served as a hospice for pilgrims heading to Abu Mina, was completed. It was partly reconstructed to prevent further damage and also for didactic requirements. The remains of buildings at the lake harbour were treated similarly Rodziewicz 2002, 2003 . During these works many previously completely unrecognised structures were identified, thus extending the possibilities of defining more features of the local topography, its culture, economy and technology. If they would have been known to Prof. ElFakharani at the time of his activities in the area, he would certainly have changed some of his first hypothetical identifications, and his general opinion of the site.

In 1978 the first archaeological excavations of the socalled site of Marea were financed by Alexandria niversity under the direction of Professor of Archaeology, Dr Fawzi El-Fakharani, a very energetic organizer and gifted speaker. In a very short time his discoveries at the site became very popular both nationally and internationally. Prof. El- akharani published in 1983 an extensive report on the survey and excavations of a vast area of Mareotis, about 6 km long and 4 km broad. ther publications of the site of Marea by Sadek 1978: 67 and Petruso and Gabel 1983: 62 followed, although analysis of available material and structures excavated there were not completed. At that time, there arose the notion of a very large and prosperous city of Marea around the existing remains of the lake harbour, which some supposed to be equal in size to ancient Alexandria. The Mareotic settlements 1.5 km south of the port, and the Ptolemaic rock-hewn tombs situated about 5 km west at the site called uassimyia , previously investigated by Müller-Wiener, were included by El- akharani in the city of Marea El- akharani 1983: 176, 186). The lack of any continuation and evolution of chronologically parallel architectural structures between these distant places was not taken into consideration by the enthusiasts of this notion of the great city of Marea .

The most required change in the general knowledge of the site concerns the extension of the urban area of Marea to the west, i.e. to the tombs of the Ptolemaic period at modern assimiyia site, ca 5 km from the city), which is still considered by some archaeologists as a part of the western necropolis of the capital of the whole Mareotis region El- akharani 1983: 176 Bagnall & Rathbone 2004: 75).2 Of three hypogea, still well preserved in the 1970s, only traces of the largest, located on the southern slope of the rocky ridge, are still recognisable. The identification by El- akharani of assimiyia site as the western necropolis of Marea is not archaeologically substantiated, because similar hypogea existed on the rocky ridge near Huwaryia village 1.5 km from Marea , and all along the same ridge were located other types of tombs such as chambers, shafts or pit tombs, which were dispersed over the area, also considered by some archaeologists as belonging to the city of Marea . urthermore, the extensive ancient ruins with a subterranean tomb, now located near the modern Huwaryia railway station, which supposedly delimited the southern urban area of ancient Marea , are according to the results of the survey of 1990s EAO Delta West Inspectorate unpublished , the remnants of an isolated structure with a church inside, which may indicate one of the numerous Christian monasteries that existed here in the early Christian period.3 There were no detectable traces of urban features such as street connections between this structure and the group of low quality houses with an associated wine press, that were excavated and published by El- akharani 1983: 183-184 . This site was earlier identified by M ller-Wiener as a separate settlement M ller-

In 1972 Fraser, the author of a major work on Ptolemaic Alexandria, wrote, Although Marea had been an important town in Pharaonic times, these remains are evidently late structures, they are not Ptolemaic and may be Byzantine or early Arab raser 1972: 146 . Some years later an essential question arose about the location of the pilgrimage lake harbour, known from the Coptic Encomium on St. Menas as Philoxenite, a place that is also mentioned in other Christian sources Drescher 1946: 147-148, 1949: 15-16 Rodziewicz 1983, 2003 . The next phase of research connected with the lake harbour, went together with the protective work undertaken by the Egyptian Antiquities Organization in 1982, which required a factual analysis of all preserved archaeological material essential for planning and intensifying the protection of the cultural heritage in this area. I was involved in this process, particularly in the protective works, which

2. El- akaharani 1983: 176: The discovery of burials in the west and south marked the limits of the town westwards and southwards since ancient cemeteries were customarily located outside of inhabited area Bagnall & Rathbone 2004: 75: In the cemetery to the west of the town …” . 3. Compare these structures to those excavated by the Swiss at Kellia, Kasser 1983, 1986 with further references in Actes du Colloque de G eneve, 13 au 15 aout 1984 , Geneva 1986.

69

LAKE MAREOTIS: REC

STR CTI G THE PAST

ing a now horizontal stratigraphy, unlike most of the Egyptian ancient settlements which grew upwards and form characteristic hills called koms . Contrary to the first and still valid identification of this archaeological site by M ller-Wiener in the 1960s in particular the water cistern belonging to the big Byzantine House that he published), as a separate construction in a rural environment, the site is still directly or indirectly considered by some archaeologists as a centre of the urban area of an extended town called Marea Bagnall & Rathbone 2004: 74-76 .5 Thus, despite the growing archaeological evidence that strongly opposes the identification of the lake port as Pharaonic Marea, we still read about the remains of an ancient town Stadtanlage located as far south as the modern village of Huwaryia Grossmann 2003: 15 . As a consequence of such interpretations of the ancient village, general conclusions were formulated concerning the location of wineries in urban environments Bagnall & Rathbone 2004: 74-76 Grossmann 2003: 15, n. 27 . urthermore, another conclusion drawn, was that the so-called by Prof. El- akharani Byzantine House at Huwaryia is the most convincing archaeological document attesting the existence of the urban centre of ancient Marea, where in reality at this location there existed a large building structured in a rural manner, surrounded by cultivated land. Judging from its huge dimensions and the church built inside, it may have served as a very comfortable and safe pilgrimage hospice from the early 6th to late 7th century AD, i.e. exactly at the time of prosperous pilgrimage activities to Abu Mina Rodziewicz 2003: 27-47 . o older or younger constructions in the area of the building have been registered. The proper understanding of the function of this large building is crucial for the interpretation of the character of the area, which was described by MüllerWiener 1967 who undertook a detailed survey of the area in the 1960s), as a rural settlement, not the centre of any larger town. He mentioned visible surface remains of a brick built water cistern that was located right on the edge of the modern road, that are still visible today see ig. 3 . In late antiquity, this cistern was situated in a garden or in an open space surrounding the large double-peristyle building the Byzantine House of El- akharani . M llerWiener thought that the cistern may eventually be connected with the bath M ller-Wiener 1967 . However, the extensive neighbouring building the Byzantine House , whose excavation was completed in the EAO in the early 1980s, appeared to be the largest residential edifice known hitherto in the whole Mareotis region. It was equipped with a church in the central, eastern wing, between the two extensive colonnaded peristyles. The church had multi-

Fig. 3: Byzantine House/ Hospitium at the pilgrimage road from the lake harbour to Abu Mina ( drawn by M. Rodziewicz) . Wiener 1967: 110 see ig. 1 . In the southern part of this settlement an underground water cistern, a group of low quality rooms and, further to the west, a medium-sized winery, were partly uncovered by EA in 1992/3. They were associated with pottery of the late 1st century BC to the middle of the 2nd century AD author s identification . This winery appeared to be much older than the winery discovered by Prof. El- akharani 1983: Taf. 36 . ther architectural remains from this settlement are younger. At the most northern part of the ruins typical Muslim burials were located, which, according to Prof. El-Fakharani, were associated with glazed pottery sherds and a probable atimid coin El- akharani 1983: 176-177 .4 urther to the north, an extensive building identified by El-Fakharani as the Byzantine House was separately built outside the settlement El- akharani 1983: 184-186 . It occupied a surface of over 1,500 m2 and was surrounded on all sides by cultivated land Rodziewicz 1983, 1988a: 267-277, 1990, 2002: 1-22 ig. 3 . The house was not covered by structures belonging to other buildings, which means that it was situated in a typically rural environment, at the far edge of the older, but still inhabited village, hav-

5. Bagnall & Rathbone 2004: 74-76 p. 76: Archaeological evidence for wine production has also come to light. South of the limestone ridge are the remains of two wine-producing establishments. The larger and more interesting is aligned with the middle of the town”. However, such installations as wine factories and pottery kilns are not known in the very centres of ancient towns, where life was regulated by city law. uite the contrary, these are very typical rural establishments.

4. El- akharani 1983: 176-177: ew pieces of Islamic pottery sherds with green glaze and some white lines, and a piece of Fatimid coin were discovered just above the dead”.

70

M. RODZIEWICZ: MAREA AND PHILOXENITE’S ARCHAEOLOGICAL EVIDENCE

Fig. 4: Byzantine pilgrimage harbour/ Philoxenite ( Nos. 5- 17) and western/ coastal rural remains ( Nos. 1-4) . O n the eastern side, structured causeway ( No. 18) ( drawn by M. Rodziewicz) . K EY 1 – J etty of pre-Byzantine date 2 – Rural estate with large water wheel 2a – Small kiln 3– nidentified structure of rural character 4– Structures with mooring place

tructures identified in the pilgrimage harbour: 5– Pier 6 – Water wheel ( saqia) 7 – Public double bath 8 – Insula with shops 9 – Public building 10 – Public lavatory 11 – K ibotos?

coloured, geometrical opus sectile mosaics on the floor and a richly ornamented northern baptistery Rodziewicz 1983, 1988a, 1988b . The baptistery floor was made of mortar and painted red in a geometrical pattern composed of triangles. After extensive cleaning of the area, it appeared that beside the large church in the centre and numerous rooms around an extensive courtyard and probably on the upper floors , a collective nine seat lavatory existed there with two smaller ones aside. They were all built in the southern part of the edifice, by the staircase leading to the upper floors, and close to the bathing equipment located in the south-western corner of the house. The bathing facilities were not sufficient to serve the needs of pilgrims eventually stationed there, but a full programme of cold and hot baths was available just about 1 km to the west, in the large public bath that was surveyed, excavated and restored by EAO Delta West Inspectorate in the early 1990s.6 The bath, datable to the Byzantine period, was not attached to

12 – Pier 13 – Transept Basilica ( after G rossmann 2002: Fig. 9) 14 – Eastern harbour 15 – Water wheel ( saqia) 16 – Public bath 17 – Rock-hewn tomb/ hypogeum 18 – Causeway leading to the island 19 – Pilgrimage track to Abu Mina

any architectural remains. It was built in an open area and surrounded by cultivated land. The area of the settlement, which extended from the uplands to the lake, bears traces of various human activities, such as limestone quarrying, various burials on the rocky ridge, clay extraction, pottery firing in the numerous kilns registered there, and cultivation down to the shores of the lake R der 1967: 118-131 . The individual buildings, whose traces have been excavated and surveyed there, were built in a rural manner. Strictly urban construction methods of building are identifiable only around the lake harbour, erroneously identified as Pharaonic Marea El- alaki 1872: 96 . Most of the buildings at the port were constructed in one phase, with similar construction principles ig. 4 . The construction methods are comparable to buildings at Taposiris Magna, Alexandria and also the pilgrimage centre of Abu Mina Grossmann, et al. 1982 Grossmann, et al. 1984: 123-151 . They are not similar to the structures of neighbouring rural settlements and the karms/ karum” located further south. West of the Byzantine passenger harbour Philoxenite Marea , on a rough natural rocky coastline, are the pre-

6. The results of which were presented by Ahmed Abd El-Fatah at the Baln orient conference, held in Alexandria, December 2006 in press .

71

LAKE MAREOTIS: REC

STR CTI G THE PAST

served traces of a small size mooring facility and a quite well preserved short but wide and relatively high, jetty that differs remarkably from the two long Byzantine quays of the pilgrimage port ig. 4.1 . The jetty is constructed from very large sandstone blocks without hydraulic mortar. In front of it a cluster of ruins ig. 4. 2 is still visible, which surround a large courtyard, water wheel saqia), cistern and channels. They probably mark structures of a coastal rural estate similar to the one at Borg el-Arab. Early Roman building material has also been recorded. In addition, there are traces of another water wheel, a small kiln ig. 4.2a , and widely dispersed traces of furnace activities. These remains depict the kind of rustic-coastal human activities with small size mooring places ig. 4.4 , created prior to the construction of a large passenger harbour at the promontory. It is worth noting that during the very last era of the passenger harbour of Philoxenite, with large buildings already abandoned and decaying, old local activities returned, motivated by the rural economy, with an oil press and a pottery kiln installed on top of the ruined constructions.

tures at the lake port Grossmann 2003: 13-20 .7 Arguments against this proposition vary from self-contradictory to those based on out of date, unproven archaeologically theories, particularly that of the existence in that place of a large town, Marea , the Pharaonic capital of the district, with a generalised opinion that the old town experienced an unprecedented economic boom in a period of well-documented and steady economic decline of Egyptian cites.8 With a new round of excavations at the lake port starting in 2000, we expected a flow of new information and objective interpretations of freshly unearthed structures and artefacts.9 However, this was not to be and the lack of a rational argument in the presentation of newly excavated material from the harbour during the conference on Medieval Alexandria in 2002, motivated me to publish a paper entitled Philoxenite-Pilgrimage Harbour of Abu Mina Rodziewicz 2003 . Three years later, I found a statement in the pamphlet published by the Polish Mission 2006 saying that there are several reasons to assume that Philoxenite was built within the city of Marea, thus supporting the interpretation of the site as that of the location of Philoxenite, which I had first suggested in 1983 Rodziewicz 1983: 202, 2002 . However, the belief that Marea lies under the Byzantine buildings of Philoxenite, despite the lack of any older urban structures there that should date back at least to the time of Psametic I, is still maintained. In the

It is the author’s belief that the lake port in Mareotis should be interpreted as a disembarkation point for Christian pilgrims travelling to Abu Mina, and not as a town with a long tradition matching the history of the ancient capital of a Pharaonic nome Rodziewicz 1983, 1988a, 1988b, 2002). This opinion was formed on the back of extensive personal study of the whole lake harbour area from the late 1970s to the mid-1990s. The text of the Coptic Encomium and other Christian sources serve to support this view Drescher 1946: 147 . Since, in the early 1980s, the water of Lake Mariout was at a very low level, all jetties and other structures at the harbour were dry, and research, as well as all necessary examination of the construction phases, was much easier than nowadays. or identification of the chronology of the site I used my knowledge of the pottery acquired on other excavations in Egypt and elsewhere, but primarily Alexandria Rodziewicz 1976 . The most informative and numerous sherds were those spotted in the mortar of the lowest parts of the jetties. They belonged to Late Roman Amphora 1, which clearly fixes the chronological horizon of the big harbour in the Byzantine period and parallels the chronology of Abu Mina. The types of the structures, the extension of the urbanised surface, three large water wells ig. 4.6, 4.15 and a very small local cemetery hewn in a rocky hill ig. 4.17 , and a very large church ig. 4.13 and extensive harbour, lead me to conclude that the site did not develop gradually, but that is was constructed in a short time and according to a very specific programme. urthermore, there were no older or younger pottery sherds collected at the site.

7. Grossmann 2003: 13-20, considers my opinion unacceptable, but his arguments are not backed by any factual documentation concerning the chronology of the artefacts in the lake harbour and pilgrimage centre in Abu Mina which are mostly not yet published . Thus, his opinion on differences in chronology of the bath in the lake port and of the pilgrimage centre at Abu Mina can be taken only as his personal supposition, not supported by the necessary documentation and study of the pottery collected there. 8. According to Bagnall 1993: 108 , despite continuous economic decay in late antiquity, not all cities in Egypt suffered regress, and Marea in the immediate hinterland of Alexandria enjoyed a period of prosperity. He says that the flourishing development at Marea, on the lake which served Alexandria´ s back door by water, points to vigorous commerce by this route in the late period”. So, the old Pharaonic city of Marea, of which we still do not have any archaeological evidence and whose location is still unknown, in Bagnall s publication flourished in the 3rd century AD. 9. The Polish Mission, active there since 2000, is headed by Dr. Hanna Szymanska from the Archaeological Museum in Krakow. With great sadness, we observe subjective methods of interpretation of the archaeological evidence obtained there. At the conference devoted to Medieval Alexandria organized by CEAlex in 2002, Mrs. E. Wipszycka from the Polish Mission spoke about the identification of Philoxenite in the light of new discoveries. Her weak documentation mobilised me to publish the 2003 the article Philoxenite-Pilgrimage Harbor of Abu Mina . In 2006, a special exhibition was prepared by the Polish team about the excavations in Marea. It accompanied the conference Baln orient at which Mrs. Szymanska presented a paper on a new bath from the harbour. This exhibition entitled Marea - Polish Excavations in Egypt 2000-2004 conducted by archaeologists from the Archaeological Museum in Kracow under the auspices of the Polish Centre of Archaeology of Warsaw niversity , was supported by a pamphlet with a short text signed K.B. in which we can find the passage: Marea or Pholoxenite Identifying the accurate name of the town is a challenge that the Polish Archaeological Mission has undertaken. We have several reasons to assume that it was within Marea that the city of Philoxenite was built.”

Thus, in 1983 at a conference in Athens Graeco-Arabica I interpreted the harbour as a specialised port for pilgrims travelling to Abu Mina, as described in the Coptic Encomium, which also mentions its name as Philoxenite Drescher 1946: 15 . The description of the site in the Coptic Encomium matches quite well the preserved Byzantine struc72

M. RODZIEWICZ: MAREA AND PHILOXENITE’S ARCHAEOLOGICAL EVIDENCE Bibliography Abd el-Aziz, M., 1998, Recent Excavations around Abou Mina. In .- . Empereur ed. , Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athè nes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole franç aise d’Athè nes,11-12 décembre 1988: 65-73. Athens, Paris. Abd el-Fatah, A., 1998, Recent Discoveries in Alexandria and the Chora. In .- . Empereur ed. , Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athè nes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole franç aise d’Athè nes,11-12 décembre 1988: 37-53. Athens, Paris. Adriani, A., 1940, Annuaire du Musée gréco-romain 1935- 1939. Alexandria. Bagnall, R.S., 1993, Egypt in Late Antiquity. Princeton. Bagnall, R.S., & Rathbone, W., 2004, Egypt from Alexander to the Copts. London. Breccia, E., 1922, Alexandrea ad Aegyptum: G uide de la Ville Ancienne et Moderne et du Musée G réco-Romain. Alexandria, Bergamo. Décobert, C., 2002, Maréotide médievale. Des Bédouins et des chr tiens. In C. D cobert ed. , Alexandrie Mediévale 2, É tudAlex 2, 127-167. Le Caire. De Cosson, A., 1935, Mareotis: Being a Short Account of the History and Ancient Monuments of the NorthWestern Desert of Egypt and of Lake Mareotis. London. Drescher, ., 1946, Apa Menas. A Selection of Coptic Texts Relating to St Menas. Cairo. Drescher, ., 1949, Topographical otes for Alexandria and District. Bulletin de la Société Archéologique d’Alexandrie 38: 13-20. Eilmann, R., Langsdorff, A., & Stier, H.E, 1930, Bericht Ü ber die Voruntersuchungen auf den Kurûm El-Tiuwal bei Amrijr. Mitteilungen des Deutschen Archaologi schen Instituts Abteilung K airo 1: 106-129. El-Ashmawi, F., 1998, Pottery Kiln and Wine-Factory at Burg el-Arab. In .- . Empereur ed. , Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athè nes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole franç aise d’Athè nes,11-12 décembre 1988: 55-64. Athens, Paris. El- akharani, .A., 1983, Recent Excavation at Marea in Egypt. In G. Grimm, H. Heinen & E. Winter eds. , Aegyptiaca Trevernsia II, Das Römisch-Byzantinische Ägypten. Akten des internationalen Symposions 26.-30. September 1978 in Trier, 175-186. Mainz. El-Falaki, Mahmoud Bey, 1872, Memoire sur l’antique Alexandrie, ses faubourgs, ses environs. Copenhagen. Empereur, .- ., & Picon, M., 1998, Les ateliers d amphores du Lac Mariout. In .- . Empereur ed. , Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athè nes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole franç aise d’Athè nes, 11-12 décembre 1988: 75-91. Athens, Paris.

same pamphlet, the author presents important archaeological evidence, previously published by Grossmann 1993 . The pamphlet informs us that: The most interesting structure in the town is a basilica situated on an eminence overlooking the lake shore near the longest of the harbour piers… The greatest surprise however was concealed under the apse. About 1.80 m below the surviving tops of the apse walls, the floor of the firing chamber of a great kiln for the production of amphorae was discovered. The part of the church where liturgical ceremonies were held had been founded on these manufacturing remains. The kiln, which is one of only a few known from the Delta so far, turned out to be 8 m in diameter and the firing chamber floor was ca. 50 cm thick. Inside it, fired amphorae of the 2nd-3rd century were discovered next to clay stacking rings that had separated the vessels during firing.

Because no further evidence was presented, I feel obliged to conclude that, firstly, this kiln alone proves that the hill is to a great extent artificial, because it was formed by industrial refuse produced by the ruins of a vast pottery workshop but that secondly, these ruins should be understood as part of an extensive rural establishment of Early Roman type, comparable to several groups of similar remains hitherto uncovered around the lake Rodziewicz 2002: 12, ig. 3 Abd-el atah 1998 El-Ashmawi 1998 Abd El-Aziz egem 1998 . ne site discovered in the 1980s by the Marsa Matruh Inspectorate of the Egyptian Antiquities Organization, was located on the southern shore of this same Lake Mariout, near modern Borg el-Arab village El-Ashmawi 1998 .10 Thus, the kiln discovered under the church in Marea ig. 4.13 belongs to a chain of Early Roman rural enterprises located all along the southern shore of the Mareotis Lake and has nothing to do with the great city of Marea. Therefore, we should retain the view that the Byzantine port of Philoxenite, created for pilgrims travelling to Abu Mina, which was built not upon the urban ruins of Marea ancient capital of Mareotis as described by Herodotus II.149 but on the ruins of a nameless, extensive Early Roman rural estate, similar to many others in the region, a number of which have been located along the shores of Lake Mariout.11

10. El-Ashmawi 1998. I had the privilege to work there in 1987/88 as an adviser in the protective works of this large Early Roman rural estate, situated on Lake Mariout. 11. Preliminary reports on excavations in Marea are scarce see Polish Archaeology in the Mediterranean XII-XVII, 2001-2007, Warsaw niversity Press . The exhibition illustrating results of the Polish excavations in Marea presented in the Bibliotheca Alexandrina, December 2006, was supplied with texts describing photos. A pamphlet was also presented. Both ignored all previous extensive research and restoration contributions to the site done by individuals and institutions. Reconstruction done at the site in the 1980s shown there on big panels that was carried out by the Egyptian Antiquities Organization was not properly described and the suggestion was that the works had been done recently by the Polish Mission. See Szymanska & Babraj 2007, with references to previous reports.

73

LAKE MAREOTIS: REC

STR CTI G THE PAST

Forster, E.M., 1922, Alexandria: A History and a G uide 5th edn.). New York. Fraser, P. M., 1972, Ptolemaic Alexandria, Vol. I. Oxford. Grossmann, P., Kosciuk, ., Severyn, G., & Severyn, H.G., 1984, Vorl. Bericht. Mitteilungen des Deutschen Archaologischen Instituts Abteilung K airo 40: 123-151. Grossmann, P., aritz, H., & Romer, C., 1982, Vorl. Bericht. Mitteilungen des Deutschen Archaologischen Instituts Abteilung K airo 38: 131-154 Grossmann, P., 1986, Abu Mina. A G uide to the Ancient Pilgrimage Center. Cairo. Grossmann, P., 1993, Die uerschiffbasilika von Hauwariya und die Bauten dieses Typus in Ä gypten als Repr sentanten der verlorenen r hchristlichen Architektur Alexandreias. Bulletin de la Société Archéologique d’Alexandrie 45: 107-121. Grossmann, P., Arnold, ., & Kosciuk, ., 1997, Report on the Excavations at Abu Mina in Spring 1996. Bulletin a la Societe d´ Archeologie Copte 36: 83-98. Grossmann, P., 2002, Christliche Architektur in Ägypten. Leiden. Grossmann, P., 2003, ochmals zu Marea und Philoxenite. Bulletin a la Societe d´ Archeologie Copte 42: 13-20. Herodotus, The Histories. A. de S lincourt transl. , 1971. Middlesex. Kasser, R., 1983, Survey Archeologique des K ellia ( BasseEgypte) . Rapport de la Campagne 1981, Vol. I-II. Louvain. Kasser, R., 1986, Le site monastique de K ellia. Sources historiques et explorations archeoloqigues. Geneva. Kaufmann, C.M., 1906, Die Ausgrabung der MenasHeiligtüm er in der Mareotiswüs te. Cairo. Kaufmann, C.M., 1908, La Découverte des Sanctuares de Ménas dans le Désert de Maréotis. Cairo. Kaufmann, C.M., 1924, Die Heilige Stadt der Wüs te. U nsere Entdeckungen, G rabungen und Funde in der altchristrlichen Menasstadt. Kempten. M ller-Wiener, W., 1967, Siedlungsformen in der Mareotis. Archäol ogischer Anzeiger 82.2: 103-117. Petruso, K., & Gabel, C., 1983, Marea. A Byzantine Port on Egypt´ s Northwestern Frontier. Archaeology, Sept/ ct: 62-63, 76-77. Röde r, J., 1967, Antiken Steinbrüche der Mareotis. Archäol ogischer Anzeiger 82.2: 118-131.

Rodziewicz, M.D., 1976, La céramique romaine tardive d´ Alexandria, Alexandrie I. Varsovie. Rodziewicz, M.D., 1983, Alexandria and District of Mareotis. G reco-Arabica 2: 199-216. Rodziewicz, M.D., 1988a, Remarks on the Domestic and Monastic Architecture in Alexandria and Surroundings. In E.C.M. van den Brink ed. , The Archaeology of the Nile Delta, Proceedings of the Seminar held in Cairo, 19-22 O ctober 1986: 267-276. Amsterdam. Rodziewicz, M.D., 1988b, Remarks on Peristyle House in Alexandria and Mareotis. Praktika 1983: 175-178. Rodziewicz, M.D., 1990, Taenia and Mareotis, Archaeological Research West of Alexandria. Acta of the First International Colloquium of the Egyptian Society of G reek and Roman Studies I: 62-81. Cairo. Rodziewicz, M.D.,1995, Eco-Archaeology of Ancient Alexandria and Mareotis. In A.A. Hussein, M. Miele & S. Riad eds. , Proceedings of the Seminar on G eosciences and Archaeology in Mediterranean Countries, Cairo, November 28-30/ 1993: 127-139. Cairo. Rodziewicz, M.D., 1998a, Classification of Wineries from Mareotis. In .- . Empereur ed. , Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athè nes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole franç aise d’Athè nes,11-12 décembre 1988: 27-36. Athens, Paris. Rodziewicz, M.D., 1998b, From Alexandria to the West by Land and by Waterways. In .- . Empereur ed. , Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athè nes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole franç aise d’Athè nes, 11-12 décembre 1988: 93-103. Athens, Paris. Rodziewicz, M.D., 2002, Mareotic Harbours. In C. Décobert ed. , Alexandrie Médiévale 2, É tudAlex 8: 1-22. Rodziewicz, M.D., 2003, Philoxenit Pilgrimage Harbour of Abu Mina. Bulletin de la Société Archéologique d’Alexandrie 47: 27-47. Sadek, M., 1978, The ancient port of Marea. Cahiers des É tudes Anciennes VIII: 67. Szymańska, H., & Babraj, K., 2007, Marea. Sixth Season of Excavations. Polish Archaeology in the Mediterranean XVII. Reports 2005: 55-66.

74

K. BABRAJ & H. SZYMANSKA: MAREA OR PHILOXENITE?

Marea or Philoxenite? Polish Excavations in the Mareotic Region 2000-2007 Krzysztof Babraj & Hanna Szymańska

Harbour installations of all kinds can be identified all along the south-western shores of Lake Mareotis. The most imposing ruins of Byzantine date located just 45 km southwest of Alexandria, extend along the coast for 1.5 km and are currently under exploration by a Polish mission ig. 1.

gion remained the most important agricultural producer in northwestern Egypt Haas 2001: 47 . The lake was an important communication route in Ptolemaic and Roman times. Goods were transported down the channels from inland, reloaded in Marea and sent to Alexandria, from where they were shipped to other parts of the Mediterranean. The information contained in the Periplus of Pseudo-Scylax Desanges 1978: 404-405 , dated presumably to the early 4th century BC, indicates that Lake Mareotis may have been navigable in Pharaonic times and that the water was potable. Strabo VII, 1, 7 reports that the lake port in Marea handled more goods than ever came to Alexandria by the sea route oyotte & Charvet 1997: 83 . South of Alexandria, there was a harbour, no longer visible, called Portus Mareoticus it handled the goods that were shipped via the lake. rom there they were transported down a canal named Kibotos through Alexandria to the Mediterranean. ne of the canals linking Mareotis with the so-called Canopic Branch of the ile Rodziewicz 1983, 1998: 101 . The fertile Mareotis region was celebrated in antiquity for its vineyards, olive groves, fruit orchards and papyrus plantations. It produced food for the capital city and rich landowners had their estates here. In the mid-7th century it was still a place for Egyptians to come to appease hunger History of the Patriarchs 1. 14, 501 . Ancient writers spoke warmly of the quality of the local wine exported to Rome Virgil, G eorg. II, 91 Strabo, VII, 1, 14 Horace, O des I, 37, 14 Athenaeus, Deipn. I, 33 . Even today one of the most renowned Egyptian grapevine growers and wine producers is based on the southern shores of the lake. Almost 30 furnace sites for firing amphorae from the early Hellenistic to late antiquity have been identified on the southern lake shore Empereur & Picon 1998: 75-91 El-Ashmawi 1990: 55-64 . In 2003, the Polish expedition unearthed beneath the basilica at Marea, a pottery kiln for firing amphorae of 2nd-3rd centuries date see below . This evidence, when considered together with the numerous remains of glass workshops in the region, gives an idea of the crafts that were an important part of the economic life of the region Kucharczyk 2008: 129-143 .

In 2000, excavations at the site were undertaken by a Polish team working under the auspices of the Polish Centre of Mediterranean Archaeology of Warsaw niversity and the Archaeological Museum in Krak w. The license granted by the Supreme Council of Antiquities SCA of Egypt, covers an area of c. 19 ha. A hypsometric grid was put in place and a surface survey of architectural remains carried out 16 objects were found Szymańska & Babraj 2008: ig. 1 . In the course of the following eight field campaigns, the team has fully cleared a Byzantine bath of the 5th to 7th century connected with a well that supplied it with water, and a funerary chapel. The present focus is on the exploration of a huge Christian basilica standing on the promontory above the port. History of the Site The site was identified as ancient Marea Μαρεία by Mahmud Bey El- alaki, court astronomer to Egypt s Viceroy Ismail Pasha El- alaki 1872: 96 Kees 1930: 16761678 Am lineau 1973: 241-243 Calderini 1980: 233 Goma 1980: 647 , but today this identification is being reevaluated as indicated in the discussion below. The extensive town ruins situated on Lake Maryut ancient Mareotis , 45 km from Alexandria, near the village of Hauwariya, have fascinated researchers for a long time. De Cosson prepared one of the first maps of the coast presenting the location of mainly architectural remains De Cosson 1935: 131 . The history of this imposing harbour town with four grand jetties for ships to dock and port basins to protect against wind and waves, can be traced in the ancient sources. Herodotus II, 30, 2 reported that in the times of Psammetich I of the 26th Dynasty, Marea was a strategic position garrisoned by troops defending the border with Libya. General Amasis defeated the armies of Pharaoh Apries at Marea and took the throne in 570 BC Diodorus Siculus I, 68 . nder Persian rule, the town was allegedly the capital of an independent Libyan-Egyptian kingdom extending from the Canopic Branch of the ile to Cyrenaica. Its ruler by the name of Inaros, challenged the Persians and lost his life after the fall of Memphis in 454 BC Thucydides I, 104 Winnicki 2006: 135-142 .

The channels that fed the lake silted up in the 8th and 9th centuries due to lack of maintenance, and the fact that fresh water from the ile, mainly the Canopic Branch, was cut off. The lake began to dry up Rodziewicz 2002: 9 leading to the decline of towns particularly along its southern shores. In 1801, the lake bed was filled with sea water after the English opened the sluices in order to cut off the apoleonic army from fresh water. The present surface of the lake is c. 90 km2, and its depth averages at 1.50 m Blue & Ramses 2005: 7 .

Marea never lost its importance as a commercial harbour despite the founding of Alexandria, and the Mareotis re75

LAKE MAREOTIS: REC

STR CTI G THE PAST

Fig. 1. Remains of structures along the southern shore of Lake Mareotis. K EY 1. 1a. 2. 3. 4. 5.

Tholos baths Stone pier 41 m long Stone pier 111 m long Stone pier 125 m long Harbour facilities nidentified structure presumably dry dock uncovered by Petruso

6. 7. 8. 9. 10.

ra n by M. iepok lc ycki and . as c yk .

teps belonged to the unidentified structure y antine baths Polish e ca ations Well operated with saqiyah ( Polish e ca ations Double baths uncovered by F. El-Fakharani Decumanus along the shops cleared by F. El-Fakharani

11. Remains of saqiya 12. G ranary uncovered by F. El-Fakharani 13. asilica Polish e ca ations 14. Funerary chapel ( Polish e ca ations 15. Tombs carved in the rock 16. Remains of the ancient road leading to the island

time of Byzantine domination the region had lost its leading role. The town fell due to lack of water as discussed above, but the decline was also due to political and social factors. Mareotis was overrun by semi-nomadic Bedouin tribes who plundered whatever stood in their way D cobert 2002 . At the beginning of the 8th century, in the rule of the Abbasids which coincides with the period when all life seems to have disappeared from the site, a new wave of Arab nomads appeared in the region. These people were politically active and they increasingly took control of local church institutions D cobert 2002: 137 . bviously, the atmosphere of the place was no longer favourable to

Mareotis was Christianised eventually but the relations with the Patriarchate in Alexandria passed through different phases Timm 1988: 1593-1603 . The first evidence of Christianity is linked to waves of repression, the effect of which resulted in adherents of the new religion being banished to the region. In 538, by the power of ustinian s edict, the entire Mareotis region, previously part of the Roman province of Aegyptus Prima, was incorporated into Libya Edict 13.1, 9, 17-22 . In the period immediately prior to the conquest of Egypt by the Arabs in 641, Mareotis was a Byzantine province. The infrequent reference to Mareotis in Coptic writings is proof that by the 76

K. BABRAJ & H. SZYMANSKA: MAREA OR PHILOXENITE? Christians, particularly as a stopover for pilgrims. The wine from Mareotis, however, still retained its reputation through to the 7th century at least this is what we are led to believe from a story about the abstinence of monks cited by ohn Moschus 1946 .

difficult in view of continued industrialization and numerous tourist villa complexes being built along the shore. Remains of harbour facilities can be seen along both the southern shore and the northern shores of the lake. The biggest of these is Marea. The quay at the waterfront in the eastern end of the site preserves stone blocks that were used in its construction. Beside it there are four almost wholly preserved jetties where the ships docked. These jetties formed the eastern, central and western harbours. They were built of several courses of stone blocks of different size three or four courses have still been preserved and can be studied thanks to the overall drop of the level of water in the lake. The dimensions of the jetties are counting from the west to east : I L. 41 m, W. 6.50 m, blocks averaging: 0.50 m x 1.35 m II L. 111 m, W. 5 m III The longest jetty is located by the promontory with the ruins of the basilica, L. 125 m, W. 7 m, blocks averaging 0.65 m x 1.70 m IV L. 0.35 m presumably only the part on land has survived , W. 4 m found next to the island that was connected in antiquity with the mainland by an artificial causeway.

However, identification of early harbour remains at Marea was questioned mainly on the strength of field survey results, which revealed nothing earlier than 6th to early 8th century pottery. The architectural remains also proved to be essentially Byzantine in nature raser 1972: 146 . n these grounds Rodziewicz 1983: 202-204 proposed to identify the ruins recorded on the surface with a town established as a transfer point on the southern shores of the lake for pilgrims on their way to the sanctuary of Abu Minas some 20 km away. The Coptic Encomium of St. Menas written by the Patriarch Ioannes IV 775-789 , mentions numerous conveniences like hospices by the lake and rest-houses for those wishing to rest Drescher 1946: 147148 , a market place, porticoes and even a facility that we would refer today as a left-luggage office. This impressive rest stop was said to have been founded by Philoxenos, Prefect of the Emperor Anastasius 491-518 , hence the town s name of Philoxenite. Wipszycka 2002 has identified this official as the consul Philoxenos Soterichos, who held office in 525 and who is known also from consular diptychs Martindale 1980: 879-880 .

Marea or Philoxenite? The size of the harbour installations, which are currently being researched by a Greek-Egyptian team of underwater archaeologists headed by ikos Lianos from the Hellenic Society for the Study and Preservation of Marine Cultural Heritage and the Department of nderwater Archaeology in Egypt when considered in connection with the considerable size of the basilica on the lakefront and the innumerous but conclusive evidence of Early Roman potsherds, indicates that a large urban agglomeration flourished on the spot already in pre-Byzantine times. either can the extensive sewage network be considered as belonging to a town of little consequence. Philoxenos may have founded his city adjacent to existing harbour installations that had formerly been part of Marea, before it was abandoned as an urban centre prior to the silting up of the fresh-water channels that supplied the lake, but there is little doubt that settlement existed here prior to its foundation. Already in the 2nd century AD Athenaeus 33d wrote of Marea as the source of a wine of fine quality he called Marea a village κώμη . Athanasios 85 col. 400b-c had the same impression of the town 200 years later. Grossmann 2003: 16 thinks that the Hellenistic architecture of Marea was of mud-brick which is easily disintegrated. It should be noted that at the margins of the settlement mud-brick is a building material used in the lower layers of the basilica, chronologically contemporaneous with the pottery kiln of the 2nd-3rd centuries AD.

Location of the Site The geographical coordinates of the site are: 30o 59 32.85 and 29o 38 58.34 E to 30o59 48.84 and o 29 39 28.32 E. The excavation area measures 19.7 ha. To the west it is limited by a double-lane road linking the waterfront with Shakush settlement the lake is a natural boundary on the north and east, and to the south there is the desert which extends as far as Shakush settlement. The overall height difference does not exceed 8 m anywhere on the site. The site grid consists of 42 geodetic points, mostly dictated by the natural topography. The coordinates were assumed locally, while the heights are listed by the Egyptian state geodetic service. The topographical plan of the site so far only covers the area excavated in 2000-2003 see ig. 1 . rom a geological point of view, Marea lies in a broad stretch of land between Mallahet Mary t, the long westward arm of Lake Mareotis, and a parallel ridge of oolithic limestone. The soils on either side of the ridge are mostly calcarenite, very favorable for cultivation with proper watering Warne & Stanley 1993 El- akharani 1983: 175 Rodziewicz 1995 Mycielska-Dowgia o & Woronko 2008: 17-18 . A similar ridge lies on the opposite side of the lake, separating Lake Mareotis from the sea. This is the Taenia ταινία Ridge which extends along the coast from Alexandria all the way to Abu Mina and Libya. Today these ridges are mostly destroyed as a result of heavy exploitation of limestone quarries. In the 4th and 5th century, Christian monasteries appeared in the Taenia they were called after the mile-stones that served to measure the distance from Alexandria. Their localisation today is very

There is however, opposition to the idea that this was the site of Philoxenite. The distance from Philoxenite to Abu Mina is one area to question. According to Drescher s 1946 translation of the Encomium, the distance amounted to about 30-35 km. The text also mentions rest stops 77

LAKE MAREOTIS: REC

STR CTI G THE PAST

where water was available, which the Prefect had arranged every 10 miles or so 14.80 km between the two towns. However, the distance in reality is some 20 km, not enough for even two water stops. et the key argument against the identification with Philoxenite is Grossmann s research on the frequency of pilgrimages to Abu Mina. Pilgrimages to the sanctuary did not peak until the second half of the 5th century, and afterwards surges in the number of pilgrimages accompanied the feast days of St Menas Grossmann 2003: 18 . The ruins currently being investigated on the site are considered earlier the double baths explored by Elakharani 1983 see both Haggag s papers in this volume originate from the end of the 4th century and the pottery kiln under the apse of the basilica operated in the 2nd-3rd centuries AD. The size of this latter facility is unmistakable proof of a flourishing centre existing on the spot possibly Marea prior to the Byzantine foundations, perhaps producing the wine that Mareotis was famous for and expediting it to other towns of the Greco-Roman world in amphorae produced at the site. One of the sherds from the fill of the pottery kiln bore the stamp of a potential owner of the vineyard ιονυσίου Szymańska & Babraj 2005: 130, n. 41 . Many artifacts discovered in the neighborhood of Marea confirm the presence of pottery kilns for firing amphorae in Roman times Empereur & Picon 1998 .

block-joining technique that went out of use in Italy in the 1st century AD, evidently point to an earlier origin Adam 2008: 57 . Moreover, the apsed funerary chapel, discovered by the Polish mission, with its three chambers intended for an affluent family, could not have been erected by and for pilgrims transiting through the town. n the other hand, there can be no doubt that pilgrimages to the temple of the holy martyr must have passed through this area, if anything because of the close proximity to the lake. More evidence is provided by a Late Roman villa rustica, uncovered in Hauwariya village, transformed into a dormitory in the middle of the 6th century and furnished with a small church inside it Rodziewicz 1988: 271-273, ig. 2 . A large urban centre with developed harbour facilities must have been perfectly suited as a stopover for wayworn and surely often ailing pilgrims. D cobert 2002: 129 identifies ancient Marea with the locality of kinj Mary t to the east of the site, where excavations are currently being conducted. et the only arguments in favour of this theory is the strategic position of this locality, affording control of the road from Cyrenaica and Egypt to Alexandria and being accessible also from the opposite side of the lake thanks to a considerable narrowing of its width at this point.

The great jetties are of key importance for the dating of the port. The structural variety of jetty construction different size blocks suggests several building stages, but the characteristic imprints left by swallow-tail metal clamps, a

Fig. . Marea. y antine baths.

78

ra n by . arara .

K. BABRAJ & H. SZYMANSKA: MAREA OR PHILOXENITE? Baths Szymańska & Babraj 2008a: 27-83 The freestanding building follows an orthogonal plan, encompassing the baths itself ig. 2 and the well, which is furnished with a s iyah installation drawing water for the baths. The actual building of the baths, covered an area of 642 m2, including two courtyards and was encircled by a wall of stone blocks. The baths itself was built entirely of red brick on a brick foundation ig. 3 . Shops and a latrine lined the north wall on the outside. Colonnaded courtyards were located to the east and west.

The underground part of the building included the service area cellars, two heating furnaces and one for heating water in a boiler, finally, there were four hypocaust cellars: two in the men s part and two in the women s part. The interior decoration of the building consisted of marble pavements, columns with Corinthian capitals, and a few coatings of painted wall plaster featuring a colorful vegetal frieze. The floors and selected walls of pools were faced with marble slabs. This was evidently a luxurious complex and definitely not the only one in town to judge by the solid water channel system to which it was connected.

The building was a double bath, consisting of two uneven parts separated by a brick wall: the southern one, for men, bigger than the northern one for women. Each of these parts was heated by a separate furnace supplying warm air to four hypocaust systems. The entrance for men was in the western facade it led from the courtyard, following widely accepted circulation principles, that is, directly from the apodyterium into the tepidarium, and then into the two caldaria. The women s part was not as extensive, accessible from a rather spacious courtyard to the east side. It consisted of three rooms: apodyterium and two caldaria. ourteen small pools were used for bathing, eight of which were located inside the two parts of the baths, by the outer wall of the caldaria. Two of them were accessible from the courtyard, leading to the men s part. our pools were located outside the baths, symmetrically in the two courtyards, abutting the walls. The pools were differentiated in shape, from semicircular through rectangular to round, depending on the rebuilding phase. Calculations made on the basis of six fragments of brick arches found in the debris permit a reconstruction of the width of baths halls, which were presumably covered with barrel vaults. The hypothetical height of the building, derived from the proportions of the surviving walls, was 3.50 m.

owhere in the building was a clear stratigraphy of the construction and occupation of the baths visible and the fact that the building had collapsed vide brick arches from the vaults lay immediately on the floors precludes precise dating. An analysis of the pottery finds has provided a timeframe for the operation of the complex between the first half of the 6th century and the first 30 years of the 8th century, that is, before the appearance of Islamic Early Lead Glazed wares, of which not unsurprisingly, not one sherd has been found at the site Majcherek 2008 . Arab coins already issued after the monetary reform, that is, after 686, and remaining in circulation for 750 years, confirm the continued operation of the baths after the damages incurred during the Persian occupation and the Arab invasion Malarczyk 2008 . The most intensive use of the baths occurred in the first half of the 7th century, a fact confirmed by both the pottery finds and the prevalence of coins issued by Chosroes II Lichocka 2008 . Sāqiyah Szymańska & Babraj 2008: 85-99 A well operated by a s iyah stood about 5 m north of the baths. The well itself was 5 m deep, rectangular in plan and built of stone blocks igs. 3 & 4 . The water in the

Fig. 3. Marea. Baths and western courtyard. View from the south-west ( photo P. us ek .

79

LAKE MAREOTIS: REC

STR CTI G THE PAST

Fig. 4. M area. Well operated with sāqiyah. View from the south photo by . alarus .

ity for collecting water in case of drought. n top of this installation was a treading circle for the animals turning the s iyah wheel. A very simple hydraulic system based on connected vessels was employed. Water was collected in the reservoir of the s iyah and carried to the baths via a system of terracotta pipes

well came from a still functioning underground spring in its northwestern corner. The mineral content of this water is much more abundant compared to the lake water. South of the well and level with its rim, there was the floor of a reservoir made of fired bricks in waterproof mortar. The south side of this basin, which was filled with water from the well, was reinforced with a solid platform built of stone blocks. A capacious cistern closed off the complex of the s iyah to the north. It was presumably intended as a facil-

s

nly one conduit reaching Basin in the baths south of the iyah see ig. 3 has been preserved by the west wall Fig. 5. M area. Funerary chapel. View from the west photo P. us ek .

80

K. BABRAJ & H. SZYMANSKA: MAREA OR PHILOXENITE? Fig. 6. Marea. Plan of the basilica ( drawn by . arara .

of the baths see plan in Szymańska & Babraj 2008: ig. 1 . The pipe, a fragment of which is visible under the last block of the s iyah reservoir with two blocks of the limestone bedding, ran to the north wall of the baths. The pipe turned into Room B1, which it crossed until it reached the south wall, where it was raised c. 1.50 m in order to empty into Basin . There is every reason to believe that there were more s iyahs operating in the town complex of Marea. A ground survey of the site has identified the remains of at least one other facility of the kind in the area to the west of the baths Szymańska & Babraj 2008: 88 .

Basilica Szymańska & Babraj 2006: 107-117 The most interesting building at the site, however, is a basilica situated on the hill near the longest harbour jetty. It was discovered by M ller-Wiener 1967: 106, n. 16 , former director of excavations at the nearby sanctuary of Abu Mina. But it was Grossmann who first determined the plan and dimensions of the building in 1986 Grossmann 1993: 107-121 . Surprisingly, none of the ancient written sources mention this church, which was one of largest buildings of its type yet known in Egypt see Grossmann 2002, passim .

Funerary chapel Szymańska & Babraj 2008: 177-185 . One hundred meters to the south there were the ruins of a building which were identified as a funerary chapel. It was furnished with an east-oriented apse and three masonry grave chambers containing 23 burials associated with very poor grave goods ig. 5 . They contained both genders of different age, all clearly of family character. It functioned for a little over a hundred years and consequently must have belonged to the town s inhabitants. A Gaza amphora under the floor of the apse verified a 6th century date for the construction of this complex.

It was a squat-shaped building, divided by columns into three aisles, furnished with a wide transept terminating in rounded arms and a relatively tiny apse Szymańska & Babraj 2004: 53-56, 61-64, 2005, 63-67, 2005a, 43-54 . The liturgical rooms were not on the outside, as was the rule at other basilicas, but incorporated into the body of the building on either side of the apse ig. 6 see Grossmann 2002, passim . The baptistery with round baptismal font apparently remained from an earlier building, possibly a small chapel, which preceded the great basilica at the site. 81

LAKE MAREOTIS: REC

STR CTI G THE PAST

Fig. 7. Marea. Apse of the basilica. View from the west. ‘ A’ and ‘ b’ locate the two burials photo by . alarus .

A kind of money exchange building intended presumably for pilgrims visiting the basilica, was discovered beyond the southwestern corner of the building. Rare examples of bronze weights used to verify coin weight were found inside this room Szymańska & Babraj 2005a: 54 .

rated columns of various sizes, all of them in Proconesian marble and probably imported via Alexandria. The interior decoration of the basilica also included pavement mosaics of which small marble cubes found in the building are the only surviving evidence.

Two burial chambers with multiple burials were discovered under the floor of the apse ig. 7, a & b . Anthropological examination identified over 100 individuals: men, women, children and even unborn children. They appear to have been buried there during the invasion of Chosroes II in AD 619 when Persian troops torched Alexandria and ravaged the region.

Amphora Kiln Szymańska & Babraj 2004a: 53-56, 6164, 2004b: 26-28, 2005: 63-67 . A complete surprise awaited archaeologists digging under the church apse. At a depth of 1.80 m below the preserved tops of the walls, a grate belonging to a large amphora kiln was subsequently discovered. The part of the church intended for liturgical practices used this earlier kiln as a foundation ig. 8 . More of the grate appeared once the floor of the burial chambers under the apse had been cleaned.

The rich interior decoration of the basilica included several fragmentary column shafts and Corinthian capitals deco-

Fig. 8. Marea. Basilica. Amphora kiln. View from the west ( photo by J. ucy .

82

K. BABRAJ & H. SZYMANSKA: MAREA OR PHILOXENITE? Bibliography

The kiln had a diameter of 8 m, and the thickness of the grate was 0.50 m. ne of only a few discovered in the ile Delta, this kiln still contained upon discovery the last batch of amphorae. These vessels were dated to the 2nd and 3rd centuries AD. The wall of the kiln still stood 0.93 m high and was built of mud-bricks. The structure of the kiln is not unlike others from the period with pillars under the grate pierced with holes to support easy circulation of hot air El-Ashmawi 1998: 55-64 .

Ancient Sources Athenaeus, The Deipnosophists, Ch.B. Gulick transl. , 1927. Cambridge, MA. Athanasios, Apologia contra Arianos, H.G. pitz transl. , 1934. Berlin, Leipzig. Diodorus Siculus I, Library of History, C.H. ldfather transl. , 1989. Cambridge, MA. erodotus Book II, A.B. Lloyd transl. , 1976. Leiden. History of the Patriarchs of the Coptic Church of Alexandria Patrologia rientalis 1.14, B.T.A. Evetts ed. , 1904. Paris. Horace, Odes and Epodes . Rudd transl. , 2004. Cambridge, MA. John Moschus, Pratum Spirituale = .H. Baynes, 1947, The Pratum Spirituale. O rientalia Cristiana Periodoca 13: 404-414. Justinian, Edict Corpus Iuris Ci ilis. Code Justinianus Vol. II, P. Kr ger ed. , 1877. Berlin. Strabo, G eography. In Strabon. Le voyage en Egypte. U n regard Romaine . oyotte & P. Charvet transl. , 1997. Paris. Thucydides, History of the Peloponnesian War, Ch. orster Smith transl. , 1928. Cambridge, MA. Virgil, Eclogues. G eorgics, H. Rushton airclough transl. , 1999. Cambridge, MA.

C onclusions The site of Marea with its Late Antique agglomeration remains a puzzle, generating considerable controversy as to its chronology and even its very name. The identification of the town ruins lying on the southern shore of Lake Mareotis, just 45 km south-west of Alexandria, is amongst the priorities of a team from the Polish Centre of Mediterranean Archaeology of Warsaw niversity and the Archaeological Museum in Krak w, which have conducted excavations here since 2000, based on a concession granted by the Supreme Council of Antiquities of Egypt that covers an area of 19.7 ha. The Polish expedition has concentrated on the extant ruins, investigating the role and importance of this Byzantine town as a religious center built around a huge basilica. During nine field seasons 2000-2008 , three independent architectural complexes have been investigated: the baths with s iyah-well, a funerary chapel, and the Christian basilica. All the structures were dated to the 6th end of 5th through early 8th centuries. A site survey included an inventory of all the ruins discernible on ground level within the limits of the SCA concession see ig. 1 .

Secondary Sources Adam, J.P., 2008, La construction romaine. Paris. Am lineau, E., 1893, reprint 1973 , G éographie de l’É gypte à l ’époque copte. snabr ck. El-Ashmawi, ., 1998, Pottery Kiln and Wine- actory at Burg el Arab. In .- . Empereur ed. , Commerce et artisanat dans l’Alexandrie hellénistique et romaine. ctes du collo ue d th nes organis par le C R le Laboratoire de céramologie de Lyon et l’Ecole franaise d th nes d cembre 55-64. Athens, Paris. Babraj, K., & Szymańska, H., 2004, Die Vierte Grabungssaison in Marea, Aegypten. Grabkapelle und Basilike. K emet 13.3: 61-64. Babraj, K., & Szymańska, H., 2005, nfte Grabungssaison in Marea, gypten: Basilika. K emet 14.3: 63-67. Babraj, K., & Szymańska, H., 2008, iyah. In H. Szymańska & K. Babraj eds. , Byzantine Marea. Excavations in 2000-2003 and 2006, Marea Vol. 1: 85-99. Krak w. Babraj, K., & Szymańska, H., 2008, Funerary chapel. In H. Szymańska & K. Babraj eds. , Byzantine Marea. Excavations in 2000-2003 and 2006, Marea Vol. 1: 177-185. Krak w. Blue, L., & Ramses, S., 2005, Lake Mareotis Research Project Preliminary Report from Pilot Survey August 2004. Newsletter of the Archaeological Society of Alexandria February : 5-16. Calderini, A., 1980, i ionario dei nomi geografici e topografici dell Egitto Greco Romano Vol. III, fasc. 3. Milano.

This large port operated in Roman and later Byzantine times and perhaps also in the Ptolemaic era. However, the question of the actual identity of the site remains still uncertain. The current excavated ruins of Byzantine date extend along the coast for 1.5 km. The great jetties are of key importance for the dating of the port. The size of the harbour installations, the small but important quantity of Early Roman potsherds, as well as huge sewage network of substantial build, indicates that a large urban agglomeration flourished on the spot from pre-Byzantine times. Philoxenos may have founded his city prior to its abandonment as an urban center due to the silting up of the sweet-water ile channels supplying the lake. However, evidence suggests that a site existed prior to the establishment of Philoxenite, being located next to existing harbour installations that may have been part of Marea. The recent discovery of the pottery kiln under the apse of the basilica that operated in the 2nd-3rd centuries AD, is unmistakable proof of a flourishing center existing on the spot possibly Marea prior to the Byzantine foundations. Perhaps this site was were the wine that Mareotis was famous for was being produced, and from here the product was being exported to other towns of the Greco-Roman world in amphorae produced at the site. 83

LAKE MAREOTIS: REC

STR CTI G THE PAST

De Cosson, A., 1935, Mareotis eing an ccount of the History and Ancient Monuments of the North-West Desert of Egypt and Lake Mareotis. London D cobert, Ch., 2002, Mar otide m di vale. Des B douins et des Chr tiens. In Ch. D cobert ed. , Alexandrie Médiévale 2, É tudes Alexandrines 8: 127-167. Le Caire. Desanges, ., 1978, P riple du Pseudo-Scylax 90 107 M P riples libiques, 7. In . Desanges ed. , Recherches sur l acti it des M diterran ens aus confins de l’Afrique, Collection de l’É cole Franç aise de Rome 38: 404-405. Rome. Drescher, . ed. , 1946, Apa Mena. A Selection of Coptic Texts relating to St. Menas. Cairo. Empereur, .- ., & Picon, M., 1998, Les ateliers d amphores du lac Mariout. In .- . Empereur ed. , Commerce et artisanat dans l’Alexandrie hellénisti ue et romaine. ctes du collo ue d th nes organis par le C R le Laboratoire de c ram ologie de Lyon et l Ecole fran aise d th nes d cembre 75-91. Athens, Paris. El- alaki, Mahmoud Bey, 1872, Mémoire sur l’antique le andrie ses faubourgs et et en irons d cou erts par les fouilles sondages ni ellement et autres recher ches. Copenhagen. El- akharani, .A., 1983, Recent Excavation at Marea in Egypt. In G. Grimm, H. Heinen & E. Winter eds. , egyptiaca re ernsia II as R misch y antin ische Ägypten. Akten des internationalen Symposions 26.-30. September 1978 in Trier, 175-186. Mainz. raser, P.M., 1972, Ptolemaic le andria I. Oxford. Goma , ., 1980, s.v. Kom el-Idris Marea , Lexicon der Äegyptologie III: 647. Grossmann, P., 1993, Die uerschiffbasilika von Hauwar ya-Marea und die brigen Bauten dieses Typus in gypten als Repr sentanten der verlorenen fr hchristlichen Architektur Alexandreias. In N. Swelim ed. , In Memoriam Daoud Abdu Daoud ( Bulletin de la Societé Archéologique d’Alexandrie 45 : 107-121. Grossmann, P., 2002, Christliche Architektur in Aegypten. Leiden. Grossmann, P., 2003, ochmals zu Marea und Philoxenite. Bulletin de la Societé d’Archéologie Copte 42: 13-29. Haas, Ch., 2001, Alexandria and the Mareotis Region. In S. Burns & .W. Eadie eds. , U rban Centers and Rural Context in Late Antiquity: 47-62. Ann Arbor, MI. Kees, H., 1930, s.v. Marea, Mareotis. Real Encyclopedie IV.2: 1676-1678. Kucharczyk, R., 2008, Glass finds from the baths and s iya. In H. Szymańska & K. Babraj eds. , Byzantine Marea. Excavations in 2000-2003 and 2006, Marea ol. : 129-143. Krak w. Lichocka, B., 2008, The early Byzantine coins. In H. Szymańska & K. Babraj eds. , Byzantine Marea. Excavations in 2000-2003 and 2006, Marea ol. : 145-152. Krak w.

Majcherek, G., 2008, The pottery assemblage from the baths and s iyah. In H. Szymańska & K. Babraj eds. , Byzantine Marea. Excavations in 2000-2003 and 2006, Marea ol. : 105-127. Krak w. Malarczyk, D., 2008, mmayad coins. In H. Szymańska & K. Babraj eds. , Byzantine Marea. Excavations in 2000-2003 and 2006, Marea ol. : 153-154. Krak w. Martindale, .R. ed. , 1980, The Prosopography of the Later Roman Empire, II. Cambridge. M ller-Wiener, W., 1967, Siedlungsformen in der Mareotis. Archäol ogischer Anzeiger 82.2: 103-117. Mycielska-Dowgia o, E., & Woronko, B., 2008, Evolution of the natural environment in the region of Marea. In H. Szymańska & K. Babraj eds. , Byzantine Marea.Excavations in 2000-2003 and 2006, Marea Vol. 1: 17-26. Krak w. Rodziewicz, M.D., 1983, Alexandria and District of Mareotis. G raeco-Arabica II: 199-216. Rodziewicz, M.D., 1988, Remarks on the Domestic and Monastic Architecture in Alexandria and Surroundings. In E.C.M. van den Brink ed. , The Archaeology of the ile elta Proceedings of the eminar held in Cairo 19-22 O ctober 1986: 267-276. Amsterdam. Rodziewicz, M.D., 1995, Eco-Archaeology of Ancient Alexandria and Mareotis. In A.A. Hussein, M. Miele & S. Riad eds. , Proceedings of the Seminar on G eosciences and Archaeology in the Mediterranean Countries, Cairo o ember : 127-139. Cairo. Rodziewicz, M.D., 1998, rom Alexandria to the West. In .- . Empereur ed. , Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du collo ue d th nes organis par le C R le Lab oratoire de céramologie de Lyon et l’Ecole franaise d th nes d cembre 93-103. Athens, Paris. Rodziewicz, M.D., 2002, Mareotis Harbours. In Ch. D cobert ed. , Alexandrie Médievale 2, É tudes Alexandrines 8: 1-22. Le Caire. Rodziewicz, M.D., 2003, Philoxenit Pilgrimage Harbor of Abu Mina. Bulletin de la Societé Archéologique d’Alexandrie 47: 27-47. Szymańska, H., & Babraj, K., 2004a, Marea. ourth season of excavations, 2003. Polish Archaeology in Mediterranean V: 53-63. Szymańska, H., & Babraj, K., 2004b, The ancient port of Marea, Egypt. our seasons of excavations by the Polish Archaeological Mission. Minerva April 2004 : 26-28. Szymańska, H., & Babraj, K., 2005, ouilles arch ologiques de Marea en gypte. Saisons 2002-2003. Archeologia LV: 119-130. Szymańska, H., & Babraj, K., 2005a, Marea. ifth Season of Excavations, 2004. Polish Archaeology in Mediterranean VI: 43-54.

84

K. BABRAJ & H. SZYMANSKA: MAREA OR PHILOXENITE? Warne, A.G., & Stanley, D. ., 1993, Late uaternary Evolution of the orthwest ile Delta and Adjacent Coast in the Alexandria Region, Egypt. J ournal of Coastal Research 9.1: 26-64. Winnicki, .K., 2006, Der libysche Stamm der Bakaler in pharaonischen, persischen und ptolem ischen gypten. Ancient Society 36: 135-142. Wipszycka, E., 2008, Remar ues sur l identification de Piloxenité à la lumiè re de fouilles récentes. Paper delivered at the Third Conference on Medieval Alexandria at the Centre culturel fran ais in Alexandria, 8-10 ovember, 2002 unpub. .

Szymańska, H., & Babraj, K., 2006, Polish Excavations in the Basilica at Marea Egypt . Bulletin de la Societé Archéologique d’Alexandrie 45: 107-117. Szymańska, H., & Babraj, K., 2008, Baths. In H. Szymańska & K. Babraj eds. , Byzantine Marea. Excavations in 2000-2003 and 2006, Marea ol. : 27-83. Krak w. Szymańska, H., & Babraj, K. eds. , 2008, Byzantine Marea. Excavations in 2000-2003 & 2006, Marea I. Krak w. Timm, S., 1984, Das christlich-koptische Ägypten in arabischer Z eit. Wiesbaden. Timm, S., 1988, Teil 4 M - P , Reihe B Geisteswissenschaften o. 41/4, Wiesbaden.

85

LAKE MAREOTIS: REC

STR CTI G THE PAST

86

M.-F. BOUSSAC & M. EL AMOURI: LAKE STRUCTURES AT TAPOSIRIS

The Lake Structures at Taposiris Marie-Françoise Boussac & Mourad El Amouri

Introduction Among the harbours of Mareotis, which in recent years have been the subject of renewed attention, Taposiris on the north shore is of particular interest. Amongst many features of interest, the city offers the only example, besides that of Alexandria, of a closed basin which allows controlled maritime traffic management ig. 1 . These structures were briefly mentioned at the beginning of the 19th century by Pacho, who in 1824 devoted a few lines to the dam running from east to west ... built in the south of the city , a structure intended, he said, to prevent floods Pacho 1827 and sketched by Coste1 in 1820 ig. 2 . Thereafter, Breccia 1914 , De Cosson 1935: 110-111 ig. 3 , chsenschlager 1979, 1999 and Rodziewicz 2002 , amongst others, have pointed out the peculiarities of the closed basin, speculating on its chronology and its connection with the wall of the Barbarians which closes, to the west of the city, the very narrow spit of land that separates the Mediterranean from Lake Mareotis at this point, and its relation to the causeway which blocked the lake to the south. At some point in their use, both systems acted as locks and seem related, but we do not know if they are part of the same feature, nor do we know during which period they were used.

When the French Archaeological Mission of Taposiris3 was launched in 1998, it was determined that an understanding of the harbour system would be a research priority. What was the chronology of the currently visible structures? To which phases of the site did they correspond? Because of the dimensions of the harbour basin the eastwest artificial levee or causeway is about 1,700 m long , areas that were deemed likely to provide some answers with limited resources were selected for excavation, supplemented with an environmental study and geophysical survey. Since most of the data acquired has already been the subject of articles Boussac 2007, 2009 , this paper will briefly present the harbour structures and the conclusions reached concerning the southern causeway. The east side of the system will be the main focus as this gives the latest possible date for all the studied lake constructions. Ove rvi ew of the Remains The ancient city of Taposiris, located on the south side of the taenia ridge, is organised into three sectors linked by a network of north-south routes: the upper town with the Brescia terrace and the temple, the middle town, and the lower town on the shores of Lake Mareotis. To the west, the Wall of the Barbarians forms the western limit of the city see ig. 1 .

To answer these questions, the American mission from Brooklyn College carried out several soundings in 1975, during a one month campaign. Anxious to locate the harbour area,2 chsenschlager worked on two sectors, north and north-west of the dug-out channel and the causeway. either gave the anticipated results: in the north one A , an elevated area, interpreted as a lake front, proved to be an accumulation of rubble waste from amphorae workshops, as shown thereafter by Empereur & Picon 1998 . In the north-west one C , American archaeologists concentrated their efforts on what initially seemed to be a warehouse along the shore. They described it as a platform which had been redesigned several times the so-called platform building , they did not understand its function, but they dated it to the 3rd century AD. They also uncovered a sophisticated system of water tanks carrying water towards the harbour from a terrace further north. No research was published to specify the chronology, and no interpretation was suggested unpublished .

The topographical survey of the lower town area carried out in 1999 and 2000 ig. 4 , and supplemented in 2009, extends from the Wall of the Barbarians in the west to Plinthine in the east, and highlights the irregular outline of the lake shore. A partially anthropogenic spit of land promontory divides the lake shore into two marshy plains, one of which, to the east, forms a bay which is interpreted as a harbour basin.4 This basin covers an area of approximately 8 ha and is closed off to the south by a causeway approximately 1,700 m long , which extends east-west in the shape of a ridge, but it is not straight as the sketches of Coste 1820 see ig. 2 or De Cosson 1935: 110 see ig. 3 imply. It is also interrupted by at least six openings that may be deliberate or may be related to the disappearance of the retaining walls which in places supported the north slope of the causeway.

3. Since 1998 La Mission ran aise in Taposiris Magna has been supported by the Commission des fouilles of the French Ministry of Foreign and European Affairs. Authorisation has been kindly given by SCA and wee appreciate their assistance. 4. Analysis by C. laux, PhD niv. Aix-Marseille, in 2008, in a PhD thesis on the geomorphology of Lake Mareotis.

1. Thanks to the Biblioth que Municipale Vocation R gionale in Marseille, especially o lle Colombi , who allowed us to reproduce two sketches by Pascal Coste from their archives. 2. Thanks to the American mission, especially M. Venit, who shared their unpublished documentation esp. Preliminary Report of the 1975 s eason .

87

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig. 1. G eneral map of the site and location of areas under study at Taposiris Magna and Plinthine. © Archaeological Mission in Taposiris.

88

MAFT, French

M.-F. BOUSSAC & M. EL AMOURI: LAKE STRUCTURES AT TAPOSIRIS

Fig. 2. Drawing by Pascal Coste, made in 1820. 2a: G eneral plan of Alexandria to Abousir. 2b: Detail of plan showing the Temple of Taposiris ( 1) , the Tower of Arabs ( 2) , bridge ( 3) , eastern j etty ( 4) and causeway across the lake from the levee to the south shore ( 5) . 2c: Plan and isometric view ( south) of the bridge of Taposiris. © BMVR Fig. 3. Map of the remains of the ancient city of Taposiris by A. De Cosson, 1935 .

89

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig. 4. L ocation of the different sectors of the site of Taposiris that were studied along the lake shore. © MAFT

The Elements of the Lake System The Causeway and the Channel Several operations 1998, 2000-2005 were conducted at the west end of the causeway, on both sides of the channel and near the bridge, where the basin forms a bend: it

expands to the nortwest to form a loop against which is aligned by a series of buildings, whose different orientations and especially superposition indicate different phases. This visible state has significantly altered the previous interpretation of the phases: the channel was dug in the first half of the 2nd century AD Boussac 2009 7 in a location where a densely occupied area was previously developed whose architectural design suggests prosperity. This area was used for storage and trade, judging from the quantity of imported amphorae material found there Boussac 2009 .8 n the north shore Sector 1 there is a row of shops, abandoned at the end of the Hellenistic period un-recovered items left in place perhaps as a result of flooding,9 before the digging of the channel. On the other side of the channel Sector 3 , earth that had been dredged from the lake covered buildings last occupied at the end of the 1st century BC and which had been in use since at least the 2nd century BC. The excavated building has the same

5. Breccia 1914. He only mentions the 1 km-long dyke, running parallel to the hills, and a fine bridge whose construction certainly dates to Roman times. It seems clear that the lake extended to Taposiris, and that the dyke blocked the water in a kind of harbour . 6. Prospecting by C. Benech, C RS. See the map in Boussac 2007: 452.

7. See the analysis by S. Marqui in Boussac 2009. 8. Boussac 2009 see the appendix by S. Marqui and the study by K. Senol . 9. Study suggests during the 1st century BC: see the analysis by S. Marqui in Boussac 2009: 137.

Topographic mapping and geophysical survey revealed the density of structures in the lower town, briefly mentioned as walls and ruins by Coste 1820 and ignored by Breccia 1914 .5 Although the causeway that blocked the lake has virtually disappeared only a stretch of about 10 m could be uncovered , its outlet was located near a warehouse built on the causeway. Most importantly, the geophysical survey revealed in the West Bay and south of the bridge many ruins which are currently silted over.6 These results underscore the extent of changes in the landscape, the mobility of the shoreline and the extent of the harbour area.

90

M.-F. BOUSSAC & M. EL AMOURI: LAKE STRUCTURES AT TAPOSIRIS architectural features as the north shops foundations and architectonic elements in stone, mud-brick walls covered with a coating . It was supplied with drinking water by rainwater collected and discharged into a cruciform tank located to the north. The angle of the buildings on both sides of an east-west axis indicates the prior circulation of water in this area, which was later replaced by the channel. The level of the tank indicates that the site chosen to dig the channel was originally an area not liable to flooding.

harbour traffic to pass, while minimizing the size of the humpback bridge. This would partly explain the asymmetry of the two arches, with a humpback passage which best fits the top of the vaults, which are all of different widths and heights approximately 4.5 m and 4 m . Furthermore, the construction of the bridge can only be explained by the existence of the causeway and channel system. A limited investigation 6.2 m by 3.2 m conducted in the foundation pebble bed radier of the north pier of the bridge, to the flood level of the lake, confirms that the construction of the bridge is contemporary with the digging of the channel, however it does not allow precise dating. Indeed, the stratigraphic study of this area shows that the bridge was built over a level that dates back to the Imperial Roman period it confirms that the channel, the bridge and the closed harbour system can be dated to the Imperial Roman period.

The whole area was disrupted by major works after a period of stagnation and neglect. The objective was to create a navigation canal by digging a channel, and ensure access by strengthening and shoring the banks formed. The channel was dug and the waste was deposited to the north and south creating two sedimentary causeways, of unequal length but of a similar configuration the causeway to the north extends for about 150 m, to the south for 1,700 m.

The East Part of the Lake System The lake basin is closed to the east by a built jetty, which was the subject of two campaigns in 2005 and 2006, following a preliminary phase of topographical mapping in 1999 and 2000. The cancellation of the following two campaigns 2008 and 2009 , due to an exceptional rise in the water level, meant that certain points remain to be determined, in particular the chronology: while the date of abandonment is well attested by the coins and ceramics first half of 7th century AD , uncertainties remain concerning the previous phases.

An anthropic embankment was formed by successive heaps of backfill piled up, stabilised by a gypsum screed. The most extensive works were carried out to the west, where the edge of the promontory needed to be cut back. This explains why there are only two artificial embankment slopes in this sector, to the north and the south, and why the causeway ends near the bridge at a height of some 5-6 m and then drops in irregular steps towards the east. The two sides are asymmetrical: the north slope, which follows the dug-out channel is steeper and supported by a series of retaining walls to prevent the collapse of the earth and therefore seal the passage, the south side has a shallower slope.

The jetty is located approximately 1,700 m to the east of the bridge west entrance of the harbour basin below the Temple of Taposiris. Between the end of the causeway and the tip of the jetty, there is an opening which is currently about 100 m wide. This probably allowed boats to enter the harbour basin from the east. We have not yet observed a structure that could act as a checkpoint like the bridge that closes access to the west or like the causeway running north-south which blocked the lake between the harbour complex and the southernmost bank Boussac 2009 .

The Bridge A dressed stone10 bridge11 guards the west entrance of the channel. Its present appearance has hardly changed since Coste s sketch in 1820 see ig. 2-c : 47 m long including the ramps, and 9 m wide. Its maximum height is 2.55 m. It consists of two massive abutments of equal length 6.02 m for the south abutment and 6.55 m for the north abutment . The space of approximately 8.35 m between the two abutments is divided by a narrower intermediate pier 1.20 m . This pier is off-centre compared to the axis of the abutments, thus forming two passages of different widths 4.10 m to the south and 3.05 m to the north which only allowed access to the channel to small boats Bougia 1996 . Study of the structure leads to several conclusions. First of all, the bridge was probably supported by masonry vaults igs. 5-7 .12 The flatness of the natural terrain required the creation of a passage sufficiently high for the

Excavations have identified several structures: the jetty, which is interrupted by three openings two flushes and a water supply channel which cross its entire width two basins, one square, the other in the shape of a hoop, located inside the dock and attached to the jetty, between the north flush and the water supply channel. The East Je tty The east jetty extends north-south and has a visible length of 230 m and a width varying of 5-6 m.13 It ends in the lake and it is possible to follow its extent for a further

10. Rodziewicz 1998: 102 only mentions a structure . In fact, the bridge was necessary to cross the channel for those coming from the south of the lake using the causeway and following the road north. 11. As is the case for most stone buildings in the region, it is calcarenite, white dune limestone, extracted from the quarries on the taenia. 12. This architectural analysis, and the model ig. 7 are the work of students at M2Pro niv. Lyon II, especially R. Brunier and . imenez, under the supervision of T. ournet C RS, IRAA/HIS MA .

13. It is longer than the jetties built on the south shore at Marea 60 and 40 m long at the site referred to as Kibotos Blue & Ramses 2005: 10 and is similar to structures noticed during prospecting on an island of Mariout, at a site which played a major role in the management of lake traffic ibid., Sites 23, 12 .

91

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig. 5. V iew of the Taposiris bridge from north. ©

MAFT

Fig. 6. ( above) Plan and elevation of the north pile of the bridge, November 2004. L ay out, drawing and DAO by M. El Amouri. © MAFT

Fig. 7. ( left) Model and reconstruction of the bridge in Taposiris Magna ( students of the Master Pro-Lyon II) . © MAFT 92

M.-F. BOUSSAC & M. EL AMOURI: LAKE STRUCTURES AT TAPOSIRIS

Fig. 8. O verall plan and detailed outline of the eastern j etty. Lay out, drawing and DAO by M. El Amouri. © MAFT

20 m or so under water ig. 8 . To the north it gradually disappears first towards the west and then the east. This asymmetry could indicate that the point of attachment of the causeway on dry land is different than at the southern lake end from that at the lake end. urther north, the silting of the area and recent constructions prevent the correct reading of its route. Situated on the axis, about 50 m to the north, is a masonry structure ST900114 which may or may not be related to the jetty.

Fig. 9. The southern end of the eastern j etty. View from north. © MAFT

The jetty ig. 9 is built in blocks of regular size 0.50 x 0.20 x 0.25 m on average arranged in headers, and the design is comparable to that of the harbour structures present at the neighbouring sites at Gamal and useir igs. 10 & 11 . Excavated sections of the eastern jetty revealed four well preserved courses, and two additional courses which have now disappeared. Over its entire length, both sides of the jetty are punctuated with buttresses ig. 12 . These 27 buttresses 16 on the east side and 11 on the west side are maintained on several courses three or four . Every second course has two headers while the alternate blocks are stretchers on their edge. They are on average 0.50-0.60 m wide and protrude by about 0.30 m. They stabilised the entire jetty, while encouraging micro areas of

14. An investigation was carried out in 2006 on this structure. The report on surface ruins shows an angle similar to that of the jetty. If it belongs to the jetty this would alter our interpretation of the extent of the closed harbour basin.

93

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig. 10. J etty at G amal. View from north-west. ©

MAFT

Fig. 11. J etty at Q useir. View from south-east. ©

Fig. . Plan and ele ation of the inner south face of the north ush and buttress. Lay out dra ing and Amouri. © MAFT

94

MAFT

O by M. El

M.-F. BOUSSAC & M. EL AMOURI: LAKE STRUCTURES AT TAPOSIRIS Fig. 13. View of the north ush during excavation, from east. © MAFT

In addition, the east part of the flush is designed in order to circulate as much surface fresh water as possible McCann 2003: 32 16 towards the interior of the harbour basin: a base of grey mortar17 at the base of the projection formed by the paving of the western part, has a slope to the east which hinders the entry of sediment into the harbour basin, while allowing the passage of a stream of fresh water. The western part of the bottom of the flush, paved in worked stone, is thus located more than 0.40 m above the mortar.

sediment accumulation, thus reducing the risk of sapping erosion due to currents and saline seepage.15 It is telling that particular attention was given to the construction of the eastern face of the jetty where the buttresses are more numerous. Indeed one notices a difference in treatment between the east side facing the open lake the outside of the basin that is subject to currents, and the west wall facing the interior of the basin. Investigation of the north flush revealed four courses of blocks to the east whereas there are only two to the west. A row of worked stones and a stone blockage line the east facing, but do not exist on the west side. In addition, the level of the upper beds of the blocks on the east side is slightly higher than that of the west side. These differences in the construction show that the jetty is a rampart against the open waters of the lake located to the east. To the west, the water in the basin is calmer and the construction of the jetty does not require these precautions.

Carved into the faces of the flush conduit on the paved western half, there are several vertical notches with corresponding horizontal grooves in the paving, which allow the recreation of the system of closing and filtering of the flush. It would appear that three of these locking systems could operate simultaneously. They probably held filter sluice gates or simple marteliè res type gates, which allowed the control of the flow of water according to currents and seasons while limiting the entry of sediments to the basin.18 Two twin notches have no corresponding grooves which implies the redesign of the paving and locking systems. A lightweight and retractable system to cross the flush probably existed for passage across the jetty, but no archaeological evidence shows its design.

The North Flush There are two flushes, designed to bring as much running water as possible into the basin and thus prevent silting one near the south tip of the jetty, the other to the north, more than 70 m from the present shore line, only the second flush has been the subject of investigation igs. 12 & 13 . Its east opening is shaped like a funnel with two facing walls built in a quadrant. The west opening of this passage is straight and 1.2 m wide. The axis of the flush is not perpendicular to the jetty but shifted 15 degrees to the north, which probably gives an indication of the direction of the wind and/or dominant currents in the region during antiquity.

16. Mc Cann 2003: 32 the author states that the constant flow of water helps control the temperature, oxygenation and salinity of the water, adding that the proximity of a source of fresh water is an advantage. 17. Grey mortar seems to have been chosen for the bed of the east entrances of the north flush and the water supply channel. However, pink hydraulic mortar is used around the upper areas of the north flush and water supply channel. 18. Does this system indicate seasons with high levels of sedimentation Are these seasons connected to ile floods Indeed, the ile is joined to Lake Mareotis via canals but underground water must also have raised the level of the lake during floods, bringing a great deal of sediment.

15. There is also a risk of erosion of the sides of the causeway facing the channel this was prevented by the stone constructions, the ramp and the retaining wall.

95

LAKE MAREOTIS: RECONSTRUCTING THE PAST Fish Farming Systems ne of the most interesting findings is the discovery of an original system for breeding/keeping fish. At the current stage of excavations, this consists of a water supply channel which crosses the jetty, a hoop-shaped basin and a fish-tank attached to the jetty inside the harbour basin see ig. 8 . This complex, which has possibly not been entirely uncovered, and which shows signs of repairs, indicates intensive fish farming Lafon 2001: 161, 165 , i.e. breeding in artificial pools. rom a typological point of view, these installations built on the lake are one of the types of fish farming installations analyzed in various recent studies, mostly devoted to the Roman world Kron 2008 : basins built by digging in the sand and masonry work Type III see below Lafon 2001: 171 . Plato tells us that fish farming was practiced on a large scale on the banks of the Nile and papyri from the Hellenistic and Roman eras reveal much information about ichthyotropheia, piscinai and other apodocheia Chouliara-Ra os 2003 Brewer & riedman 1989 of which our system is an example. The region of Canope is home to several piscinae dugout of the rocks, some of which have long been known Breccia 1926 Abd El-Maguid forthcoming but the installation uncovered in Taposiris is unique to this day around the shores of Lake Mareotis Lafon 1998: 573 .19 However, the association of harbour installations and fisheries is not surprising and the facilities found at Taposiris are similar for example to those found at Kenchreai dating to around AD 80 Scranton, et al. 1978 :20 basins are created near the jetty Rothaus, et al. 2008 Lafon 2001: 162 21 and connected to one another and with the sea through channels.

tion basin. As with the flush, the grooves holding the filter sluice gates or locking marteliè res were modified during the rehabilitation of the water supply channel. The walls of the water supply channel are built in worked stones bound with pink mortar. The blocks which form the entrance to the channel are chamfered to facilitate the entry of water. It is a simple funnel system. At the entrance, the bottom of the channel, like the bottom of the flush, is made of grey mortar sloping towards the east. The water supply channel, like the flush, might have been covered by a crossing system, but we did not find any trace of it, although we noticed the last course on its north wall was slightly battered. The western part of the south interior face of the channel ig. 14 appears to have undergone repair with small unworked stones which contrast with the usual blocks of worked stone of the walls of the channel. ther repairs are visible in the south wall of the hoop-shaped basin. The meticulous cleaning of the water supply channel led to the discovery of seven coins wedged between the paving stones of the channel, and thus protected from the cleaning work ield nit 9240 . ive of them belong to the Later Roman Empire AE4 and two are Byzantine dodekanoummia of the 6th-7th century AD, including one dated to Heraclius AD 610-641 .23 Thus, the water supply channel must have functioned at least from the 4th century to the 7th century AD.

The Water Supply Channel The water supply channel crosses the jetty perpendicularly over its entire width igs. 14 & 15 . This channel is 0.70 m wide, 6 m long and 0.85 m deep at the point where the third course is still preserved. It consists of a small decantation basin, a ledge and three closing systems with notches carved opposite one another, 0.50-1 m apart: they were used to create a kind of lock, like in Kenchreai Scranton, et al. 1978 .22

Its period of abandonment is suggested by traces of blocking up. Chamfered blocks the same size as those still in place, and a block with a groove identical to that found on the blocks still in place, were discovered during excavations. These blocks thus come from the courses of the water supply channel. The chamfered blocks were placed at the entrance of the water supply channel to block it: they fit perfectly in the entrance of the water supply channel while keeping one layer horizontal, indicating that they were probably deliberately placed there. The backfilling of the channel thus seems a deliberate action indicating the stopping of the water supply or even of fishing, but ensuring the continued functioning of the jetty.

One of these systems of notches still has its corresponding grooves cut in the bottom of the paving blocks ig. 16 . This is the first paving block to the east and acts as a dam to stop the sediment accumulated in the small decanta-

The Hoop-shaped Basin The west end or exit of the water supply channel flows into a hoop-shaped structure about 6 x 3 m the base of which is paved igs. 8, 17 & 18 . It is built in blocks of worked stone, arranged in a single row and preserved to two or three courses. The blocks are arranged in stretchers and their bases are covered with a hydraulic coating that forms a slight slope on the inside edge of the basin see ig. 18 . This poorly preserved coating is only applied at the junction between the base pavement and the elevation.

19. Lafon 1998: 573, n. 3, notes that coastal speculation has caused the disappearance of many installations in the western Mediterranean over the past fifty years. Lake Mareotis is also threatened by intensive property speculation: during our first works on the east jetty area, it was divided up into plots of land, shown by boundary makers. 20. Scranton, et al. 1978. Their construction dates back to around AD 80. 21. Also see Lafon 2001: 162, n. 89, for Cosa: the author notes that in the port of Cosa, “canals joining the lagoon which has various dams, flow directly into the port, and form part of the same project . 22. Scranton, et al. 1978. ote the presence of several sluices per channel, necessary to form intermediary basins and fish traps: the dimensions of the basins 1 x 0.7 m are similar to those found at Taposiris.

23. Study by T. aucher, Appendix 2, infra.

96

M.-F. BOUSSAC & M. EL AMOURI: LAKE STRUCTURES AT TAPOSIRIS

Fig. 14. L ay out, elevations and sections of north and south walls of the water supply channel through the eastern j etty. Lay out, drawing and DAO by M. El Amouri. © MAFT signed so as to create a slope down towards the fish-tank basin.

The hoop-shaped basin, fed with water through the supply channel, is closed by a double sluice gate system, to the north-east and south: to the north-east, vertical and horizontal grooves were cut into the blocks forming the junction between the channel and the basin. To the south, the funnel-shaped end is fitted with the same system, indicating the method of control of the flow of water into the fish-tank, a large square structure enclosed by walls with double facing ig. 19 . The junction was also de-

The north wall of the basin is an extension of the north wall of the water supply channel and seems to be a renovation or addition. Indeed, this wall is not linked to the channel and is built with smaller, roughly hewn stones, some of which are reused. Perhaps the link with the channel did not initially exist, at least in this form. 97

LAKE MAREOTIS: RECONSTRUCTING THE PAST The Fish-tank The fish-tank is a square basin whose interior dimensions are about 8 m x 8 m. It features walls angled north-south and east-west, all of different form. The west face of the east jetty serves as its east wall. The northeast corner of the basin corresponds to the end of the west face of the jetty. Investigations carried out on the extension of this face showed that no repairs had been made at this location and that the jetty had its original form. It appears that the south and west walls of the basin were entirely built under water. The south wall of an average width of 1.5 m, consists of two parallel faces built in stone and filled with green clay sediment 9221 to the east and 9225 to the west . This wall has an opening, built on at least two courses and equipped with a sluice gate system to control the flow of water. The proximity of the flush installed in the jetty provides a fresh water supply to this southern sluice gate Leatham & Hood 1958-59: 265 .24 The position of this pool is therefore also chosen because of the flow of fresh water into he harbour basin. Sluice G ates and Water Flow A system of fresh water flow was set up between the various structures composing this fishery. The flow or the blocking of water, or at least of fish in the case of filter sluice gates, is ensured by a system of sluice gates or marteliè res type gates. These are all designed in the same way: the vertical grooves cut in the facing blocks allow their positioning across passages, while horizontal grooves cut in the paving slabs ensure a good grip. Some of these gates must have let water filter through to ensure a good regeneration of water in the fish tank. However, no archaeological evidence indicates their construction material.25 The size between 0.10-0.20 m wide for the different grooves in the facing or paving blocks suggests that at least the external framework of the sluice gates was made of wood. They may have been solid and entirely made of wood, or with openings, made with a wooden frame and metal grid.

Fig. 15. W ater supply channel through the eastern j etty opening into the hoop-shaped basin. View from east. © MAFT Fig. 16. G rooves and notches in the water supply chanel. View from north. © MAFT

The water from outside enters from two openings located to the south and north of the whole system: In the south, fresh water arrives through the flush located in the jetty. It enters the harbour basin and indirectly feeds the fish tank by the passage located in its south wall. It is closed by a sluice gate installed upstream of the paving inclined towards the fish-tank. In the north, the water supply channel feeds fresh water into the hoop-shaped basin. The arrival of water is regulated by four successive sluice gates, which leave water to flow towards the hoop-shaped basin while preventing fish from escaping. The frequency of these sluice gates also facilitates maintenance work cleaning, repairs, etc. . A rearrangement may explain the position of the sluice gate located to the west end of the water supply channel, which opens directly onto the northeast corner of the hoop-shaped basin. The space partitioned off by this sluice gate and the one before it, is the largest in the channel. The last sluice

24. Leatham & Hood 1958-59: 265: n. 8, concerning the need for a flow of fresh water into the fish tanks the authors cite . Hortensius criticising M. Luculus whose fish tank does not have enough. 25. Sciallano 1997: 17 notes bronze sluices with holes, like those recommended by Columella viii.17,6 because they are fine nets which allow water to pass through while trapping the fish Leatham & Hood 1958-59: 265, n. 9 . But other systems exist in wood or composite wooden frame and metal grid for example , or in stone like at Mochlos Leatham & Hood 1958-59: 275 and plate 63 a .

98

M.-F. BOUSSAC & M. EL AMOURI: LAKE STRUCTURES AT TAPOSIRIS Fig. 17. View of the hoopshaped basin, the water supply channel and the fish tank. ie from north. © MAFT

Fig. 18. The pavement in the hoop-shaped basin. View from north. © MAFT

Fig. 19. The opening between the hoop-shaped basin and the fish tank. ie from south west. © MAFT

99

LAKE MAREOTIS: RECONSTRUCTING THE PAST gate is fitted across and at an angle to the axis of the channel and is therefore directed towards the centre of the loop, which may also be due to a repair. This all suggests that this sluice gate dates from a later date.

9216/9220 filling the hoop-shaped basin, principally uncovered storage containers Amphorae LRA 1, 4 LRA and LRA 5/6 , indicate a final phase of use of the fish tank during the first half of the 7th century. This hypothesis is supported by the absence of Egyptian amphorae Egloff 167, well documented on Alexandrine and Mariout sites from the mid-7th century. Similarly, among the coins wedged between the paving stones on the channel bed, there are two Byzantine dodekanoummia from the 6th-7th century AD, including one dated to Heraclius AD 610-641 see Appendix 2 .

The passage between the hoop-shaped basin and the fishtank, which is also blocked by a sluice gate, is equipped with the same system. Thus, all these adjustments help to prevent the basins from silting up while encouraging the constant flow of water, which is a key element for the survival of fish in captivity Sciallano 1997: 17 . urthermore, the shallow and muddy bed is the perfect environment for fish farming Lafon 2001: 159 .26

This last phase of use of the system also relates to the last phase of the warehouse excavated in 2000 on the causeway, strategically located near the causeway which crosses the lake Boussac 2009: 129 . The finds included three Heraclius dodekanoummia and late amphorae LRA 1, 4, 5/6 and 7 .

The Fish The fish species identified in the ield nit related to the use of the fish tank 9215 and 9220 in particular at this stage of the study are mainly marine fish or fish which live in brackish water Lafon 1998: 575 .27 A preliminary diagnosis carried out using photographs to visually identify the species28 in advance of a comprehensive study, indicates the presence of sturgeons and garfish Belone belone . At the present stage of work, all the excavated structures seem too small to have housed real breeding farms Guest-Papamanoli 1986: 301 . We suppose it was for the temporary storage of live fish, like at Kenchreai.

The intensity of lake activities during late antiquity corresponds to what the texts and archaeology tell us of the role of Taposiris during this period: the identification in 2009 of late antique thermal baths in the middle town, comparable to those at Marea, further strengthens these findings. However, nothing is yet known about the Hellenistic installations, although we are collecting more data on the city from the 2nd - 1st century BC, and even from the end of the 3rd century: the temple, according to Dr. . Hawass, was founded by Ptolemy IV, suggesting that the city s development is linked to or accelerated by this royal intervention Hawass 2008: 29 . However, it implies nothing about lake constructions. evertheless, in the west, the area covered by the backfill from the digging of the channel in the first half of the 2nd century AD Boussac 2009: 137-141 , revealed many imported amphorae Rhodians and Cnidians which show the intensity of trade from the outside via Alexandria, at least between the mid-2nd century BC and the early 1st century BC. Although their fragmentation prevents us from being precise, the site where these containers were found, close to the shore, suggests that we are not dealing with land transport, but transport on the lake Bernard 2009 . This storage area suggests specific installations of which nothing remains.

C hronology In the east sector of the harbour system, the various phases of installation of the fish farming facilities and the chronological link between the visible system and the construction of the jetty are currently difficult to specify. Whereas the west harbour system bridge and channel can be dated to the 2nd century, the state of visible installations in the east remains unclear: the traces of repairs in the system between the hoop-shaped basin and the channel do not make it possible to date its construction, nor that of the visible jetty. All that can be said about the jetty is that it cannot be later than the 4th century. The presence in the filling of the south wall of the fish tank, of amphorae remains LRA 1 and fragments of containers LRA 4, some of which are dated between the last third of the 5th century and the middle of the 6th century,29 provides an approximate dating for its construction terminus post quem .

othing is known about the structures from the beginning of the Imperial era: in describing Taposiris as the gate eisbolè ) of Egypt in his Letter to the Alexandrians in AD 41, Claudius refers to the fiscal and military role of the city, which was probably associated with specific facilities Boussac 2001 . However, the oldest phase of the harbour that we were able to identify is that which includes the digging of the channel, which is not earlier than the first half of the 2nd century AD Boussac 2009 .30 If one considers that the warehouses and the fish farming facilities operated until the 7th century, then one understands to what extent the intermediate stages of evolution still escape us.

However, ceramic and monetary data provide consistent dates for the last phase of use of the system: ceramic material discovered in the last archaeological layer

26. Lafon 2001: 159: breeding requires very sheltered water with a shallow sandy or silty bed: lagoon shores provide the most favourable environment”. 27. Lafon 1998: 575. The author insists that brackish water is favourable for breeding as it helps the development of young fish and attracts fish. 28. S. Cravhino, pers. comm. PhD niv. Aix-Marseille I . 29. Study by D. Dixneuf, I A . See Appendix 1, infra.

30. See S. Marqui s study in Boussac 2009.

100

M.-F. BOUSSAC & M. EL AMOURI: LAKE STRUCTURES AT TAPOSIRIS

Appendix 1 The C eramics in the Harbour Area: the Eastern Je tty ( Sector 9)

Fig. 20. Main types of ceramics found in Taposiris Magna ( Sector 9) during the Late Roman era.

Archaeological excavations carried out in the area of the harbour basin, and more precisely on the eastern jetty Sector 9 uncovered a few ceramics often very fragmentary and eroded. They amount, except for fragments of bodies, to more than 243 sherds, or 119 individuals.31 However, most of them, i.e. more than 66% of the ceramic assemblage MI 79 , consist of storage and transport containers, both locally produced and imported, characteristic of the Late Roman period, dating which is confirmed by the study of coins by Thomas aucher see Appendix 2 this paper .

pereur & Picon 1989: 236-243 ,33 and were used for the storage and transport of wine, and possibly olive oil. Fragments found in Area 9 ig. 20, a are related to the B1 type according to the classification of Late Eastern Amphorae established by Dominique Pi ri 2005 . This form, generally attributed to the 6th and 7th centuries, is characterized by thick rolled rim an edge to mid-neck recalls the moulding of the previous type Pi ri 2005: 75 . The handles are now almost unribbed. The clay is rather dense and a general trend emerges: marl clay, fine texture and a red to light brown section. Inclusions are of large size, and consists mainly of grains of quartz, some white and sometimes red.

Delphine Dixneuf, I A , Cairo

1. The Imported Amphorae and C eramics he LR mphorae32 Among the imported containers at the site, LRA1 amphorae are the second most common type MI 10 . This evidence finds parallel in all Late Roman Egyptian sites, from the second half of the 4th century until the mid-7th century, or slightly beyond. These jars were produced on the southern coast of Turkey, Cyprus and Rhodes Em-

he LR mphorae LRA 4 productions from southern Palestine represent 32.7 of MI, i.e. 39 individuals. The general form of these amphorae ig. 20, b is morphologically close to Pieri’s type B and dated between the middle of the 6th and the 7th century Pi ri 2005: 106-107 . Many of these amphorae have a sandy and gritty marl clay, with a medium dense fabric and contain several grains of quartz, some grey and white particles of various sizes. The section is usually homogeneous, from buff to orange. Shaping and surface treatment are quite coarse, and clay accretions are frequently observed on the neck. The macroscopic charac-

31. All the sherds are analysed using the criteria of the clay and shape to determine the M I Minimum umber of Individuals by context first, then for the whole sector. Rims, handles, bottoms and bodies fragments were counted. NMI or the global minimum, estimated for the sector, is the highest number among the different morphological features, except the bodies. Regarding the amphorae, the number of handles is divided by two. 32. The term Late Roman Amphorae LRA 1 to 7 refers to the typology of Late Eastern amphorae established by Riley 1981: 85-122 for the material found in Carthage.

33. However, only two workshops have been excavated in Cyprus: Paphos and iggy. See Demesticha & Michaelides 2001: 289-296 Demesticha 2003: 469-476.

101

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig. . Ceramics from the filling of the hoop-shaped basin ( FU 9216/ 9220) .

teristics help us to locate the region of production in southern Palestine, especially in the area of Gaza, Ashkelon and Ashdod.34 However, it is interesting to note the similarity between the clay of Palestine amphorae and ceramics manufactured from Mareotic clays.

Sector 9 are morphologically close to Pi ri’s “bag-shaped amphora, type 4 Pi ri 2005: 117, fig. 76 . It is a globular amphora of small size the neck is short and cylindrical or convex, ending with a small round and/or flared rim, with a sinuous profile ig. 20, c . n the Kellia site, rançoise Bonnet-Borel sets the appearance of these containers during the second quarter of the 7th century Bonnet 1983: 442 , however it is not impossible that they started as early as the late 6th century. They are still present during the Fatimid era, with no major change in the shape Vogt 1997: 256, 258 . As regards the identification of site productions, only one group of clay could so far be identified: marl clay, medium density fabric, sandy the section is usually brown to buff. The inclusions consist of several grains of quartz, some grey and white particles of various sizes. The macroscopic examination of the amphorae, which can logically be assumed to be local or at least regional productions, questions the distinction between Palestinian and Mareotic productions. Thus, two hypotheses can be put forward: they are either Egyptian or Palestinian containers exported with the LRA 4 Amphorae. nly macroscopic and petrographic analysis would answer this question.

In addition to transport containers of types LRA 1 and LRA 4, a few fragments of amphorae and ceramics could not be identified with the exception of a fragment of the body of an African sigillata and two elements of Cypriot sigillata. It is a ring-base low bottom and a rim which seems to correspond to Hayes 9B shape and date from late 6th to late 7th centuries Hayes 1972: 378-382 . 2. Egyptian Production Egyptian Production in Marl Clay With the exception of 12 residual fragments MI 6 of AE 3 amphorae, marl clay productions are mainly illustrated by globular or ovoid containers, more commonly referred to as “bag-shaped”, which were produced in many eastern Mediterranean sites, more specifically in the Levant, Palestine and Egypt MI 15 . Containers found in

Quite logically, common ware is illustrated primarily by marl clay products, some with alluvial content, manufactured without any doubt in Mareotic workshops, on the site

34. or a state of question on LRA 4 amphorae and bibliography, see Dixneuf 2005: 54-62.

102

M.-F. BOUSSAC & M. EL AMOURI: LAKE STRUCTURES AT TAPOSIRIS

Appendix 2 The C oins from Sector 9

or at least nearby Majcherek 2001: 60-64 . The shape repertory includes dishes with rounded rim ig. 20, d , some of them with a high-placed carination in their external wall, cups with flattened carination ig. 20, e , fragments of storage jars and vases for liquids and some cooking pots ig. 20, f .

Thomas Faucher, Sorbonne, Paris A number of coins whose study is currently in progress were found during the excavations in Sector 9. After some preliminary restoration, it is possible to make a few remarks here.

Egyptian Production in llu ial Clay These productions from the Delta and/or the ile Valley are only a small part of the ceramics: 17 individuals, i.e. 14.2 of all ceramics identified. Apart from some fragments of three AE amphorae, a bottom and a carinated shoulder of a wine LRA 7 container ig. 20, g have been identified. Common wares are illustrated by bowls ig. 20, h ,35 dishes with rounded rims, and two cooking pots with a C shaped rim dated at the Kellia to the 7th century Egloff 1977: 103, Pl. 52 no. 4 & 6, type 138 and to the first half of the 7th century in Baouit.36

Most of the coins belong to the Late Roman and Byzantine periods. o coin dating to earlier than the 4th century AD was found, whether in or out of stratigraphical context. All the Late Roman coins 12 items are small denominations ca 10 mm commonly called AE4 AE for bronze and 4 for the size . The identification of these coins is not possible until the restoration is completed but it is unlikely that more information about the type and date of these coins will be gained, for two reasons. First, the proximity of the lake water has facilitated the corrosion of these coins which are often in poor condition. Secondly, these coins, commonly dated to the 4th-5th centuries, were often molded coins, unlike struck coins, produced more or less legally to overcome the lack of currency. This technique allowed for the production of a large number of coins in a short time, but makes coins almost illegible, even when they are just produced. Therefore, it is logical that their circulation and subsequent deposition in the ground has erased any information.

3. The Assemblages he Fishery Comple The southern wall of the fish-tank consists of two faces with a filling of green clay 9221/9225 . ew ceramics were found in the filling but they provide a chronology for the construction of the fish-tank and probably the whole complex. They consist of bodies of some LRA 1 amphorae and fragments of LRA 4 containers including a bottom and a rim close to Pi ri s B1 form this form is dated from the last third of the 5th to the middle of the 6th century Pi ri 2005: 105-106 .

In addition to these Roman coins, there are also three dodekanoummia a coin of 12 noummion easily recognisable by the letters IB on the reverse, produced using the same molding process. Among the latter, were identified one coin from the reign of ustin I 518-527 and a second from the reign of Heraclius 610-641 . It is difficult to give a precise date for molded coins since the coin might have been molded later than the date of production of the model.

The ceramic material discovered in the last archaeological layer 9216/9220 of the hoop-net shaped basin consists mainly of storage containers, i.e. LRA 1, LRA 4 and LRA 5/6 Amphorae, and fragments of common wares which might be linked to food consumption by the fishermen ig. 21 . Thus, this material attests to a final phase of use of the basin during the first half of the 7th century. This hypothesis is confirmed by the absence of Egyptian Egloff 167 amphorae, well documented on Alexandrine and Mariout sites from the mid-7th century onwards.

The Alexandrian excavations emphasised the importance of this phenomenon whose magnitude was often underestimated in the past. This manufacturing technique is attested as early as the Ptolemaic era and continues in Roman and Byzantine periods despite its apparent restricted use during the Early Roman Empire in any case it is difficult to say if it was outlawed by the state and successfully repressed.

To conclude, the study of Late Roman pottery discovered at Sector 9 offers some interesting insights into the commercial activities and trade in the harbour area, which was the western customs of Alexandria on the Mareotis Lake since Roman times if not before. In addition to local marl clay productions, the repertory includes several imports, mainly of wine containers from the eastern Mediterranean and some ceramics from the Delta and/or the ile Valley, reflecting the economic vitality of the region and fishing activities.

When fully restored, the coins from Sector 9 will hopefully provide a more precise datation and help discern more accurately the different periods of use of the eastern jetty area.

35. ne should tentatively compare this rim fragments with a series of dishes found at Kellia and dated to the first half of the 7th century however these are of alluvial clay and with a painted decoration on white slip. See Bonnet 1994: 378-379, fig. 231 o. 138 . 36. npublished material.

103

LAKE MAREOTIS: RECONSTRUCTING THE PAST Bibliography Bernard, S.G., 2009, Alexandrian Tainiai and Land Traffic Patterns: A ote on the Amnesty Decree P. Teb. I 5 in Light of the Topography. Z eitschrift für Papyrologie und Epigraphik 168: 265-270. Blue, L., & Ramses, S., 2005, Lake Mareotis Research Project Preliminary Report from Pilot Survey August 2004. Newsletter of the Archaeological Society of Alexandria 15 ( February) : 5-16. Bonnet, ., 1983, Poteries, verres, monnaies, d cors et inscriptions. In R. Kasser & .-M. Al s eds. , Survey archéologique des K ellia ( Basse-É gypte) . Rapport de la campagne 1981: 423-480. Louvain. Bonnet, ., 1994, Le mat riel arch ologique r colt en 1977, 1982 et 1983 aux ou o r er-Roub iy t. In R. Kasser & . Bonnet eds. , Explorations aux Q ouç oût er-Roubâ’ î yât . Rapport des campagnes 1982 et 1983: 349-406. Louvain. Bougia, P., 1996, Ancient bridges in G reece and coastal Asia Minor. PhD thesis, niversity of Pennsylvania. Boussac, M.- ., 2001, Deux villes en Mar otide: Taposiris Magna et Plinthine. Bulletin de la Société franç aise d’égyptologie 150: 42-72. Boussac, M.- ., 2007, Recherches r centes Taposiris Magna et Plinthine, gypte 1998-2006 . Comptes rendus des séances de l’Académie des inscriptions et belles-lettres ( j anvier-mars) : 445-479. Boussac, M.- ., 2009, Taposiris Magna: la cr ation du port artificiel. In . Dumasy & . ueyrel eds. , Archéologie et environnement dans la Méditerranée antique: 123-142. Paris. Breccia, E., 1914, Alexandria ad Aegyptum. Bergamo. Breccia, E., 1926, Monuments de l’É gypte gréco-romaine, I: Le rovine e i monumenti di Canopo. Bergamo. Brewer, D. ., & riedman, R. ., 1989, Fish and fishing in ancient Egypt. Warminster. Chouliara-Ra os, H., 2003, La pê che en É gypte à la lumiè re des papyrus grecs. Ioannina. De Cosson, A., 1935, Mareotis: Being a Short Account of the History and Ancient Monuments of the NorthWestern Desert of Egypt and Lake Mareotis. London. Demesticha, S., 2003, Amphora production on Cyprus during the Late Roman Period. In C. Bakirtzis ed. , De Rome à Byzance; de Fostat à Cordoue. É volution des faciè s céramiques en Méditerranée ( Ve – IX e siè cles) , actes du VIIe congrè s international sur la céramique médiévale ( Thessalonique, 11-16 octobre 1999) : 469476. Ath nes. Demesticha, S., & Michalides, D., 2001, The Excavation of a Late Roman 1 Amphora kiln in Paphos. In E. Villeneuve & P.M. Watson eds. , La céramique byzantine et proto-islamique en Syrie-J ordanie ( IVe - VIIIe centuries AD. J .-C.) . Actes du colloque d’Amman ( 3-5 décembre 1994) : 289-296. Beirut. Dixneuf, D., 2005, Production et circulation des biens Gaza durant l Antiquit tardive: le t moignage des amphores. In C. Saliou ed. , G aza dans l’Antiquité Tardive. Archéologie, rhétorique et histoire: 54-62. Salerne. 104

Egloff, M., 1977, K ellia. La poterie copte. Q uatre siè cles d’artisanat et d’échanges en Basse É gypte. Genve. Empereur, .- ., & Picon, M., 1989, Les r gions de production d amphores imp riales en M diterran e orientale. Amphores romaines et histoire économique: dix ans de recherché. Actes du colloque de Sienne ( 2224 m ai 1986) : 236-243. Rome. Empereur, .- ., & Picon, M., 1998, Les ateliers d amphores du Lac Mariout. In .- . Empereur ed. , Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athè nes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole franç aise d’Athè nes, 11-12 décembre 1988: 75-88. Athens, Paris. Guest-Papamanoli, A., 1986, Arch ologie, ethnographie ou ethnoarch ologie des ressources marine de sites cotiers. Le cas de la pê che aux muges dans les lagunes de la Gr ce ccidentale. In . Gerard ed. , VIè mes rencontres internationales d’archéologie et d’histoire d’Antibes. L’exploitation de la mer de l’Antiquité à nos j ours. II: La mer comme lieu d’échanges et de communication : 281-303. Valbonne. Hawass, ., 2008, In search of Cleopatra and Mark Antony. Horus August/September : 26-29. Hayes. .W., 1972. Late Roman Pottery. London. Kron, G., 2008, Reconstructing the Techniques and Potential Productivity of Roman Aquaculture in the Light of Recent Research and Practice. In E. Hermon ed. , Vers une gestion intégrée de l’eau dans l’empire romain: 175-185. Rome. Lafon, ., 1998, Piscinae et pisciculture dans le bassin occidental de la M diterran e. J ournal of Roman Archaeology 11: 573-582. Lafon, ., 2001, Villa Maritima, recherches sur les villas littorales de l’Italie romaine ( 3è me s. av. J .-C. / 3è me s. ap. J .-C.) . Rome. Leatham, ., & Hood, S., 1958-59, Submarine Exploration in Crete, 1955. The Annual of the British School at Athens 53-54: 263-280. El-Maguid, A., forthcoming, Maamourah Surveys: Preliminary report. In H. Tzalas ed. , Tropis VIII: Eighth conference on ship construction in antiquity, Hydra, G reece, August 2002. Athens. Majcherek, G., 2001, Marea 2001: ote on the Pottery. Polish Archaeology in the Mediterranean 13: 60-64. McCann, A.-M., 2003, he Roman Port and fishery of Cosa, a short guide. Rome. chsenschlager, E.-L., 1979, Taposiris Magna: 1975 season. First International Congress of Egyptology: 503-506. Berlin. chsenschlager, E.-L., 1999, Taposiris Magna. In K.A. Bard ed. , Encyclopedia of the Archaeology of Ancient Egypt: 759-761. xford. Pacho, .-R., 1827, Relation d’un voyage dans la Marmarique et la Cyrénaï que. Paris. Pi ri, D., 2005, Le commerce du vin oriental à l’époque byzantine ( Ve – VIIe siè cles) . Le témoignage des amphores en G aule. Beyrouth.

M.-F. BOUSSAC & M. EL AMOURI: LAKE STRUCTURES AT TAPOSIRIS Riley, .A., 1981, Pottery from the Cisterns 1977.1, 1977.2 and 1977.3. Excavations at Carthage 1977 conducted by the U niversity of Michigan 6: 85-122. Rodziewicz, M.D., 1998, rom Alexandria to the West by land and waterways. In .- . Empereur ed. , Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athè nes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole franç aise d’Athè nes, 11-12 décembre 1988: 93-103. Athens, Paris. Rodziewicz, M.D., 2002, Mareotic harbours. In C. D cobert ed. , Alexandrie médiévale 2: 1-22. Le Caire. Rothaus, R.M., Reinhardt, E.G., & oller, .S., 2008, Earthquakes and Subsidence at Kenchreai: sing Recent Earthquakes to Reconsider the Archaeological

and Literary Evidence. In W.R. Caraher, L. ones Hall & R.S. Moore eds. , Archaeology and History in Roman, Medieval and Post-Medieval G reece: Studies on Method and Meaning in Honor of Timothy E. G regory: 53-66. Aldershot. Sciallano, M., 1997, Poissons de l’antiquité. Catalogue d exposition, 1er avril-31 ao t 1997. Mus e d Istres. Scranton, R., Shaw, .W., & Ibrahim, L., 1978, K enchreai Eastern Port of Corinth, Vol. I. Leyde. Vogt, C., 1997, Les c ramiques ommeyyades et abassides d Istabl Antar- ostat: traditions m diterran ennes et influences orientales. In G. D mians d Archimbaud ed. , La céramique médiévale en Méditerranée. Actes du VIe congrè s de l’AIECM2 ( Aix-en-Provence, 13-18 novembre 1995) : 243-260. Aix-en-Provence.

105

LAKE MAREOTIS: RECONSTRUCTING THE PAST

106

M. BERGMANN, ET AL.: SCHEDIA ON THE CANOPIC NILE

Schedia, Alexandria’s Harbour on the Canopic Nile. Interim Report on the German Mission at Kom el Giza/Beheira (2003-2008) Marianne Bergmann, Michael Heinzelmann & Archer Martin Historical Sources and Research History Shortly after the foundation of Alexandria, the new metropolis was connected to the Canopic ile by an artificial, 30 km long channel. At the point where the canal branched off the Canopic Nile a new town, Schedia, was founded (see in general: RE 1921: 401-403; RE 1899: 2030; Fig. 1).1

to display them. At the same time they are an indication of the formation of a prosperous citizenry with an understanding of itself as inhabitants of Schedia. In Roman times Christianity seems to have established itself early under Alexandrian influence. Schedia received its own episcopal seat. In spite of its importance the site as well as the Nile Branch, is still virtually unknown. The river and the canal have now vanished from the landscape. Schedia itself was still a prominent stretch of hills at the beginning of the 20th century with a length of about 2.5 km and a width of 1.5 km. Intensive sebakh extraction and the increasing encroachment on the ancient settlement site by the villages of Kom el Giza, Kom el Hamam and Kom el Nashwa have, however, made the topography so unclear that the place is hardly to be noticed in the landscape. Only in a few places can the flat remains of the koms and single structures be recognized. The greatest height in the territory of Kom el Hamam is occupied by the present cemetery.

In its important river harbour all goods coming from Upper Egypt had to be transferred to smaller vessels. It was also a customs station wherefore a pontoon bridge was installed (σχεδία), which gave its name to the place. Schedia seems to have been a flourishing Greek-dominated polis throughout Ptolemaic and Roman times and was inhabited until the early Byzantine period. Ancient sources suggest the existence of ample harbour and storage installations with the personnel and administrative infrastructure that belonged to them (Strabo, G eogr. 17,1,16; Procopius, Aed. 6,1). The inscriptions found in the area of Schedia provide further evidence of the character of the place (Bernand 1966: 45 ff., 1970a: 329 ff.; Botti 1902). Thus, cults of Zeus Soter and Athena Polias are attested, along with a ship and military station already in the late 4th century BC. About 115 BC the garrison station at Schedia dedicated a Kleopatreion. These cults, as well as early Ptolemaic grave inscriptions with Greek names, appear to indicate a high percentage of Hellenic inhabitants or at least of strongly Hellenized people in the city founded ex novo. There was, however, a Jewish colony as early as the time of Ptolemaios III Euergetes (246-221 BC) with one of the oldest known synagogues attested by an inscription – evidence of the early multicultural character of this harbour city. In the Roman Imperial period Schedia calls itself a polis in inscriptions and apparently has its own magistrate (archon). Various honorary statues and bases for private citizens and emperors found in the area of Kom el Giza suggest that there must have been appropriate public spaces in which

After the first vague attempts to localize Schedia in the 18th century, it was the Napoleonic expedition of 1798-1802 that first concerned itself with Schedia and the Alexandrian canal (Jomard & Jacotin 1818/25; Bernand 1970b: 422). In this connection the place is shown on the first accurate map as one of largest tells in the Nile Delta. A participant in the expedition, ean Michel Le P re, was the first to propose the localization of the beginning of the canal at El-Karyun and the identification of the group of mounds of ashwa and Kom el Giza as Schedia. Subsequently, Ptolemaic-Roman finds begin to be reported in the area later identified as Schedia, mostly chance finds by farmers or people engaged in quarrying sebakh. These objects, which include a series of honorary and funerary inscriptions and a Nilometer, finished in part in private collections but mostly in the museums of Cairo and Alexandria.

1. This project is being carried out by the Institute for Classical Archaeology of the University of Goettingen in cooperation with the Archaeological Prospection Service of the University of Southampton. Further support came from the German Archaeological Institute, the Department of Geodesy at the University of Stuttgart, the Geology Department of Mansoura University and the Centre d’Études Alexandrines. The project has been funded first by the ritz Thyssen oundation and since 2004 by the German Research Foundation. We thank the Supreme Council of Antiquities and its Director General, Dr. Zahi Hawass, for the permission to work at Schedia. For help and support we thank the head of the Antiquities Service for the Department of Beheira, Fawzi el Choulani, and the director of the SCA’s excavations at Schedia, Ahmed Abd el Fattah.

It was only in the 1980s, though, that the first regular archaeological work was done at the site, when the Egyptian Antiquities Service carried out some rescue excavations because of the growing threat to the ancient site through modern construction (Abd el Fattah 1988, 1998, 20002001). Several areas of varying sizes were investigated on the eastern edge of Kom el Giza and at Kom el Hamam, that is in the eastern part of the ancient settlement (see Fig. 4). Numerous building structures of the late Hellenistic, Roman and Byzantine periods were uncovered. A high number of Hellenistic objects were found, among them numerous coins. The excavated areas were restored and protected by the Egyptian Antiquities Service. 107

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig. 1: Alexandria and its hinterland in 1925, Survey of Egypt, sheet 4 ( detail) . The N ew Ev idence Since 2003 thanks to the Supreme Council of Antiquities (SCA), the Institute of Classical Archaeology of the University of Goettingen has conducted an interdisciplinary project consisting of a topographical survey, large scale geophysical surveying, geological investigations, selective stratigraphic excavations and a comprehensive study of the finds Bergmann & Heinzelmann 2007 .

boundaries can also disclose characteristics of the ancient topography, because some of them seem to document ancient boundaries, like the former edge of a river branch. G eophysical and G eo-archaeological Investigations Three seasons of geophysical surveys have been conducted at the site since 2003 by the University of Southampton, to assess the archaeological remains of the area surrounding the Schedia excavations (Figs. 2 & 3). The surveys were carried out using magnetometry. Furthermore, in 2005 a series of deep drillings and sedimentary analyses was carried out in collaboration with the Geology Department of the University of Mansoura, which will give further evidence of the course of the Canopic Nile and the canal.

Topographical Survey In order to integrate old maps and the results of the excavations, geomagnetic and geological investigations into a unified GIS system, it was necessary to establish a raster of fixed points for the whole site investigated. After the introduction of this geodetic net, relevant points of the newly established fixed point system were measured using a real time GPS to determine the parameters of transformation for the implementation into the Egypt Blue-Belt-system. Subsequently an up-to-date cadastral map was created on the basis of recent Satellite images (Fig. 2). This mapping of the modern surroundings of ancient Schedia was indispensable for several reasons. With regard to the planning of the geomagnetic surveying and the stratigraphic sondages, detailed knowledge of the modern topography especially of the canals, streets and field borders is of great importance. Furthermore, it is possible to draw some conclusions about the ancient landscape from its current shape. This holds true for the relief of the terrain, which reveals areas of settlement as well as the course of the probable harbour basin, but to a certain degree modern field

At present these results seem to suggest that the modern Kanubiye Canal follows the south-western embankment of the ancient Canopic Nile which may have passed between Kom el Giza and Kom el Hamam. In that case, its breadth amounted to 150 m. As there is also clear evidence that there was once a lagoon or a second branch of the river to the north of Kom el Giza, it can be assumed that the settlement of Kom el Giza actually was isolated on a kind of long and narrow island. Maybe it was this topographical situation which gave the stimulus to construct a pontoon bridge here. The investigations have also revealed a possible localisation of the Alexandrian Canal with a strict east-west orientation between Kom el Nashwa and Kom el Sherif and a 108

M. BERGMANN, ET AL.: SCHEDIA ON THE CANOPIC NILE

Fig. 2: Schedia. Topographical map based on quickbird satellite image ( 2005) ( A) and corings ( H) . G erman Mission at K om el G iza/ Beheira.

indicating excavated areas

Fig. 3: Schedia. Map with reconstruction of the ancient topography. G erman Mission at K om el G iza/ Beheira.

109

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig. 4. Sc hedia. Plan of the southeastern settlement ( K om el G iza, K om el Hamam) with excavation areas 1980-92 and 2003-06. G erman Mission at K om el G iza/ Beheira. possible harbour basin immediately north of it. A second canal could have existed between Kom el Hamam and Kom el Sherif, branching at a right angle from the Canopic Nile and then turning in a westward direction. We thus appear to have a complex system of canals.

the architectural remains documented as a first step. Secondly, some stratigraphic trenches were excavated in order to learn more about the chronology and function of the structures. Furthermore, a series of new stratigraphic sondages have been carried out in previously untouched areas.

As for the extension of the ancient city, it is quite clear that the ancient settlement once occupied the whole area around the possible harbour basin (today’s Nashwa), the whole southern bank of the former Canopic Nile for a length of ca. 2.5 km (today’s Kom el Sherif, Kom el Hamam, el Karyun), as well as parts of the northern bank (today’s Kom el Giza). However it seems that the built up areas did not reach a significant depth but stuck rather close to the edges of the waterways, because of the necessities of a river harbour with its ongoing transhipment of goods. Thus, the picture that emerges is that of a rather amorphous and unhomogeneous settlement, consisting of a series of elongated narrow strips along the waterways, which also separated the different parts of the settlement from each other. Its nucleus has to be assumed in the vicinity of the harbour, from where it probably grew step-by-step without over-all planning along the river embankment. Unfortunately, the former kiman in the area of the harbour have suffered most from sebakh extraction, while we have to assume that the better preserved areas of the former settlement at Kom el Giza and Kom el Hamam were actually on the outskirts.

Summing up the results the following picture appears. In the western part of the older excavations (Area 3), there is a double-tholos-bath with bath-tubs to sit in (Fig. 5). It belongs to a Hellenistic type, which is well attested in the Nile Delta, the Fayum and other parts of Egypt, where it was used as a Hellenistic heritage much longer than anywhere else in the ancient world (see Boussac, et. al. forthcoming) A special feature of the Schedia bath is the visible division between the baths for men and women, otherwise attested in the contemporary papyri. Coin finds and the deep foundations suggest that the bath was built in Hellenistic times, but it is one of the rare examples whose long duration can be shown. Provided with additional rooms with bath tubs for immersion and with new high quality terrazzo floors, it was used at least until around AD 200 (Bergmann & Heinzelmann forthcoming). In its immediate vicinity, still in Area 3, the remains of a huge foundation are visible. Its dimensions seem to indicate that they belonged to a big public building, maybe a temple. Its facade was probably oriented westwards to the Canopic Nile. Unfortunately, however, this part is inaccessible because of modern constructions. Again all findings in this area, including the coins, indicate a (late?) Hellenistic date. Both buildings, the round baths and the possible temple foundation, consist of fired bricks, which is quite unusual for Ptolemaic architecture outside Alexandria.

Excavations at K om el G iza ne important aim of the field archaeological activities was to explore further the parts of the town found during the SCA’s rescue excavations in the 80s and early 90s (Fig. 4). Therefore, these areas were cleaned and all 110

M. BERGMANN, ET AL.: SCHEDIA ON THE CANOPIC NILE

Fig. 5: K om el G iza, Area 3. Plan of Hellenistic baths. G erman Mission at K om el G iza/ Beheira. Fig. 6: K om el G iza, Area 1. Plan of Roman villa, tombs and later basins. G erman Mission at K om el G iza/ Beheira.

111

LAKE MAREOTIS: RECONSTRUCTING THE PAST Fig. 7: K om el Hamam. Plan with excavations areas 2003-2006. G erman Mission at K om el G iza/ Beheira.

Fig. 8: K om el Hamam. Plan of Area 6 with Roman storage building and later phases. G erman Mission at K om el G iza/ Beheira.

112

M. BERGMANN, ET AL.: SCHEDIA ON THE CANOPIC NILE

Fig. 9: K om el Hamam. Reconstruction of the Roman storage building ( J . Schumann) . G erman Mission at K om el G iza/ Beheira.

While these two structures of Area 3 seem to be in a certain sense public buildings, Areas 1-2 farther to the east reveal a completely different picture (Fig. 6). In both areas the earliest traces of use were numerous constructions that are possibly to be interpreted as funerary monuments dated on typological grounds to the Roman period. Especially one type of monument is very characteristic: a stepped pyramidal structure of ashlar blocks, probably with a pillar on it, which has close parallels in the necropoli of Alexandria and Marina el-Alamein (Daszewski 1998).

particular interest is a large building measuring 70 x 40 m, which is visible on the surface of Kom el Hamam. Three stratigraphic sondages were carried out down to the foundations of the building (Figs. 8 & 9). They show that it was constructed around AD 100 above earlier mud-brick structures, which were demolished for the purpose. The building was erected in an interesting building technique: the supporting structure consisted of 10 x 8 rows of massive brick pillars with cores of opus caementicium, which bore massive cross vaults of caementicium, collapsed fragments of which were found. It was an enormous fired brick building, particularly by Egyptian standards, that also demonstrates an in-depth technical knowledge of Roman buildings from central Italy of the period. However, two features indicate that the responsible architect was not really familiar with this building technique: contrary to the Roman prototypes, all pillars were constructed without foundations, instead of which huge mudbrick pylons were added already in the original phase of the building on the outside in order to absorb the pressure of the vaults. Furthermore, mud-brick walls running in a north-south direction were constructed between the pillars, which means that the building was separated into a series of nine parallel naves each ca. 5 m wide, 40 m long and 5.5 m high. All these naves opened on the south side through huge doors. The floor consisted of a simple mud-brick pavement. In the middle of the naves long low benches (ca. 0.25 m high, 1.50 m wide) made of mudbricks, seem to have been installed. Unfortunately, no specific finds were found that could give an indication of the original function of the building; obviously it was completely emptied before the changes of the second phase. However, judging by the typology of the building, it most probably formed a large storage facility for some kind of precious goods.

One of the less well preserved pyramidal structures has been investigated. Its core consisted of a cube of fired bricks. Around and underneath these bricks a stratum containing ashes, bones and pottery has been found. All bones belonged to animals of different types, such as at least four jaws of sheep as well as a complete skeleton of a young goat missing the head. They do not show signs of burning. However, a certain portion of the bones concentrated in one area was burnt and was in a very fragmentary state of conservation. They have yet to be analyzed, but it seems likely that they belong to a human cremation, while the unburnt animal bones might be connected to some funeral rites. Because of a Trajanic coin found within the ash layers this structure probably belongs to the 2nd century AD. The area must be regarded as suburban cemetery. At a somewhat later time, an isolated villa was built apparently in the midst of the sparse graves of Area 1 (Fig. 6). This villa of the middle Imperial period, was organised around a central atrium-like courtyard and was furnished with mosaics and opus sectile floors. It was in use for a long time and repeatedly rebuilt and renovated. During the 4th century AD the villa was abandoned, and numerous basins were installed over its ruins and the neighbouring graves, as well as in most of the other areas excavated up to now. Undoubtedly, these were production units that were used intensively and were repaired repeatedly up to the 6th century.

Later, the building underwent several different phases of use and renovation. Only a short time after its erection, additional mud-brick walls were added during the later 2nd and 3rd centuries AD in order to create simple apartments. These habitations show two main phases of use. In one of the sectors several rooms were examined, some of them containing cooking installations. They revealed a huge amount of pottery, mainly amphorae, as well as coarse and

Excavations at K om el Hamam New archeological examinations were carried out at Kom el Hamam and other parts of the ancient city (Fig. 7). Of 113

LAKE MAREOTIS: RECONSTRUCTING THE PAST Fig. 10: K om el Hamam, Area 9. Roman enclosure wall. G erman Mission at K om el G iza/ Beheira.

cooking wares, mostly of local production. In the late 4th or early 5th century AD these habitations were abandoned and levelled in order to install a granary. On a level ca. 1 m above the pavements of the prior habitations, a special brick floor was constructed supported by some dozens of small parallel vaulted chambers – a typical Roman system for keeping grain dry. This grain storage facility was in use until the late 5th or 6th centuries AD, when it was destroyed by fire. n its ruins simple dwellings and agrarian installations were installed, which were in use until the Arab conquest.

Once again in the 4th century, a radical change took place. As in the case of the villa suburbana at Kom el Giza, most of the earlier buildings at Kom el Hamam were destroyed or changed their function. Instead of the dwellings or monumental buildings, a whole series of simple working installations occur. Among these are several types of basins, which seem to have served various purposes. Some have the characteristic arrangement for wine production with a higher platform for treading and a lower catchment basin accompanied by further basins for fermentation (Fig. 11) (Rodziewicz 1988). Others are isolated or have furnaces nearby and appear to have served other productive processes that cannot be identified at present. Aside from their function, the basins testify to an intensive agricultural use of the territory in late antiquity. These agricultural installations were in use until the beginning of the 7th century, when Schedia was completely abandoned at the time of the Arab invasion.

Immediately south of this huge building, the remains of another monumental structure was uncovered (Figs. 7 & 10). At a distance of ca. 4.5 m a huge wall runs parallel to the southern front of the above-mentioned storage building. The wall has impressive dimensions: with a thickness of about 2 m its foundations consist of a massive foundation of opus caementicium and three layers of huge ashlar blocks, each with a height of 0.45 m, while the elevation consisted of two more layers of ashlar blocks and a massive mud-brick wall on top. The stratigraphic evidence shows that the wall was constructed shortly before the huge storage building, probably in the first half of the 1st century AD. Its function is not yet clear. The dimensions as well as the massive construction technique indicate that the wall must originally have reached an impressive height. Therefore, it seems probable that the wall actually formed a high enclosure belonging maybe to a temple or some other kind of public building. Further investigations showed that the wall actually belonged to a much bigger structure, consisting of at least two rectangular courts. Strangely enough the northern half of it was abandoned shortly after its construction and substituted by the huge pillar building, while the southern part was continuously in use until late antiquity.

Analysis of the Finds The detailed analysis of our excavated material, especially coins and pottery, will give further information in the future concerning Schedia’s economic development. However, because of the huge amount of finds we are just at the beginning. Coins Around 4,000 coins found at the ancient site of Schedia have been analyzed by Ch. Noeske. Among them are 1,400 stray and stratified finds from the recent excavations. n the whole they cover the period from ca. 320 BC (Ptolemy as a satrap) to the time of Arab conquest (Heraclius, AD 641). Hellenistic coins from the SCA excavations at Kom el Giza are numerous, while in the excavations on the northern side of the Kanubiye Canal (Kom el Hamam, Kom el Sherif) the Roman Imperial and late-antique coins 114

M. BERGMANN, ET AL.: SCHEDIA ON THE CANOPIC NILE prevail. The analysis of these coins has different aims. The dates they offer for the stratigraphic units are important, but they are also analyzed in their own right, as sources for minting policy, monetary circulation and its regions, and for economic change. Comparison to the few other places where coins have been noted, such as Elephantine, Abu Mina and parts of Alexandria, is informative in this respect.

phorae make up approximately 2/3 or 3/4 of the assemblages, which is a characteristic percentage for Mediterranean sites that engage actively in commercial exchange, while in the latter amphorae are attested at about 1/3. Consideration of some of the specific functional groups confirms the difference between the earlier and the later layers. The characteristics of the fine wares and amphorae vary markedly between the two periods. There are so few lamps that it is not possible to speak of trends in them. The utilitarian wares (coarse and cooking wares, which at Schedia appear in the same fabric) are essentially of local production throughout, and their typological development remains to be determined.

Pottery At Schedia the programme of pottery study aims at as complete a documentation as possible. The protocol used is predicated on the idea that all fragments, even typologically unidentifiable body sherds, have some information to give. Therefore, all the sherds per stratigraphic unit are attributed to a functional group and fabric, counted and weighed. For rim sherds the percentage of the diameter preserved is also registered, in order to calculate estimated vessel equivalents. Diagnostic fragments are, of course, classified typologically, and chronologically sensitive items are used for dating purposes. This degree of quantification permits our material to be compared with other assemblages throughout the Roman world, where various standards prevail. It also allows us to make statistical considerations at various levels (Martin 2008; Martin forthcoming).

Among the fine wares in the earlier layers it is noticeable that Schedia did not participate in the fine-ware boom of the Augustan period, when most parts of the Roman world wanted to import or produce red-gloss wares reflecting the repertoire of Italian sigillata, that is to present themselves as Roman at the table (Martin 2008: 268). There are some very occasional imports of Eastern Sigillata A and less frequently yet of Italian sigillata, but for the most part Schedia seems to continue to employ red and black-slipped wares in the Hellenistic tradition. In the late-antique layers a certain quantity of Cypriot and also of African Red-Slip Ware appears, and a larger amount of Egyptian products inspired by them is attested. It is only then that one can say that Schedia participates in the Mediterranean-wide trends in fine wares.

A first level concerns the overall composition of the assemblages by functional groups. There is some evidence that the percentage in particular of transport amphorae with respect to fine table wares, lamps, cooking and coarse wares, reflects the degree of integration of a site into trade networks. Here Schedia shows a decided difference between the late-antique layers up to the abandonment no earlier than the 7th century and the earlier ones. In the former, am-

In the earlier levels Egyptian amphorae greatly outnumber imported ones – the majority were the Amphore Egyptienne 3 type which were first developed at the beginning of the Empire. Other examples included the earlier Amphore

Fig. 11: K om el Hamam. Plan of Area 8 with basins for wine-production. G erman Mission at K om el G iza/ Beheira.

115

LAKE MAREOTIS: RECONSTRUCTING THE PAST Egyptienne 2 and the contemporary Amphore Egyptienne 4 types (Empereur 1998: 79). They appear most frequently in the same fabric as the utilitarian vessels, decidedly more rarely in the northwest coastal fabric associated with production around Maryut, and only occasionally in the Nile silt fabric. Imported amphorae are represented by a diffuse scatter of various Aegean and Levantine pieces. In late antiquity, on the other hand, the provenience of the amphorae is much more focused and the percentage of imported vessels much higher. Egyptian vessels appear in the Nile silt fabric associated with Egloff 172 and the Carthage Late Roman Amphora 7, while the imports come almost exclusively from Cilicia with the Carthage Late Roman Amphora 1 and Gaza with the Carthage Late Roman Amphora 4.

During the 2nd and 3rd centuries the city reaches its maximum extension. That the brick-pillar building investigated here was transformed for residential purposes contrary to its original destination, can be considered a further indication of population pressure and of an increasing process of urbanization. Parts of the city reveal a strong Greek tradition (the Hellenistic bath and the stepped bases of grave monuments and Roman influence villa suburbana). Other monuments, like the pillar building with its contemporary use of fired and mud-bricks show an interesting mixture of imported and local building traditions. At the same time, Schedia appears to have been a city with interesting cultural contrasts: while a prosperous upper class lived in houses of clearly Roman influence, a large number of inhabitants lived in simple dwellings of local Egyptian character.

In short, Schedia in the Imperial period presents, paradoxically for the major river harbour of Alexandria of which the written sources inform us, the picture of a site little integrated into the exchange networks and cultural trends of the time, while late-antique Schedia appears to participate more actively in them. How to explain this situation remains to be seen. Could it be that in the earlier period little needed to be offered in exchange for Egyptian products (e.g. grain collected in tax, stone from the Imperial quarries), while late-antique Egypt was in a similar situation to other provinces, with imports and exports. In any case, merit is due to the pottery study that serves to highlight a line of research to be pursued that might not have been highlighted in any other circumstance.

In late antiquity the explored parts of the city underwent a fundamental change. They seem to lose a large part of their urban character with the transformation of the older buildings into productive installations, of rural and other determination. At the same time the new installation of a granary in the pillar building coincides with the undiminished export of grain to Constantinople. How these changes are to be interpreted against the background of new discussions about the flourishing agriculture and renewed trade and imports of late antique Egypt (Banaji 2001; Kingsley & Decker 2001) and against the situation in Alexandria itself, remains to be seen. Although the new transformation of the pillar building into a large granary points to the still uninterrupted importance of the city as a supply base, it changes into a centre for agricultural production that possibly still has regional importance in supplying Alexandria but appears to a large extent to lose its role as an emporium, perhaps because of the overall decrease in trade. This hypothesis, which for now can take only the late Ptolemaic and Roman settlement of Schedia into account, needs to be checked by further investigation and especially to be complemented by a greater consideration of the Ptolemaic phases.

C onclusions The following preliminary hypotheses concerning the development of Schedia seem possible. After the foundation of Alexandria and the construction of the Canopic Canal the early Ptolemaic settlement of Schedia existed probably around an artificial harbour basin at the starting point of the canal. Because of the artificial bridge, the schedia’, and a very dynamic development, the settlement, extended on both sides of the Canopic Nile towards the southeast. Here, on its southern borders, monumental and public buildings (baths and temples) were erected not later than middle and late Ptolemaic times. The settlement, therefore, already had urban character, which is also documented by the inscriptions.

Bibliography Abd el Fattah, A., 2000-2001, Brief report on excavations carried on at Kom el Giza during the season 1989. Annales du Service des Antiquités de l’É gypte 76: 9-12. Abd el Fattah, A., 1998, Quelques nouveaux moules alexandrins à Kom Giza. In A. Abd el-Abd el Fattah & P. Gallo (eds.), Aegyptiaca Alexandrina. Monuments pharaoniques découverts récemment à Alexandrie 1: 65-73. Alexandrina. Abd el Fattah, A., 1988, Recent discoveries in Alexandria and the Chora. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athè nes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole franç aise d’Athè nes, 11-12 décembre 1988: 47-53. Athens, Paris.

In the course of the Late Hellenistic and Early Roman periods a substantial necropolis with monumental tombs developed along the southeastern outskirts of the city, but because of the continuous growth of Schedia, luxurious residences for a prosperous upper class soon penetrated the periphery previously used for burials. At the same time on the southern embankment a monumental building, maybe a temple complex, as well as large storage facilities like the pillar building, arise, apparently as a consequence of massive investments. On one hand, they attest to the economic attractiveness of Schedia in the Imperial period, and, on the other, their distance from the harbour suggests that the more favourable zones were already built up. 116

M. BERGMANN, ET AL.: SCHEDIA ON THE CANOPIC NILE Banaji, J., 2001, Agrarian Change in Late Antiquity. G old, Labour and Aristocratic Dominance. Oxford. Bergmann, B., & Heinzelmann, M., 2007, Schedia, Alexandrias Hafen am kanopischen Nil. Zwischenbericht zu den Arbeiten 2003-2007. Hefte des Archäol ogischen Seminars Bern 20: 65-77. Bergmann, B., & Heinzelmann, M., forthcoming, The tholos-bath at Schedia. In M.F. Boussac, T. Fournet & B. Redon (eds.), Le bain collectif en Egypte. Des balaneia antiques aux hamams contemporains: origine, évolution et actualité des pratiques. Actes du colloque Balnéorient d’Alexandrie 1-4 Décembre 2006. Alexandria. Bernand, A., 1966, Alexandrie la G rande. Paris. Bernand, A., 1970a, Alexandrian Canal. Le delta égyptien d’aprè s les textes grecs I. Les confins liby ues: 329-380. Cairo. Bernand, A., 1970b, Schedia/Menelais. Le delta égyptien d’aprè s les textes grecs I. Les confins liby ues: 381-442. Cairo. Botti, D.G., 1902, Studio sul III° nomo dell’Egitto inferiore. Bulletin de la Société d’Archéologique d’Alexandrie 4: 41-61. Boussac, M.F., Fournet, T., & Redon, B. (eds.), forthcoming, Le bain collectif en Egypte. Des balaneia antiques aux hamams contemporains: origine, évolution et actualité des pratiques. Actes du colloque Balnéorient d’Alexandrie 1-4 Décembre 2006. Alexandria. Daszewski, A.W., 1998, La nécropole de Marina el-Alamein. In S. Marchgay, M.-Th. Le Dinahet & J.-F. Salles (eds.), Nécropoles et pouvoir: idéologies, pratiques et interprétations. Proceedings of a conference Lyon 1995 : 229-241. Lyon. Empereur, J.-Y., 1998, Les ateliers d’amphores du lac Mariout. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athè nes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole

franç aise d’Athè nes, 11-12 décembre 1988: 75-91. Athens, Paris. Jomard, E.F., & Jacotin, P., 1818/25, Carte Topographique De L’É gypte Et De Plusieurs Parties Des Pays Limitrophes (1818/25), M 1: 100 000, Blatt Nr. 37. Kingsley, S., & Decker, M., 2001, New Rome, new theories on inter-regional exchange. In S. Kingsley & M. Decker (eds.), Economy and Exchange in the East Mediterranean during Late Antiquity. Proceedings of a conference at Somerville College. O xford 29th May 1999: 1-27. Oxford. Martin, A., 2008, Pottery from Schedia near Alexandria (Egypt). Rei Cretariae Romanae Fautorum Acta 40: 263-269. Martin, A., forthcoming, The pottery from a late-antique settlement at Schedia (Western Delta, Egypt). In S. Menchelli, M. Pasquinucci & S. Santoro (eds.), The 3rd International Conference on Late Roman Coarse Ware, Cooking Ware and Amphorae in the Mediterranean: Archaeology and Archaeometry ( LRCW3) ( Parma – P isa, 25- 30 March 2008) . Oxford. Procopius, e edificiis, Vol. VII, H.B. Dewing (transl. & ed.), 1961. Cambridge, MA. RE 1899, Wissowa, G. (ed.), Paulys Real-Encyclopädi e der classischen Altertumswissenschaften, Vol. I 1, 2030 s.v. αιρ ου (Sethe). Stuttgart. RE 1921, Wissowa, G., Kroll, W., &Witte, K. (eds.), Paulys Real-Encyclopädi e der Classischen Altertumswissen-schaft, Vol. II A 2, 401-403 s.v. Schedia (Kees). Stuttgart. Rodziewicz, M.D., 1988, Classification of wineries from Mareotis. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athè nes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole franç aise d’Athè nes, 11-12 décembre 1988: 27-36. Athens, Paris. Strabo, G eography, Vol. VIII, H.L. Jones (transl. & ed.), 1959. Cambridge, MA.

117

LAKE MAREOTIS: RECONSTRUCTING THE PAST

118

P. WILSON: SURVEY WORK IN MAREOTIS

Recent Survey Work in the Southern Mareotis Area Penelope Wilson

Introduction The ‘Sais and its Hinterland Project’ based at Sa el-Hagar, Sais in Gharbiyah Province, was designed partly to determine the impact of river distributary systems on settlement in the north and north-west of Egypt. In order to compare the Canopic with the Rosetta riverine systems, 70 sites were surveyed and visited in Beheira and Kafr esh-Sheikh Provinces. This paper describes and discusses sites in the southern Mareotis area, which formed an interesting group in themselves. They represent places influenced by another type of water-system, an inland lake located at sea-level, that is fed by a series of distributaries from the Canopic system. The location and extent of Lake Mareotis fluctuated over time and this may have been an influencing factor in the life-span recorded at some of the sites in the southern Mareotis area.

sides. The lower parts of its sides have been left largely untouched, but are used for animal husbandry. The village of Kom Ishu was formerly on the western, southern and eastern sides of the hill but has developed in linear streets radiating to the east. It is possible that the original heart of the village was sheltered from the north wind by the hill. The field boundaries beyond the village suggest that the kom may have covered a larger area at one time. The whole of the upper surface of the hill is now covered by a modern cemetery. Pottery gathered at the site is mostly of Late Roman date.

K om Ishu and K om el-Hagg The two sites share a similar geological nature as they both consist of rocky limestone outcrops at the edge of a limestone ridge formation and seem to have been the focus of human activity in Lake Mareotis (Embabi 2004: 260-2).

Kom el-Hagg (SCA 100155) is surrounded by a small ezbet, called Ezbet el-Bank. The mound has a sloping, rounded edge to the east and the other sides are slightly steeper at the edges. The area is used for stabling animals and as a manure store. The southern edge of the kom has been cut away, leaving a metre-high section above field level containing pottery fragments. The fields to the south of the main mound are covered in sherds and may have been part of the ancient site, perhaps representing a settlement against the rocky outcrop. It is also possible that the sherds to the south were washed or swept down off the rocky hill. It would be interesting, therefore, to define the exact nature of the site at this point because of its location and the apparent use of the local geological features as at Kom Ishu and also at Kom Lemsan, situated to the north.

Kom Ishu is approximately 12 km west-south-west of Kafr ad Dawar. The rocky outcrop has steep, almost sheer,

The pottery sample from the site gave a broadly consistent date range from the Ptolemaic through to the Early Roman

The Sites (Figs. 1-2) The sites are discussed in an anti-clockwise direction around Lake Mareotis. Although the pottery collection and survey are of a preliminary nature, some interesting observations can be made arising from a study of the results.

Fig. 1: Ancient sites in the Delta, highlighting the Mareotis region ( drawn by P. Wilson) .

119

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig. 2: Map of the sites surveyed, with 0 metre sea-level contour area highlighted in grey. By R. Dickinson, after 1:50,000 Survey of Egypt maps ( 1997) .

120

P. WILSON: SURVEY WORK IN MAREOTIS periods. The different types of vessel from the site included both amphorae and domestic vessels and so did not indicate a particular function for the site, but there certainly seems to have been some kind of settlement here. Most of the pottery was Egyptian in origin, and presumably was produced locally, with only one of the amphorae of Early Ptolemaic date perhaps from Rhodes, possibly a reused wine or water container. Both Kom el-Hagg and Kom Ishu are situated on the western edge of the main body of Lake Mareotis and could have had strategic value in guarding the lake shores or in watching over the desert areas to the west, the river to the south and marshes or fields on all sides Bernand 1970: 866). Although the two places are within signalling distance, the pottery represents two distinct phases, with that from Kom el-Hagg covering the Ptolemaic and Early Roman periods, while that at Kom Ishu is of the Later Roman period. There may be a number of reasons for this, mostly due to the nature of the collection of the pottery sample. But it could be that Kom Ishu replaced Kom el-Hagg as a watching place or simply that it continued for a longer period of time, perhaps because it was situated upon a more long-lived canal which later became the Tirat Abis. In such a case, Kom Ishu had a better strategic value in the later period because of its water channel communication.

Fig. 3: The main mound at K om el-Mahar ( photograph by P. Wilson) . 2nd century. Most of the pottery sample consists of tablewares, with some imported finewares and glass, and may have come from domestic contexts. Kom el-Mahar could have been a small town or village focussed on a series of elite houses and villas on the southern shore of the lake, taking advantage of the elevated land (Botti 1902: 55).

K om el-Mahar and Sidi G hazi Kom el-Mahar (SCA 100167) lies isolated in farmland to the east of the Nubariyah Canal and south-west of Sidi Ghazi village. The main mound is covered in pottery, including finewares ig. 3 . The name of the site in Arabic means ‘shell’ and shelly deposits are apparent all over the mound and in the sections. The shell derives from the fact that the mud-brick used to build the houses of the town was made from silt mixed with crushed shell from degraded, local oolitic limestone. When the mud-brick disintegrated, a large amount of shell was left lying upon the surface.

The modern village of Sidi Ghazi (SCA 100142) lies approximately 10 km south of Kafr ad Dawar and to the north of the Masraf al-Umum Canal which runs to the west of the Baslaqun-Luqin group of tells. There is now no tell at the location of Sidi Ghazi, but modern maps indicate that the village lies upon a small area of raised ground. It is likely that the original tell has been subsumed under the village. A site here may have formed one of the islands inside Lake Mareotis. South-western Sites Kom el-Gel (SCA 100217) is the site at Ezbet Abd el-Qadir at-Tawil and lies approximately 17 km north-west of Abu Matamir. The site consists of a mound with a flat top and gently sloping northern side and a secondary mound to the east. The top of the mound is covered with a modern cemetery, while the sides are used for animal husbandry. The houses on the western side of the tell are built on flat land at the base of the hill. The fields to the south of the mound contain noticeable quantities of pottery and some limestone fragments, suggesting that this was once a more sizeable mound. Satellite imagery suggests that the area to the south may have been part of the mound because of the shape of the fields.

The lower slopes of the eastern and south-eastern side of the mound are covered by a modern cemetery. The northern and north-western side of the mound has sheer sides, where they have been dug away and it is likely that the original site extended in all directions away from the mound. It is now surrounded by deep irrigation ditches and fields, which are encroaching on the outer parts of the site. One of the northern sections contains a substantial, stratified sequence, including ashlar limestone blocks, a wall made of sandy mud-bricks and shell-filled, mudbricks. There are also fragments of red granite on the mound and glass fragments lying upon the surface of the site. The southern area of the site has been flattened and there are building outlines visible on the surface. There are some lower, smaller mounds covered in scrub to the northwest and south-east sides of the site.

The pottery sherds from the site comprised fine wares, domestic material and amphorae. The pottery dated mainly to the Late Ptolemaic period, with some Early and Late Roman sherds included in the collection.

The main pottery component from the site ranges from Late Ptolemaic to Early Roman in date (Fig. 4), suggesting that it may be a Ptolemaic foundation continuing as a town into the Roman period, before being abandoned around the

The original site at Kom el-Farag lies under a modern town, approximately 13 km north-west of Abu Matamir. It may be the location of Kom el-Hanache (Toussoun 1922: 121

LAKE MAREOTIS: RECONSTRUCTING THE PAST Fig. 4: Late Ptolemaic to Early Roman pottery sample from K om el-Mahar ( drawings and analysis by D. G rigoropoulos) .

lake next to the village which may be a relic of an ancient river distributary or ox-bow lake, extending for about 500 m in a north-west to south-east orientation. The lake may link with some ancient depressions in the landscape to the west or have formed one of the distributary channels from the Canopic Branch. If there was once an ancient settlement here, it may have been located upon the river, lying along a raised levee as with other tells in this part of Beheira.

Pl.II), but as the Ezbet Farag grew to the south, the tell seems to have been subsumed underneath the houses and removed by the Delta Light Railway. The cemetery part of Kom el-Farag lies to the north-west and may be a remnant of the older mound, being a maximum of 1.5 m above ground level. The old part of the village upon the mound has a less organised street plan than the more recent areas of the town with their orthogonal layout. The interest of this location is that it lies on the Tirat Abdel al-Fatah and Tirat al-Hajar drainage channels and, perhaps, on the channel of one of the earlier Canopic distributaries on the western side of Beheira. Kom Farag is also at the centre of a cluster of small villages and ezbets with the word kom in their name, such as Kom es-Siwayd to the south, Kom Kifri, as Sa’ayidah and Jaradah to the south-west and to the north-west Ezbet Abd el-Qadir atTawil (Kom Makhboura), which has a cemetery upon a kom. They may well be part of a town cluster focussed on an irrigation basin network dating at least to the medieval period.

Abu Guduur (SCA 100212) is 2.5 km north-west of Abu Matamir and lies to the south of the main road, upon a low ridge of silt along an ancient distributary course, at the edge of the Mareotis basin. It once covered an extensive area, but now consists of some low, sandy mounds covered in scrub and a larger flat area with scrub and bushes. Most of the site is now used as a municipal rubbish dump, having once also been a gravel extraction and processing area. The highest mound is only about 2 m above the level of the surrounding land and the whole area is very sandy. There was pottery lying on the tell as well as some bronze objects, including a stirrup ring and a three-flanged arrowhead. Both could date to the Late Dynastic period. A small diagnostic set of pottery fragments were collected from the site. As far as can be judged, they mostly date to the Late Ptolemaic period, but there were some Late Roman sherds and some Late Dynastic pottery as well.

South of the Lake Kom el-Adda SCA 100230 or Kom Abu El-Eida lies 8 km north-west of Abu Matamir. Although there is a cemetery mound to the south-east of the main village, the main interest of the area is the curious crescent-shaped, enclosed 122

P. WILSON: SURVEY WORK IN MAREOTIS The place may have been a strategic Dynastic location for a small military establishment related to the other western delta fortified sites and to other mounds running from Abu Matamir to the south east. Southern Port on the Naukratis Canal Kom Trugi or Turuga (SCA 100251) lies 6 km due north of Abu Matamir, close to several other places with kom in their name and on the former edge of Lake Mareotis. Turuga is known to have been an important inland port on the lake for people going from Alexandria to the Wadi elNatrun and for goods and supplies coming out of the Western Desert and delta towards Alexandria for trade (Timm 1984-1992: 2545-6 De Cosson 1935: 79, 151 . It is possible that the site had two halves on either side of the waterway into the lake, one known as Psenemphaia, according to a stela from the site Bernand 1970: 899-913 and the other known as Theroge or Therange, which lends its name to the whole site after the Ptolemaic period. Roman period material was noted at the site in 1895 by D.G. Hogarth Spencer 1999: 305, 308-11 and Botti noted that marble column capitals could been seen there and that a headless white marble statue was found at Kom Trugi (Botti 1902: 58). Kom Trugi suffered greatly from sebakhin digging (De Cosson 1936 , with the authorities only intervening at the discovery of important objects or remains such as a Ptolemaic-Roman bath-house El-Khashab 1957: 127-139 .

Fig. 5: Red brick structures at K om Trugi, excavated below foundation level. The scale is 2 metres high ( photograph by P. Wilson) . Fig. 6: Section through the cemetery mound at K om elQ adi ( photograph by P. Wilson) .

It is clear that this was once an extensive site and perhaps a high mound. It now consists of flat, sandy areas interspersed with several excavated red brick structures and limestone upon mounds. To the north of the area, there is a modern village with an adjoining cemetery and both may have been built upon part of the original site. Amongst the extant red brick structures there is part of a pavement left standing upon its sand-brick foundation and a pillar of earth, as well as red brick and plastered tanks, perhaps from a winery. There is also a fragment of a limestone wall consisting of five courses of ashlar blocks for a length of 4.75 m, with some plaster still adhering to one side (Fig. 5). All of this area seems to have been dug out to below the floor level of these buildings and the site seems very denuded. On one of the mounds to the south there are the foundation walls of a limestone building. The flat, sandy area between the mounds shows building plans in wet weather, according to the local people, but is now mostly used as a football pitch. There are also a number of red granite grindstones at the site, lying upon the ground. Although it is difficult to make sense of the standing remains, they do suggest that an impressive town with large red brick and limestone buildings once stood here.

the medieval period. The visible levels date perhaps to the Early and Middle Roman periods and information about the later levels may be lost. The height of the mound suggests that there is still buried material at the site, but that it is disappearing fast. North-east of the Lake Kom el-Qadi (SCA 100165), south of the main Cairo to Alexandria road, and Kafr ad Dawar may be part of a cluster of sites around Al Baslaqun, forming a township or irrigation basin group of sites. The sprawling area at Kom el-Qadi has several distinct zones and a main archaeological area to the south of Kom el-Qadi village. It consists of a high mound, around 8 m high, the southern and northern faces having been cut away to form sheer sections. The southern sections show significant stratified sequences, including sand-brick walls, pottery deposits and human burials in both pottery coffins and brick vaults. Skulls of adults and children can be seen in the sections and on top of the mound (Fig. 6). The remains may be interpreted as

The pottery sample collected from the site dates mainly from the Early Roman period, but there are also some Ptolemaic, Middle-Late Roman and Medieval Arab period sherds in the collected sample. Because of the disturbed nature of the site, however, it is difficult to associate particular pottery with specific areas of the site. The range of the pottery agrees well with the historical material, suggesting that the site was used from the Ptolemaic through to 123

LAKE MAREOTIS: RECONSTRUCTING THE PAST Summary (Table 1) Lake Mareotis created its own special topography and landscape, with the shallow flooded area bordered by reeds and rushes forcing people onto the high ground around the lake edge, islands in the lake and between the distributaries to the south and the Canopic and Alexandria Canal to the north. The strange, thin tongues of land, with their dykes and paths, must have made the desert routes seem attractive to travellers and so created opportunities on the Western Desert edges for monastic hostels and stopping places. The topography of the area shows that the presence of the lake had a profound impact upon social organisation and economic management in the North-West Delta. Pharaonic period attempts to control the Western Desert edges may always have led to limited settlement in the north-west area, as suggested by Abu Guduur and the fortress line in the Ramesside period (Habachi 1980). The foundation of Alexandria seems to have acted like a magnet, attracting service settlements and industries upon a huge scale, as well as enabling the development of agricultural lands around the city and bringing settlement in favourable locations. The preliminary dating evidence from the sites in Beheira shows that the earliest material at the surveyed sites in the majority of cases comes from the Ptolemaic period, contemporary with the development of Alexandria Wilson & Grigoropoulos 2009: 42-43 . The sites at Kom el-Mahar, Kom el-Gel and Kom Trugi may all be part of this wave of development, with the creation of farming towns, a process which continued into the Early Roman period. This seems to have been a busy time for the lake, with many settlements established either as internal trans-shipping ports for the movement of goods and people

a reuse of the abandoned town mound as a cemetery. This may be the Roman cemetery from which three inscriptions were published (Botti 1902: 57). On the eastern side of this mound a substantial mud-brick wall is visible in the section. In the village to the west, there is a Roman bath-house, which has been restored recently and includes part of the concrete floor of a tholos, some possible underfloor furnaces and rectangular tanks. More uncovered red brick chambers as well as surface building plans, can be seen on a sandy area to the east of the village and main mound. A further sandy area lying to the east of this part of the village and a patch of extensive low scrub to the north-east of the village may have been part of the original site. There were some larger fragments of limestone lying in the mounds and the village. The pottery collected was mostly from the mound. It consisted of Early and Middle Roman material, including some imported fine wares, many cooking pots which showed evidence of burning on the outside and some Egyptian amphorae. The material seems to indicate the domestic nature of the site in the Roman period, but the relationship of the cemetery remains to the settlements was not clear from a preliminary examination of the mound sections. The site seems to be a small town and its associated cemetery, perhaps dating from the height of Roman management of this area. Table 1: Summary table of sites surveyed by the ‘ Sais and its Hinterland Proj ect’ in the region of southern Lake Mareotis. Site & current dimensions

Early Roman Ptolemaic Early Late 1st c. BC rd st 3 -late 1 c. BC to 2nd c. AD

Middle Roman Late Roman 3rd-4th c. AD 5th-end 7th c. AD

Kom Ishu 150 x 200 x 10 m

Early medieval Arab 8th-16th c. AD

X

FUNCTION Watch post

Kom el-Hagg 120 x 200 x 4 m

X

X

Watch post

Kom el-Mahar 440 x 240 x 12 m

X

X

Main Ptol-Roman town and villa site

Sidi Ghazi Under town Kom el-Gel 270 x 220 x 4 m

Island village X late

X

Town/villa

Kom el-Farag Under town

Farming village

Kom el-Adda 220 x 300 x 3 m

Riverside town

Abu Guduur 300 x 250 x 2 m

X late

Kom Trugi 350 x 300 x 3 m

X

X late X

X

Kom Hassan nder fields Kom el-Qadi 370 x 250 x 8 m

Fort?, riverside town X

Inland harbour town Unclear

X

X

Main Roman town

124

P. WILSON: SURVEY WORK IN MAREOTIS Bibliography SCA numbers refer to site registration numbers of the Supreme Council of Antiquities, Egypt.

to and from Alexandria, or as points of control around the lake, such as at Kom Ishu, Kom el-Hagg and perhaps Sidi Ghazi. The apparent abandonment of places, such as Kom el-Mahar and Kom el-Gel, may be part of the ‘Antonine Plague’ effect with depopulation of some areas (Bagnall & rier 1994: 173-8 . In the later periods, the lack of Late Roman and Early Arab continuity at the sites suggests that, in common with other parts of Beheira province, once the Canopic Branch of the Nile and its distributaries had silted up, the sustainability of some of the settlements was not possible and that people moved away from earlier settled land. They may have gone to a few major centres such as Abu Matamir or Dilingat or Damanhur or Alexandria itself, rather than carrying on trying to eke out an existence on the shores of the Mareotis Lake.

Bagnall, R.S., & Frier, B.W., 1994, The Demography of Roman Egypt. Cambridge. Bernand, A., 1970, Le Delta égyptien d’aprè s les textes grecques, I-III. Cairo. Botti, G., 1902, Studio sul IIIo nomo dell’egitto inferiore: e più specialmente sulla Regione Mareotica. Bulletin de la societé royale d’archéologie - Alexandrie 4: 4184. De Cosson, A., 1935, Mareotis: Being an Account of the History and Ancient Monuments of the North-West Desert of Egypt and Lake Mareotis. London. De Cosson, A., 1936, El Barnugi, Bulletin de la societé royale d’archaeologie - Alexandrie 31: 113-116. Daressy, G., 1929, Ménélais et l’embouchure de la branche Canopique. Revue de l’É gypte ancienne 2: 20-51, Pls. I-II. Embabi, N.S., 2004, The G eomorphology of Egypt. Landforms and Evolution. Vol. I. The Nile Valley and the Western Desert. Cairo. El-Khachab, A.M., 1957, Les Hammams du Kom Trougah. Annale du Service des Antiquités de l’ Egypte 54: 117-139. Habachi, L., 1980, The Military Posts of Ramesses II on the Coastal Road and the Western Part of the Delta. Bulletin de l’Institut franç ais d’archéologie orientale 80: 13-30. Spencer, P., 1999, Hogarth’s 1895 Report to the Committee of the Egypt Exploration Fund. In W.J. Tait & A. Leahy (eds.), Studies on Ancient Egypt in Honour of H.S. Smith: 302-311. London. Survey of Egypt 1997. 1:50,000 Maps; “Abu al-Matamir”, “Kafr ad-Dawwar”, “Ikinij Maryut” & “al-Iskan dariyyah”. Survey of Egypt/Ministry of Finances, Cairo. Timm, S., 1984-1992, Das christlich-koptische Ägypten in arabischer Z eit: Eine Sammlung christlicher Syät ten in Ägypten in arabischer Z eit, 6 Vols. Wiesbaden. Toussoun, O., 1922, Mémoire sur les anciennes branches du Nil. Mémoires de l’Institut d’É gypte 4, Premier fascicle, Cairo Wilson, P., & Grigoropolous, D., 2009, The West Nile Delta Regional Survey, Beheira and K afr esh-Sheikh Provinces. London.

Of the sites surveyed here, Kom Trugi was the largest and most important, perhaps straddling the distributary of the Canopic Branch from Naukratis and providing a secondary direct link from the south to Alexandria and from the Western Delta area to Alexandria. Kom el-Qadi seems to have been a typical Early Roman town with a bath-house, similar to other new towns in the Roman delta, and a substantial mound used as a cemetery. It may have been founded in the Early Roman period at around the same time as Schedia and part of the cluster around BaslaqunLuqin, once suggested to be the location of Menelais Daressy 1929). Further detailed work in any or all of the Mareotis catchment area would be highly desirable to explore riverine, lacustrine and human interaction during periods of profound cultural change. Acknowledgements I thank Dr Dimitrios Grigoropoulos of the Deutsches Archä ologisches Institut – Athens who studied the pottery from the sites and Dr Roger Dickinson for the mapping. I am also grateful to the Supreme Council of Antiquities in Damanhur under Fawzy Fawzy el-Kholi and Naama Suleiman, as well as Zahi Hawass and Magdy Ghandour in Cairo. or the fieldwork, I thank Dr oanne Rowland, Mohamed Osman, Ashraf abd el Rahman and Ahmed Bilal. The Department of Archaeology, Durham University and Egypt Exploration Society also supported the work. Funding from the Arts and Humanities Research Council, UK is acknowledged. The survey is now published by the Egypt Exploration Society’s Delta Survey Project under the title The West Nile Delta Regional Survey, Beheira and K afr esh Sheikh Provinces.

125

LAKE MAREOTIS: RECONSTRUCTING THE PAST

126

D. DZIERZBICKA: WINERIES OF THE MAREOTIC REGION

Wineries of the Mareotic Region Dorota Dzierzbicka

Sea, it was by far the best in the Nile Valley. Sources from the Pharaonic and Persian periods testify to wine production in this area long before the coming of the Greeks (cf. Forbes 1955: 70-124). The wineries date from the Roman and Byzantine periods. Some are of considerable size and testify to extensive cultivation of grapes and large scale wine production in this area, especially in correlation with the local pottery kilns which indicate a mass production of amphorae (Empereur 1993: 39-47).

Wineries or remains of wine making installations have been found in various parts of the Nile Valley: the coastal areas and the Delta, the Fayum and Bahariya Oasis, as well as Upper Egypt (Dzierzbicka 2005: 11-24). However, their concentration in the Mareotic region, especially along the southern shore of the lake and around Abu Mina, is unmatched in other parts of the country (Fig. 1). In addition to the identified wineries, there are also elements of wine-making installations that are no longer in situ – primarily lion headed spouts through which freshly pressed or trodden must flowed into the vat. umerous examples, including those found around Lake Maryut, can be seen in Alexandria in the Greco-Roman Museum or at the Kom el-Shukaqa necropolis (Rodziewicz 1998: 28, n. 6).

Literary Sources In literary sources wine produced in the Mareotic region during the Greco-Roman Period is well attested. It was known to Horace (O des I 37.14), who wrote that the mind of Cleopatra was made mad by Mareotic wine (mens lymphata Mareotico). According to Strabo (XVII 1,14) the Mareotic region produced good wine and in such quantity that it was racked to be matured. The white Mareotic grape variety (Mareotides albæ ) is mentioned by Virgil in G eorgica (G eor. ii. 91). Athenaeus (Deip. I 33 d-e), in turn

Many factors contributed to the large concentration of wineries in the vicinity of Lake Maryut. Although in antiquity the climate in the Delta may have not been as beneficial for vineyards as on the northern shores of the Mediterranean

Fig. 1: A map of the Mareotic region with marked locations of wineries. Plotted on the map are excavated and published structures, as well as unexcavated ones mentioned by scholars in other works or featured on SCA maps as protected archaeological sites. The wineries designated as NN were plotted by Rodziewicz ( 1998: 28, Fig. 1) , but the name of their location was not provided ( D. Dzierzbicka) .

127

LAKE MAREOTIS: RECONSTRUCTING THE PAST claims that Mareotic wine was excellent it was white and pleasant, fragrant, easily assimilated, thin, did not go to the head and was diuretic. According to this author better yet was the wine from Taenia, from the same region. This was pale yellow and had an oily quality, which disappeared after it had been diluted with water. Besides being sweet, this wine was rather aromatic and mildly astringent.

There are three other wine making installations in the vicinity of the modern village of Huwariya. One of them is a simple Early Roman Type I unit built of dressed stone blocks bound with mortar unexplored, mentioned by Rodziewicz 1998: 31). A winery in southern Huwariya (reported by Rodziewicz 1998: 34), built of irregular stone blocks, clay and waterproof plaster, consisted of one press unit a treading platform and vat connected by two flights of steps. The round base of a mechanical press was set in a niche in the wall of the treading platform. A channel in the floor of the platform led from the press directly to the vat. A third installation, located 3 km southeast of Huwariya (Rodziewicz 1998: 35), consisted of a treading platform, a vat and a mechanical press separated from the treading floor by a low, thin wall. A channel under the floor led from the mechanical press to the vat. The walls of the structure were built of irregular stone blocks bonded with mortar and covered with plaster reinforced with potsherds.

Wineries1 The South Coast of Lake Maryut The wineries in the immediate vicinity of Lake Mareotis are located on the strip of land between the lake and the Mediterranean Sea, and along the southern shore of the lake. The majority are found on the southern shore of the lake. A large winery near Burg el-Arab (30o55’37”N 29o 31’48.50”E),2 is located in the northeast corner of a large building (El-Ashmawi 1998: 55-73). A mechanical press was set up on the treading platform and closed off from the surrounding area by a thin wall. The must from the mechanical press enclosure and from the treading floor flowed to the adjacent collection basin through two separate channels.

A small, presumably Roman winery of George Nesim (30° 57’18.30”N 29° 36’15.20”E) was found by the Lake Mareotis Research Project (Blue & Khalil forthcoming). It appears to have consisted of a treading platform and vat Type I Rodziewicz 1998: 31 , but the presence of a mechanical press unit with a separate, small collection basin within the same complex cannot be excluded, although it requires further investigation.

The winery located at Taher el-Masry (30o56’59.40”N 29o34’58.20”E)3 was built of stone blocks and stone rubble embedded in mortar. The uncovered press unit consisted of a square treading platform 3.75 m x 3.75 m and a vat 2.37 m x 2.00 m , 1.40 m deep. utlines of structures surrounding the unit suggest that the winery may have been part of some kind of a larger complex.

The North Coast of Lake Maryut To the north of the lake, at Abu Talaat (unpublished, mentioned by Rodziewicz 1998: 29), an Egyptian archaeological mission uncovered a winery with two connected treading platforms and one collecting vat. The wine-making unit was entirely hewn out of the bedrock except for one wall, which was built of stone blocks.

A restored wine-making unit near Marea (30o58’49.10”N 29o40’12.90”E) consisted of a large treading platform, a smaller room with a base for a mechanical press, and a collection vat (El-Fakharani 1983: 182-184). Two separate channels led from the two rooms to the vat. The rooms where the pressing took place were raised higher than the room of the vat and two flights of steps connected the two levels. Nearby, on the same site another unpublished winemaking unit was found.4

The winery at Site 215 recorded by the Lake Mareotis Research Project (30° 53’30”N 29° 25’22”E) has two parallel treading platforms visible on the surface. The channels in their western sides indicate the presence of two collection basins, which have not yet been uncovered. Thus, this winery represents a compound of two Type I wineries according to the typology established by Rodziewicz (1998: 31-35). Further investigation is needed to identify its other features.

1. This part of the paper includes elements previously published in Dzierzbicka 2005. 2. GPS coordinates of the wineries as well as some additional information – dimensions, architectural details – were collected on my study tour of the Mareotic wineries in November 2005 during my scholarship at the Polish Centre of Mediterranean Archaeology in Cairo, Egypt. The study tour included the wineries near Burg el-Arab (also called Abu Sir), at Taher el-Masry and Marea, two of the installations at Abu Mina, and the wineries at Karm el-Shewelhy and Karm el-Baraasi. I am thankful to Mr. Adli Rushdy from the West Delta Inspectorate and to Mr. Saber Selim from the Islamic Inspectorate of the Supreme Council of Antiquities for allowing me to see the sites. For GPS locations and photographs of the wineries at George Nesim, Kom Trouga and Site 215 I am greatly indebted to Dr. Lucy Blue and Dr. Emad Khalil of the Lake Mareotis Research Project. 3. Explored by Mr. Mustafa Rushdy of the Egyptian Antiquities rganisation in 1998/99 (as it was communicated to me by Mr. Maged Ahmed, Inspector of the SCA), unpublished. The survey I conducted in 2005 yielded the GPS coordinates of the site, as well as some preliminary observations based on the remains visible on the surface. 4. I am thankful to Prof. Mona Haggag for bringing it to my attention.

Area to the East of Lake Maryut Much fewer wineries have been located in the eastern part of the Mareotis region. A winery was uncovered in 1917 by Daninos Pasha near Abu ir Bay east of Alexandria (published as a private bath in Breccia 1926: 47-49). The Ptolemaic dating established by Breccia on the basis of the quality of plaster used in the structure seems rather uncertain. A large treading platform 3.25 m x 5.80 m covered with waterproof plaster was raised 0.30 m above the floor level of the building. Two channels led through a wall separating the platform from a collection basin 1.60 m deep, 4.20 m long and 2.20 m wide. Twin flights of steps connected the two parts of the winery. 128

D. DZIERZBICKA: WINERIES OF THE MAREOTIC REGION There is also a Roman Period winery at Kom Trouga (30° 57’46.90”N, 30° 10’27.10”E). This unpublished winery mentioned by El-Fakharani (1983: 184) and Rodziewicz 1998: 31 was a complex that consisted of several scattered wine making units, today badly damaged due to quarrying activity in the area. Remains of basins and adjacent raised treading platforms are visible. All of the units appear to have consisted of a single treading platform and vat, covered with waterproof plaster. The best preserved one was built of stone blocks, while others were brick structures. Possible wine-making installations have also been found at Schedia.5

two treading platforms, two vats, one mechanical press, and storage rooms. The structure was built of mud-brick and re-used stone blocks. The winery located at Karm el-Shewelhy (30o49’37.70”N, 29o38’65”E) consisted of a treading platform and vat, two storage rooms, and a courtyard. The wine-making unit was hewn in rock except for the eastern wall, which was built of limestone blocks. The floor of the treading platform and the walls of the vat were covered with waterproof plaster (Abd el-Aziz Negm 1998: 73). A sizeable winery at Karm el-Baraasi (30o50’41.40”N, 29o40’49.50”E) consisted of several large treading platforms with vats and mechanical presses. The excavations published by Abd el-Aziz Negm (1998: 63-70) brought to light a large treading platform and vat. In the same room as the vat there were two small enclosures (4 m2 each) for mechanical presses, each equipped with a small vat of its own. In November 2005 another treading platform was visible semicircular in shape, with an adjacent vat. This unit was located further to the west and separated from the first platform by a room. The floor level of this room was lower than the two treading platforms and connected to them by two short flights of steps. urther to the west traces of three more treading platforms were visible not yet excavated . The complex was built of mudbrick with corners reinforced with limestone blocks. The entrance to the unit uncovered by the excavators led from a courtyard. The winery was part of a larger complex across the courtyard from the winery there was a house, and the courtyard itself was surrounded by rooms. Around 300 m to the north-west of this complex 30o50’50.70”N, 29o40’40.20”E) there are probably more installations yet to be uncovered, as one can tell from the outlines of basins covered with waterproof plaster that are discernible on the surface.

The Region of Abu Mina A number of wineries were found in the vicinity of Abu Mina, not far from Lake Mareotis. The region was a thriving grape-growing and wine-making centre. There are three wineries within the town area. A large winery uncovered in 1907 by Kauffmann to the east of the double bath (Grossmann, et al. 1984: 148 Grossmann, et al. 1997: 87-90 behind the portico of the colonnaded street (30o50’52.90”N, 29o39 76.10 E was expanded in five phases until it became a complex of five treading floors with four mechanical presses. It was built of mud-brick and stone blocks covered with waterproof plaster. Another winery was located behind the apse of the basilica (30o50’45.80”N 29o39 82.30 E . It consisted of two rooms, the first housing one large treading platform flanked by two mechanical presses, a collection vat, and one smaller raised platform. In the second room there was another treading floor and vat Grossmann 1999: 82-83, fig. 9 . The third winery (Müller-Wiener & Grossmann 1967: 468-473) within the area of Abu Mina was uncovered in the so-called K umring A, in three rooms of Building 17, entered from a courtyard. Two of these rooms each contained a set of one large and one small treading platform and vat. The third room on the opposite side of the courtyard was used for storage.

Elements of Mareotic Wineries The Mareotic evidence provides an abundant and diverse range of evidence for wineries, a brief overview of the basic elements of a winery (the treading platform, the collection basin, and the mechanical press), and their function, are provided here.

The winery at Izbat Mohamed Farid, 1 km north of Abu Mina, had two phases of use (Grossmann 2002: 31). In the first phase, at the beginning or first half of the 5th century AD it was a simple pressing unit consisting of a treading platform and vat, with walls built of baked brick and covered with waterproof plaster. When it was no longer in use, another winery was built (later part of 6th century AD) partly on top of the old one. This new structure comprised a treading platform, a vat, and two mechanical presses. During both phases of use the winery stood next to a multistory country house.

Taking the winery at Marea as an example, it can be seen that a basic press unit was a compound consisting of a raised treading platform and a collection basin, built within an enclosed and most probably roofed, space (El-Fakharani 1983: 182-184 . The treading floor was separated from the vat by a low wall. It was covered with waterproof plaster, preserved, for instance, in the Burg el-Arab winery (El-Ashmawi 1998: 62), and was often quite large, as in the case of the Karm el-Baraasi winery (Abd el-Aziz Negm 1998: 66). The room in which the press unit was located was sometimes entered from a courtyard, or had a large window giving easy access to the treading floor. Grapes were thrown through this window, or simply carried through the door, to the treading platform, where they were crushed.

Three wineries near Abu Mina were explored in 1986 by the Supreme Council of Antiquities (Abd el-Aziz Negm 1998: 65-73). The winery at Karm Gadoura consisted of

5. For this information I am indebted to Prof. Dr. Marianne Bergmann and Prof. Dr. Michael Heinzelmann of the Swiss-German mission excavating at Schedia.

129

LAKE MAREOTIS: RECONSTRUCTING THE PAST Fig. 2: A schematic drawing of a Type 4 w inery with names of its elements used in G reek papyri from Egypt ( D. Dzierzbicka) .

Wineries in Papyri Archaeological finds are not the only available sources on wineries (for more details on wine production in papyri see Dzierzbicka 2005: 24-89). Greek papyri of the Hellenistic and Roman period allow for the identification of words used to call the elements of a winery. The reconstruction drawing of Type 4 winery (Fig. 2) shows the basic parts of a winery and gives the terms by which they were called in papyrus documents. Matching the architectural features with their Greek names in papyri makes it possible for the researcher to use hundreds of texts that provide invaluable information on the function of the winery as a whole and on its elements, permitting a better understanding of the way these structures were used and provoking more questions to ask oneself in the field.

Most wineries were also equipped with a mechanical press either a movable bag press, or a more sophisticated, fixed screw press. Their preserved remains were found in Abu Mina (Grossmann, et al. 1997: 87-90) and in Burg el-Arab (El-Ashmawi 1998: 60-64), where one can still clearly see a raised, circular base of the screw press. These devices were used to squeeze the remaining liquid out of the trodden grape pulp either directly on the treading floor, or in an area separated from it, as in Burg el-Arab. In the latter case, a channel connected the place of the mechanical press with the main collection basin, or with a smaller, separate vat. The must flowed into a plastered vat or vats dug into the ground. It remained there through the first turbulent phase of fermentation, during which the basins were covered with wooden planks for protection against contaminants.

One of these questions, especially important in the case of the Mareotic wineries, concerns their surroundings. Papyrus documents containing references to buildings and other immovable property found in vineyards, provide interesting information on other installations forming part of a wine-making complex Table 1 . The table below is a comparative chart of the facilities mentioned in extant texts, showing how frequently they appear and in what combinations. The documents mention elements already discussed: the treading platform (lê nos) and collection basin (pithos), in a later period referred to by a common term – lê nopithos, a mechanical press (stemphylourgikon organon or mê chanê ), as well as a sunning yard (hê liastê rion), a storeroom (thesauros) and a cellar (oinothê kê ). In fact, spaces that may have been storerooms were uncovered adjacent to wineries, for instance in Abu Mina (Grossmann, et al. 1997: 89). Large wineries where mass production took place even had a potter’s workshop (keramikon ergastê rion) on the premises. There was indeed a large pottery kiln in the vicinity of the winery near Burg elArab, suggesting this may have been such a more developed production estate (El-Ashmawi 1998: 62).

Generally speaking, wineries exhibit a rather uniform structure, dictated by practical needs, that hardly changed regardless of whether the winery was located on the Mareotis lakeshore, in the Bahariya asis unpublished wine press photo in Hawass 2000: 164), in Nubia (Gardberg 1970: 41-44), or even on the Crimean Peninsula (Michalowski 1958: 49 . Mareotic wineries provide an excellent chance to study the characteristic traits of these structures. What is striking about the typology elaborated by Rodziewicz (1998: 31-36) is the diversity of wine making installations which led to the distinguishing of as many as six simple and two composite types in just over a dozen wineries. However, the distinguished types are in fact very useful as they have strong similarities with wineries in other parts of Egypt and beyond (e.g. Michalowski 1958: 49). As it was noted by Rodziewicz (1998), the wineries consist of two or three essential elements: the treading platform, the collection basin, and, in types 3-8, a mechanical press. The convenient uniformity of this category of structures makes them fairly easy to recognise in an archaeological excavation. 130

P. Tebt. III.1 814, Sale, after 227 BC, Tebtynis

X

P. Lond. II 401, Petition, 115–110 BC, Thebaid

X

P. dem. G ieben 2, Sale, 107–30 BC, Sebennytos

X

PSI VIII 918, Sale, AD 38/39, Tebtynis

X

X

X

P. Ross. G eorg. II 28, Lease, after AD 163/164, Arsinoites

X

P. Flor. III 385, 2nd–3rd century AD, Hermopolites

O ikopedon

Pyrgos

K ella

Epoikion

Epaulis

X

X X X

PSI XIII 1328, Petition, AD 201, xyrhynchos

X

P. O xy. LI 3638, Cession, AD 220, Sinary, xyrhynchites

X

P. Flor. I 50, Division of property, AD 269, Hermopolis

X

P. O xy. XXXIV 2723, Sale, 3rd century AD, xyrhynchos

X

X

X

SPP XX 58 (7) Col. II, 3rd century AD, Hermopolis

X

X

X

SB XX 14291, Lease, 3rd century AD, xyrhynchos

X

X X

X

X

X

X

X

X

X

X

X

X

X

X

X X

X

P. Vind. Sal. 12, Lease, AD 334/335, Hermopolis

X X

X

X X

P. Vind. Tand. 28, Lease, AD 576/577 Hermopolis

SPP XX 218, Lease, early 7th century AD, Hermopolites

X

X X

P. Cairo Masp. I 67097 (r), Contract, AD 571/572 (?), Aphrodites Kome (Antaiopolites)

Thesauros

X

P. O xy. XLIX 3491, Marriage contract, AD 157/158, xyrhynchos

P. Vat. Aphrod. 25 fr. A, Division of property, 6th century AD, Aphrodito

O inothê kê

Hê liastê rion

K eramikon ergastê rion

Pithos

Stem. organon

Mê chanê

Lê nopithos

DOCUMENT

Lê nô n; lê nos

D. DZIERZBICKA: WINERIES OF THE MAREOTIC REGION

X X

X X

X

X

X X

X

Table 1: Installations forming part of a wine-making complex mentioned in G reek papyri containing references to immovable property found in vineyards or to the appurtenances of a winery. 131

LAKE MAREOTIS: RECONSTRUCTING THE PAST P. dem. G ieben 2 = Vandorpe, K., & Clarysse, W., 1998, A Greek Winery for Sale in a Fayum Demotic Papyrus. In A.M.F.W. Verhoogt & S.P. Vleeming (eds.), Two Faces of G raeco-Roman Egypt. G reece and Demotic and G reek-Demotic Texts and Studies Presented to P.W. Pestman: 127-139. Leiden, Boston, Köl n. P. Flor. I 50 = Vitelli, G. (ed.), 1906, Papiri greco-egizii, Papiri Fiorentini I. Documenti pubblici e privati dell’età r omana e bizantina. Milan. P. Flor. III 385 = Vitelli, G. (ed.), 1915, Papiri grecoegizii, Papiri Fiorentini III. Documenti e testi letterari dell‘ età r omana e bizantina. Milan. P. Lond. II 401 = Kenyon, F.G. (ed.), 1898, G reek Papyri in the British Museum II, 13-14, no. 401. London. P. O xy. XXXIV 2723 = Ingrams, L., Kingston, P., Parsons, P.J., & Rea, J.R. (eds.), 1968, The O xyrhynchus Papyri XXXIV, no. 2723. London. P. O xy. XLIX 3491 = Bülow-Jacobsen, A., & Whitehorne, J.E.G. (eds.), 1982, The O xyrhynchus Papyri XLIX, no. 3491. London. P. O xy. LI 3638 = Rea, J.R. (ed.), 1984, The O xyrhynchus Papyri LI, no. 3638. London. P. Ross. G eorg. II 28 = Krüger, O. (ed.), 1929, Papyri russischer und georgischer Sammlungen II. Ptolemäi sche und frühr ömische Texte, no. 28. Tiflis. PSI VIII 918 = Vitelli, G., & Norsa, M. (eds.), 1927, Papiri greci e latini VIII, no. 918. Florence. PSI XIII 1328 = Vitelli, G., & Norsa, M. (eds), 1949-1953, Papiri greci e latini XIII, no. 1328. Florence. P. Tebt. III.1 814 = Hunt, A.S., & Smyly, J.G. (eds.), 1933, The Tebtunis Papyri III.1, no. 814. London. P. Vat. Aphrod. 25 = Pintaudi, R. (ed.), 1980, I Papiri Vaticani di Aphrodito, no. 25. Rome. P. Vind. Sal. 12 = Salomons, R.P. (ed.), 1976, Einige Wiener Papyri, no. 12. Amsterdam. P. Vind. Tand. 28 = Sijpesteijn, P.J., & Worp, K.A. (eds.), 1976, Fünf unddreissig Wiener Papyri, no. 28. Zutphen. SB XX 14291 = Bilabel, F., Kiessling, E., & Rupprecht, H.-A. (eds.), 1997, Sammelbuch griechischer U rkunden aus Aegypten XX, no. 14291. Wiesbaden. SPP XX 58 = Wessely, C. (ed.), 1921, Studien zur Palaeographie und Papyruskunde X X . Catalogus Papyrorum Raineri. Series G raeca. Pars I. Textus G raeci papyrorum, qui in libro „ Papyrus Erzherzog Rainer– Führ er durch die Austellung Wien 1894 “ descripti sunt, no. 58. Leipzig.

The winery was not just a press unit, but a building or a set of buildings, located either in the urban area or in the vicinity of a vineyard. According to some documents the complex was also equipped with farm buildings epoikion, pyrgos, epaulis), storehouses (kella) and building plots (oikopedon). Indeed, the treading platform and vat at Abu Qir Bay was part of a sizeable square structure built of limestone blocks. The layout of rooms in the building is only partly preserved (Breccia 1926: 47-49). It has not been determined where the entrance to the winery was located and whether there were passages leading to the adjacent rooms. The wine-making complex in this structure may have not been limited to only one unit, consisting of a treading platform and vat, especially that the other rooms also had an industrial function, perhaps related to wine and oil production. As we learn from papyri (Dzierzbicka 2005: 24-89), after the turbulent phase of fermentation was over, the wine was poured into vessels that had been obtained from a pottery workshop located within the complex or acquired from outside the winery. Wine that was paid for in advance was distributed straight from the vat and carried off from the winery in vessels provided by the buyer. The filled jars that remained in the winery were placed in storage rooms or in the sunning yard (hê liastê rion) for maturing. Both of these facilities were usually spaces or buildings within the wine-making complex and they were equipped with doors and locks. Maturing wine was occasionally controlled to check if it had not gone sour or spoiled. After the process of fermentation had ended, the vessels were sealed. There are still many uncertainties about the broader context of wineries. The immediate surroundings of a winemaking structure can provide information on the building that housed the installation and determine whether it was located in a rural setting or in a larger industrial or even residential area. Further archaeological investigations, as well as the study of literary and documentary sources, may provide insight into industrial buildings and structures found in the vicinity of a winery and improve our understanding of the character of wine production in the Mareotic region. Acknowledgments I am grateful to the Polish Centre of Mediterranean Archaeology of the University of Warsaw for assistance granted to me during my stay in Egypt during the conference.

Ancient Sources Athenaeus, The Learned Banqueters, books I-III.106e, S.D. Olson (transl. & ed.), 2006. Cambridge. Horace, O des I, Carpe Diem, West, D. (transl. & ed.), 1995. xford. Strabo, G eography, H.L. Jones (ed.), 1996. London, New York. Virgil, Eclogues. G eorgics. Aeneid 1-6, Vol. 1, H.R. Fairclough (ed.), 1974. Cambridge.

Bibliography Papyri P. Cairo Masp. I 67097 (r) = Maspero, J. (ed.), 1911, Papyrus grecs d’époque byzantine. Catalogue général des antiquités égyptiennes du Musée du Caire I, no. 67097. Cairo.

132

D. DZIERZBICKA: WINERIES OF THE MAREOTIC REGION Gardberg, C.J., 1970, Scandinavian J oint Expedition to Sudanese Nubia, Vol 7: Late Nubian Sites: Churches and Settlements. Astrom. Grossmann, P., Kosciuk, J., Severin, G., & Severin, H.-G., 1984, Abu Mina. Elfter Vorl ufiger Bericht. Kampagnen 1982–1983. Mitteilungen des Deutchen Archäol ogischen Instituts Abteilung K airo 40: 148. Grossmann, P., Arnold, ., & Kościuk, ., 1997, Excavations at Abu Mina 1995. Bulletin de la Société d’Archéologie Copte 36: 87-90. Grossmann, P., 1999, Report on the excavations at Abu Mina in spring 1998. Bulletin de la Société d’Archéologie Copte 38: 75-84. Grossmann, P., 2002, Report on the excavations at Abu Mina in spring 2001. Bulletin de la Société d’Archéologie Copte 41: 25-31. Hawass, Z., 2000, The Valley of the G olden Mummies. Cairo. Michalowski, K., 1958, Mirmeki I. Warszawa. Müller-Wiener, W., & Grossmann, P., 1967, Abu Mina. 6. Vorl ufiger Bericht. Archäol ogischer Anzeiger 82.4: 468-473. Rodziewicz, M.D., 1998, Classification of wineries from Mareotis. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athè nes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole franç aise d’Athè nes, 11-12 décembre 1988: 27-36. Athens, Paris.

Secondary Sources Abd el-Aziz egm, M., 1998, Recent excavations around Abou Mina. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athè nes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole franç aise d’Athè nes, 11-12 décembre 1988: 65-73. Athens, Paris. El-Ashmawi, F., 1998, Pottery kiln and wine-factory at Burg el-Arab. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athè nes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole franç aise d’Athè nes, 11-12 décembre 1988: 55-64. Athens, Paris. Blue, L., & Khalil, E. (eds.), forthcoming, A Multidisciplinary Approach to Alexandria’s Economic Past: the Mareotis Case Study. xford Breccia, E., 1926, Le rovine e i monumenti di Canopo, eadelfia e il tempio di Pnefer s. Bergamo. Dzierzbicka, D., 2005, Wineries in Graeco-Roman Egypt. J ournal of J uristic Papyrology 35: 9-89. Empereur, J.-Y., 1993, La production viticole dans l’ Egypte ptolémaï que et romaine. In M.-C. Maouretti & J.P. Brun (eds.), La production du vin et de l’huile en Méditerranée: 39-47. Athens. El- akharani, ., 1983, Recent Excavations at Marea, Egypt. Aegyptiaca Treverensia 2: 175-186. Forbes, R.J., 1955, Studies in Ancient Technology, Vol 3. Leiden.

133

LAKE MAREOTIS: RECONSTRUCTING THE PAST

134

E. KHALIL: MAREOTIS’ WATERFRONT

Waterfront Installations and Maritime Activities in the Mareotic Region Emad Khalil

Introduction Lake Mareotis represents one of the most distinctive geomorphic features in the north-west coastal region of Egypt. In antiquity, it was fed by means of a number of canals, which bifurcated off the Nile’s defunct Canopic Branch, and flowed into the southern and eastern sides of the lake (Fig. 1). Some of these canals were navigable, which enabled merchandise to be transported to and from the hinterland. By the Greco-Roman period the lake was also connected to the sea through a navigable canal that debouched at Alexandria (Strabo 17.1.7). Its connection to both the Nile and the sea resulted in Mareotis becoming a vital conduit of communication in Egypt’s internal transport system. Moreover, it supported around its shores various agricultural activities and embraced major production centres for different industries, which contributed significantly to the economy of Alexandria and to Egypt as a whole. Accordingly, this paper will look at the role that Lake Mareotis played in the ‘maritime’1 transport system of GrecoRoman Egypt. It will also examine the types and nature of the maritime and waterfront installations that were recorded along the shores of the lake and the possible spatial and functional relationship between the different sites.

(Strabo 17.1.14; Pliny 5.11.63). It comprised a main rectangular body of water which merged to the east and south with the Nile Delta Plain, and a narrow arm that extended westwards parallel to the northern coast. However, during the past two millennia Lake Mareotis has undergone dramatic changes which significantly affected its size and nature. Nonetheless, the western extremities of the lake reflect the original extant remains, and form an arm, known as the Mareotic Arm, that extends some 40 km west of Alexandria and is 2-3 km wide and is separated from the lake’s main body by causeways and shallows. It also contains an island, Mareotis Island, which is 3.7 km long and about 680 m at its widest point (see both Blue and Hopkinson this volume).

The Lake Mareotis Research Project Much evidence indicates that Lake Mareotis extended in antiquity for about 50 km south and west of Alexandria

Since 2004 the Centre for Maritime Archaeology (CMA) of the University of Southampton, in collaboration with the Department of Underwater Antiquities (DUA) of the

Archaeological investigation of the western Mareotic Arm has been ongoing for several decades; however, previous research has been largely limited to specific areas and specific issues such as work on the Byzantine port of Marea/ Philoxenité (Petruso & Gabel 1982; El-Fakharani 1983; Rodziewicz 2003 , and work on amphorae and wine production (Empereur & Picon 1986, 1998) and on the wineries of the Mareotic region Rodziewicz 1998b .

Fig. 1: The approximate ancient and present limits of Lake Mareotis ( E. K halil) .

1. The term ‘maritime’ is used in this sense to denote all aspects of waterborne activity and communication from the sea, across the lake, along the canals and on the Nile River.

135

LAKE MAREOTIS: RECONSTRUCTING THE PAST Fig. 2: The Lake Mareotis Research Proj ect survey area along the shores of the western Mareotic Arm ( Lake Mareotis Research Proj ect) .

Waterfront Sites in the Mareotic Region The waterfront sites that were recorded along the shores of the western Mareotic Arm and on Mareotis Island can be classified into four categories:

Egyptian Supreme Council of Antiquities (SCA), has conducted a comprehensive archaeological survey along the shores of western arm of Lake Mareotis (see Blue this volume). The survey revealed a wealth of archaeological sites including numerous settlements of a maritime and industrial nature, which reflect the economic activities that took place in the Mareotic region (Fig. 2). Most of the archaeological sites recorded along the shores of the Mareotic Arm are located between Marea/Philoxenit and Taposiris Magna, with a concentration on Mareotis Island. The sites identified include maritime structures, such as harbours, jetties and quays, in addition to what appear to be waterfront warehouses and storage facilities. Sites pertaining to industrial activities are also evident and include amphora kilns, ceramic slag and kiln wasters, as well as a number of structures relating to water management, such as cisterns, wells and water wheels (sakkia). This correlates with much archaeological and textual evidence for viniculture and wine production in the region Rodziewicz 1998b .

1 - Harbours in the form of harbour complexes of significant design and constitute substantial structures. However, the only two sites that fit this description are Marea/Philoxenité and Taposiris Magna. The complexity and magnitude of their structures are unrepresented elsewhere in the entire Mareotic region. The two harbours date, however, to quite different periods. Taposiris Magna is essentially Hellenistic in date (see Boussac & El Amouri this volume), while Marea/Philoxenit mostly dates to the Late Roman period see Haggag, Pichot, Rodziewicz and Babraj & Szymańska this volume . However, the two harbours are associated with relatively large towns, and much historical and archaeological evidence indicates that these two towns where probably among the largest and most active along the shores of Lake Mareotis in antiquity.

The dating of the sites discovered relies primarily on the ceramic assemblages collected during the survey. Accordingly it was realised that the majority of sites date from the Hellenistic period until the 7th century.

Taposiris Magna has extensive archaeological remains that date from the Hellenistic to the Late Roman period, including evidence for thriving maritime and commercial 136

E. KHALIL: MAREOTIS’ WATERFRONT activities (El-Fakharani 1974; Empereur 1998: 225-7; Rodziewicz 1998a Boussac & El Amouri this volume . The town which is located on the northern shore of the western sub-basin of the lake has one of the best preserved harbours in Lake Mareotis (see Boussac & El Amouri this volume). Commercial activities in Taposiris Magna were mainly focused on handling products transported across the lake as well as receiving goods arriving from the west through overland routes, and shipping them to Alexandria (Empereur 1998: 225). Under the Romans, Taposiris Magna was a customs station where duties were levied on products coming from the Mareotic region and from Cyrenaica heading east towards Alexandria or to the Nile Delta (Empereur 1998: 225-7; Vör ös 2001: 15-6). Alternatively, river vessels could have travelled on the lake through the harbour of Taposiris Magna to the west as far as the lake extended.

ings to allow water to flush away the silt and sediments that might accumulate in the harbour basin. This arrangement was supplemented by the construction of a 1,700 m long wall that extends southwards from the artificial ridge to the southern shore of the lake. It was also supplemented by the construction of a solid limestone wall that extended from the northern shores of the lake to the seashore. This wall ensured that even caravans travelling overland had to go through the town of Taposiris Magna (see Fig. 1 in Boussac & El Amouri this volume) De Cosson 1935: 111 Rodziewicz 1990: 72-4 . Another prominent structure in Taposiris Magna is a 17 m high tower that stands on the coastal ridge to the north of the harbour over looking the Mediterranean coast to the north and the lake to the south (Kadous 2001: 457-60). The function of the tower and its relation with the town and harbour of Taposiris Magna is disputed (El-Fakharani 1974; Vör ös 2001: 37). Nonetheless, it is generally accepted that the tower, which represents a 1:4 or 1:5 replica of the Pharos lighthouse of Alexandria, was in fact a funerary monument for a Hellenistic necropolis that occupied the area around and below the tower (Empereur 1998: 225). The utilisation of such a structure as a landmark by navigators on both the sea and the lake is a possibility that cannot be overlooked (Fig. 3)

The harbour of Taposiris Magna was constructed to control the movement of vessels travelling through it. This was achieved by digging a channel c. 1,700 m long and 50 m wide parallel to the northern shore of the lake. The spoil resulting from the digging was piled up to form an artificial ridge which delimits the channel from the south. At the western end of the channel stood a limestone structure that took the form of a double-opening gate or bridge, through which all boats wishing to go through Taposiris Magna had to pass Empereur 1998: 225-7 Rodziewicz 1998a: 102, n. 32; Vör ös 2001: 15-6) (see Figs. 5-7 in Boussac & El Amouri, this volume). The total width of the gate is about 8.3 m, however, it is divided by a 1.2 m thick wall into two openings; one is 4.1 m wide and the other one is 3 m wide, thus indicative of the maximum possible width of the vessels that passed through. The eastern entrance of the channel is partially obstructed by a quay which is c. 230 m long extending from north to south perpendicular to the shoreline. The distance between the southern end of the quay and the eastern end of the artificial ridge, c. 100 m, forms the eastern entrance of this semi-closed harbour basin of Taposiris Magna. The eastern quay of the harbour includes at least two de-silting open-

As a result of the recent excavation of the area, it is now believed that the digging of the channel as well as the construction of this harbour system, took place the during the Early Roman period rather than during the Hellenistic period, as was previously believed (El-Fakharani 1974; Boussac & El-Amouri this volume). Although the northern shoreline at Taposiris Magna was occupied during the Hellenistic period, as evidenced by houses and shops from the 2nd and 1st century BC, it seems that it was abandoned by the end of the Hellenistic period as a result of a rise in the lake level. The flooded area was then excavated in the Roman period to create the closed harbour system (see Boussac & El Amouri this volume). Marea/ Philoxenité is located about 15 km east of Taposiris Magna, on the southern shore of the lake. El-Falaki 1966: 96 identified this settlement and its associated harbour as the town of Marea, the capital of the Mareotic region. According to Herodotus (2.31), Marea was a post of Egyptian soldiers guarding the Libyan border during the time of King Psammetichus of the 26th Dynasty Rodziewicz 2003: 27 . In the Ptolemaic and Roman periods, Marea functioned as a major trade centre, second only to Alexandria. In the Byzantine period, in addition to its involvement in commercial activities and internal trade, Marea flourished as a stopover for pilgrims heading to the holy Byzantine shrine of St. Minas Abu Mina , 20 km south of the lake Gabel & Petruso 1980 Kucharczyk 2002 Rodziewicz 2003 . ntil recently, most archaeological research carried out in the area has revealed no evidence earlier than the 5th century AD. However, recent archaeological investigation at Marea has revealed material

Fig. 3: Taposiris Magna tower ( photo: E. K halil) .

137

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig. 4: The middle quay of the harbour of Marea ( photo: E. K halil) . Excavations on the peninsula at the easternmost part of Marea, resulted in the discovery of a large Early Roman building which consists of a courtyard surrounded by numerous rooms of relatively similar size. The building is connected to a quay to the north through stairways cut into the rock. Therefore, it has been suggested that the building could have been used for storage and trade. The remains of 1st century BC to 1st century AD workshops for metalwork were also discovered on the peninsula (see Pichot this volume). The recent archaeological discoveries at Marea, particularly of pottery and coins, would indicate that the area was thriving before the 5th century and possibly as early as the Hellenistic period

from the Hellenistic and Roman periods (see both Pichot and Babraj & Szymańska this volume . Amongst the most significant archaeological remains in Marea are four quays that extend into the lake and divide the 1.5 km long shoreline into eastern, central and western harbour basins (Figs. 4 & 5). The dimensions of the quays from west to east are: 41 m x 6.5 m, 111 m x 5 m, 125 m x 7 m, and 35 m x 4 m Szymańska & Babraj 2008: 1115). Judging from the construction technique of the quays at Marea, in which large regular limestone blocks (1 m x 0.5 m x 0.03 m) were used for their construction, it seems reasonable to suggest that the four quays were constructed earlier than the Byzantine city. However, they were probably subject to several building phases in subsequent periods since evidence of Byzantine hydraulic mortar opus signinum) can still be seen between many of the building blocks of the quays.

2 - The second category of waterfront sites that was recorded along the shores of the western Mareotic Arm consists of different types of anchorage facilities such as quays and jetties, which form the majority of maritime Fig. 5: The quay of Marea harbour was made of large regular limestone blocks without evidence of the use of mortar ( photo: E. K halil) .

138

E. KHALIL: MAREOTIS’ WATERFRONT

Fig. 6: The box-shaped harbour on the southern shore of Lake Mareotis ( photo: Lake Mareotis Research Proj ect) . installations in the region. More than ten different anchorage facilities were recorded on the northern and southern shores of the lake and on the northern shore of Mareotis Island. Possibly the most substantial of them is a Kibotos or box-shaped harbour (Site 09) which is located at the end of a promontory on the southern shore of the lake, about 2 km to the west of the Sidi Kerir-Borg El Arab road. The harbour, which is constructed of large limestone blocks, consists of two parallel moles enclosing an area some 60 m long (N to S) and 36 m (E to W) wide (Fig. 6). The eastern mole is 60 m long and at its northern extremity it returns to the west for a distance of 12 m, while the western mole, which is less well preserved, extends for 40 m and returns to the east at the northern end for a distance of some 6 m. A gap of 18 m between the two ends of the two moles equates to the entrance of the harbour on the north side. The moles are constructed of up to three courses of single and double breadth limestone blocks. Although the dating for this harbour is still uncertain, judging from its construction technique and from the large size of blocks used c. 1.10 x 0.7 x 0.5 m), it seems that the harbour is pre-Roman in date. However, the existence of lumps of coarse red mortar opus signinum with lime inclusions between some of the blocks indicate that it remained in use at least until the Byzantine period. ne carved mooring ring was noted on the upper course of one of the blocks, which would have helped facilitate the mooring of vessels to the outside of the harbour (El-Fakharani 1984; Blue & Ramses 2006).

They are Sites 204 (Gamal) and 208 (Quseir). The jetties at each of these sites, which are located approximately 2 km apart, are about 50 m long and 8 m wide. Evidence of red mortar opus signinum is noted between some of the blocks. Also the jetty at Site 208 (Quseir) had mooring stones extending from the upper course of blocks at the eastern side (the lee side) of the quay. Other jetty-like features, although not as substantial, are located along the southern shore of the lake and the northern shore of the Mareotis Island. However, it was realised during the survey that the anchorage facilities along the northern shore of the lake are mostly associated with civic and residential sites, while those along the southern shore of the lake, particularly on Mareotis Island, are associated with sites of a commercial nature. The dating of these sites based on ceramic collections is quite problematic since the jetties are continuously washed by water in the winter, which, in many cases, does not leave any ceramics to be dated. However, judging from the ceramics dated from adjacent sites, it was realised that most of the sites could have been used for quite a long period of time, probably from the Hellenistic to the Late Roman or Byzantine period. 3 - The third type of maritime installation identified can be described as seawalls or more accurately lake walls. Unlike the jetties, which are perpendicular to the shore, lake walls parallel the shore and they were intended to define the shores and protect them from the effects of silting and sedimentation ig. 8 . At least five lake walls were discovered in the survey region. These kinds of structures are mainly found along the southern shore of the lake and the northern shore of Mareotis Island, those shorelines most subject to silting and the deposition of sediments as a result of the prevailing northwest winds that would carry sediment from the coastal ridges and deposit it into the lake. Besides acting as a form of protection against silting, the lake walls could also have been utilised as docking facilities for merchant vessels. Another possible function for such structures was to retain rainwater for use in agricultural purposes.

With the exception of this square harbour on the southern shore of the lake, all the other anchorage facilities along the shores of the lake take the form of jetties and quays that extend into the water perpendicular to the shore (Fig. 7). The technique used for the construction of most of the quays was building two parallel single or double breadth piers of limestone and filling the distance between them with rubble. With the exception of the substantial structures already noted at Marea/Philoxenit and Taposiris Magna the two most substantial anchorages are located on the northern shore of the lake opposite Mareotis Island. 139

LAKE MAREOTIS: RECONSTRUCTING THE PAST Fig. 7: O ne of the quays extending into the lake at the northern shore of Mareotis western arm ( photo: Lake Mareotis Research Proj ect) .

Fig. 8: ( below) A lake wall extending parallel to the southern shore of the lake. Evidence for red mortar (opus signinum) can still be seen between the blocks ( photo: Lake Mareotis Research Proj ect) .

The longest of these lake walls is located on the north shore at the western end of Mareotis Island (Site 21). It is c. 245 m long and 1 m wide and is constructed of a series of limestone blocks laid as stretchers along the lake edge (see Hopkinson this volume). Other examples include a wall located in the middle of the island on the north shore, the extant remains of which extended some 70 m in length and was made of one course of large limestone blocks of 0.60 m x 0.30 m x 0.25 m dimension, that were arranged as headers facing the shoreline (Site 123). Similarly, on the southern shore of the lake further substantial walls of over 250 m in length were identified at both Sites 109 and 44. 4 - The fourth and final type of waterfront structures does not necessarily have a maritime function. At a number of sites in the survey region the remains of several multi room buildings were recorded very close to the present waterline. Some structures even extend into the water (Fig. 9). Examples of this type of structure are found in Sites 117, 118 and 119 which are located at the north-eastern shore of the island. At Site 117 there are the remains of a multi-roomed building that measures 12 m EW x 17 m NS which is divided on the inside into at least four smaller rooms. Site 118, about 25 m west of Site 117, contains the remains of at least two multi-roomed structures which measure 18 m EW x 15 m NS and 20 m EW x 20 m NS. Each of them contains the remains of numerous walls which belonged to a number of internal rooms of different shapes and sizes. About 40 m to the west of Site 118, a further Site 119 contains the remains of a rectangular building that extends from the shoreline southwards for about 40 m and measures about 25 m EW. Limited excavation carried out in the middle section of this building revealed the remains of two 140

E. KHALIL: MAREOTIS’ WATERFRONT

Fig. 9: Some of the multi-room square buildings located at Site 118 along the northern shore of the Mareotis Island. The buildings could have been used for storage purposes ( image: Lake Mareotis Research Proj ect) . significant structures. The first structure is a rectangular enclosure that measures c. 9 m x 4.5 m that contained two rooms. Both rooms have almost the same dimension c. 3 m EW x 2.7 m NS. The second structure constructed on the same alignment as the first and adjacent to it to the west, is a rectangular building that measures c. 2.5 m NS and at least 5 m EW, in which remains of imported Hellenistic amphorae were discovered.

Marea and Taposiris Magna, only a distance of some 15 km, but out of a total of more than ninety archaeological sites recorded along the shores of the Mareotic Arm west of Alexandria, only four substantial sites were recorded to the west of Taposiris Magna. Thus, judging from the nature and extent of these sites, it is evident that the navigable limits of Lake Mareotis in antiquity extended west of Taposiris Magna for at least 12 km.

Accordingly, it seems reasonable to suggest that this type of waterfront structure were used as storage facilities for different merchandise and products that were traded along the Mareotic Arm.

Likewise, it is noticeable that maritime installations located on the southern shore of the lake from Marea to the eastern end of Mareotis Island, are in fact located on a ridge that extends for about 800 m from the present southern shoreline to the west. It is noteworthy that no sites were recorded on the southern shore of the lake opposite this ridge. This actually raises a question about the nature of the area between the ridge and the southern shore of the lake, and whether or not it was actually land in antiquity, that has subsequently been artificially excavated or subject to inundation due to the changes in ground water level (see Flaux forthcoming). Similarly, all the archaeological sites of a maritime nature that were recorded on Mareotis Island, where located along its northern shore, with essentially no evidence for sites either on the southern shore of the island or on the northern shore of the lake opposite. This also raises a question about the nature of this island and whether or not it was actually an island in antiquity.

By examining the remains of these structures, particularly on Maerotis Island, it becomes evident that they have undergone construction phases over successive periods. Moreover, it seems that the sections of the structures closest to the waterline were subject to accumulated sediments, and hence had to be rebuilt. In other words the different phases of building and modification of structures could be the result of adapting to changes in the waterline. a ns an S n an By looking at the distribution of archaeological sites in general and maritime sites in particular along the shores of the Mareotic Arm, it becomes evident that not only is there an apparent concentration of sites in the area between 141

LAKE MAREOTIS: RECONSTRUCTING THE PAST N avi gation in Lake Mareotis In the 1st century BC, when speaking about the water supply for Lake Mareotis, Strabo (17.1.7) mentions that it is filled by many canals from the ile, both from above and on the sides, and through these canals the imports are much larger than those from the sea, so that the harbour on the lake was in fact richer than that on the sea”. On another occasion (Strabo 17.1.22) speaks of “…s everal canals, which empty into Lake Mareotis”. As a result, it has been assumed that there was intense maritime traffic passing through the lake carrying various products and cargoes to Alexandria. Merchandise which would have been transported to Alexandria for local consumption and for transshipment to other Mediterranean harbours would have included Egyptian products such as papyrus, textile and grain (Rickman 1971: 300-6, 1980: 231-5; Lewis 1983: 165-7), as well as quarried stones from the Eastern Desert (Peacock 1992: 5-7; 2002: 426-7). It would have also included products imported via the Red Sea from India, Arabia and East Africa such as spices, tortoiseshell, frankincense, ivory, cotton, silk and gems (Strabo 2.5.12; Casson 1991: 200-212; Peacock 2002: 432-3). At the same time, Alexandria was receiving from the Mediterranean, for local consumption and for transshipment to the south, various products such as wine, oil, fine pottery, glass, timber, copper and tin. et, the role that Lake Mareotis played in this internal transport system is some what unclear.

The different nature of maritime installations located along the northern shore of the lake and those along the southern shore and on Mareotis Island, is also noteworthy. As mentioned earlier, maritime installations such as quays and jetties located the northern shore of the lake are mostly associated with large tell sites of a civic and residential nature. These tells were up to 60,000 m2 in area, and were densely occupied. They are mainly covered with building stones and the foundations of buildings, as well as the remains of several wells, cisterns and red brick basins lined with opus signinum, which could have been used in baths, houses or other urban structures. However, the situation on the southern shore is quite different since the southern shore is where most industrial and commercial sites were recorded, and hence maritime installations were mostly associates with those sites. In antiquity, almost all the amphora and wine production sites in the region, as well as sakkia installations (Empereur & Picon 1998 Rodziewicz 1998b see also the individual contributions of Blue, Hopkinson and Dzierzbicka this volume), were located on the southern shore of the lake, the focus of agricultural and industrial activities. The reason that agricultural and industrial activities where concentrated on the southern shores of the Mareotic Arm is mainly due to the difference in the topography between the northern and southern shores of the lake. The western arm of Lake Mareotis is delimited from the north and the south by two limestone carbonate ridges, of average elevation 25-35 m and average width 300 m (Said 1990: 499; Warne & Stanley 1993; El-Raey, et al. 1995: 191; Frihy, et al. 1996: 282). The northern ridge is known as the Abusir Ridge, and to the south a longitudinal depression 3-4 km wide known as Al-Alamein-Maryut Depression, extends roughly E-W, partly occupied by the western arm of Lake Mareotis. This depression is delimited to the south by another coastal ridge, known as Gebel Maryut Ridge, which is located 5-9 km south of the Abusir Ridge.

Although it is well known that Lake Mareotis was fed by means of a number of canals, which branched off the Canopic Branch, and flowed into the southern and eastern reaches of the lake, there is a considerable degree of uncertainty about the exact number, location and the routes of these canals. However, the most important of these canals was Schedia Canal see Bergmann, Heinzelmann & Martin this volume) It bifurcated off the Canopic Branch of the Nile at the town of Schedia, originally a Hellenistic foundation that was later known as Chaereu, currently located in the region of the villages of Kom El-Giza, Kom El-Nashw and Kom El-Hamam, some 30 km south-east of Alexandria Bergmann & Heinzelmann 2004 . While the Canopic Branch continued north to debouch into the Canopic Bay (Abukir Bay), the Schedia Canal turned northwest towards Alexandria and followed a course close to the present course of the Al-Mahmoudeyah Canal. In a statement by Strabo (17.1.16) in which he describes the town of Schedia, he mentions that it has “…t he station for paying duty on the goods brought down from above it and brought up from below it; and for this purpose, also, a schedia float has been laid across the river, from which the place has its name”. Accordingly, Schedia was the main Nile emporium, customs harbour and checkpoint east of Alexandria, where custom duties were imposed on imported and exported goods (Empereur 1998: 225; Bergmann & Heinzelmann 2004 . Moreover, it seems that the Canopic Branch at Schedia was obstructed by some kind of a pontoon that prevented boats from sailing past it until duties were paid on merchandise travelling both ways. Additionally, it was at Schedia where exported commodities

Accordingly, the distance between the Gebel Maryut Ridge and the southern shore of the lake is far greater than the distance between the Abusir Ridge and the northern shore of the lake. In antiquity, the area south of the lake was a fertile plain that flourished with agricultural activities and was known for the quality of its vines, olives and fruits as well as for the cultivation of flax and papyrus Athenaeus 1.33.d-e; Pliny 13.22.71; Strabo 17.1.14; Empereur & Picon 1998; Horden & Purcell 2002: 353; McGovern 2003: 121-3). Moreover, the abundance of calcareous clay, particularly suitable for amphora production, resulted in a thriving large-scale amphora industry along the southern shore of the lake (Empereur & Picon 1986: 103-9, 1992, 1998 Rodziewicz 1998 Blue & Ramses 2007 . Thus, the focus of wine and amphora production in Hellenistic and Roman times was the southern shores of the Mareotic Arm. Thus, it is reasonable to suggest that those maritime installations located along the southern shore of the lake were very much involved in commercial activities including the transport of Mareotic products to Alexandria and possibly also to the southern limits of the lake. 142

E. KHALIL: MAREOTIS’ WATERFRONT brought from upriver were transferred from large Nile boats to smaller boats that could travel easily through the canals to Alexandria (Procopius 6.1.3; Hassan 1997: 365 n. 13).

sequently, the lake’s southern and eastern shoreline was unstable and subject to constant change, and was therefore unsuitable for the establishment of substantial harbours and waterfront installations. A recent survey conducted along the ancient southern and eastern limits of the lake (Wilson 2007; see Wilson this volume) revealed that most settlements established in this area during the Ptolemaic and Roman periods, were located on high ground around the lake’s edge. Also it revealed that many settlements were involved in agriculture and industrial activities mainly in the service of Alexandria. However, as yet there is no evidence for substantial maritime structures or significant waterfront installations.

As it approached Alexandria, the Schedia Canal bifurcated into two branches in the Alexandrian suburb of Eleusis El- ozha . The first branch turned towards the north-east leading to Canopus, east of Alexandria, while the other branch continued south of Alexandria and parallel to the lake’s northern shore, until it debouched into Lake Mareotis south-east of Alexandria. According to Strabo (17.1.7), boats also sailed from the Nile to the Canopic Branch and through the network of canals that fed the lake from the south and east, then across the lake northwards to Alexandria. This indicates that navigation on Lake Mareotis was intense and operated in many directions. It also raises a point about the practicalities of sailing in Lake Mareotis from south to north against the prevailing north-westerly wind. The predominant winds along the north coast of Egypt are north-westerly and they prevail more than 40% of the time throughout the year and more than 70% of the time during the summer sailing season (El-Zouka 1979: 125-7; El-Gindy 1999: 17). Accordingly, merchant vessels sailing in Lake Mareotis from south to north would have faced a direct headwind, which meant that the boats had to tack in order to reach Alexandria. Tacking in Lake Mareotis was possible considering the large area of the water body; however, tacking from the southern limits of the lake to Alexandria would have meant that boats would have to travel several times the direct distance across a water body full of shallows and marshlands and against prevailing winds. In the 5th century St. Palladius (7.1) mentioned that he sailed across Lake Mareotis from north to south, from Alexandria to the monastic settlement of Mount Nitria, a distance of about 50 km, in a day and half. Accordingly, sailing in the lake in the opposite direction would have taken much longer, possibly as long as four to five days.

C onclusion There were two ways for river vessels to travel to and from Alexandria, either across the lake, or along the Schedia Canal. Considering the arguments outlined above, particularly in relation to the prevailing winds, it seems reasonable to suggest that the main northbound traffic probably went via the Canopic Branch and the Schedia Canal, rather than across the lake. However, sailing south across the lake would have been conducted with considerable ease. Along those stretches of the canal where boats had to maneouver against the wind, they could have been towed along from the shore, a standard procedure for moving river boats in rivers and canals around the world. In that respect, Strabo’s statement (17.1.7) about the lake harbour south of Alexandria being richer than the seaport of Alexandria, would still be valid. At the time of Strabo, the Schedia Canal debouched into Lake Mareotis, so all the canal traffic had to pass through the lake. Moreover, boats coming from the western arm of the lake also arrived at the lake harbour. Therefore, it is possible that the lake harbour was quite busy receiving river vessels from the south as well as from the west. At the same time, it is not unreasonable to suggest that east–west commercial traffic along the western Mareotic Arm was probably more intense and more regular than the north south traffic that passed through the main body of the lake. Recent archaeological investigation in the Mareotic region have revealed that the number, nature and extent of archaeological sites along the shores of the Mareotic Arm, is unparalleled any where else in the Western Deltaic Region (see Blue this volume). Settlements in this region were located far from the silting effects of the Nile sediments, the coastline was more stable, prevailing winds were more favourable for east-west movement, and settlements were in close proximity to Alexandria. Therefore, it is understandable why so many shoreline settlements and associated maritime installations were established along its shores. Thus, the contribution of the western Mareotic Arm to the economy of ancient Alexandria and hence of Egypt as a whole, was probably far more significant than any other area along the shores of Lake Mareotis. Thus, the shores of the western arm of Lake Mareotis appear to have been one of the most active areas of economic activity in the Western Deltaic region during the Hellenistic, Roman and Byzantine periods.

Furthermore, the extended period of travel across the lake would have laid boats venerable to another challenge that prevailed on Lake Mareotis in antiquity. Achilles Tatius (4.12) in the 2nd century and Heliodorus (1:14) in the 3rd century, spoke of piracy and bandits on Lake Mareotis. The marshes and islands of the lake provided excellent hideouts for groups of bandits and their vessels. Also, the large size of the lake made it quite difficult to guard and control, therefore, it is possible that sailing across the lake with valuable commodities was quite risky. Moreover, settlements located on the southern and eastern shores of Lake Mareotis were far more susceptible to sedimentation from silts deposited via the nearby Canopic Branch of the ile, particularly during flood seasons, as well as sediment which had been carried by the prevailing winds across the lake from the north-west to the southeast. All this would have contributed to the build up of sediments against the southern and eastern shores of the lake, thus preventing settlements in this region from being as actively involved in across lake transportation. Con143

LAKE MAREOTIS: RECONSTRUCTING THE PAST Bibliography Achilles Tatius, The Adventures of Leucippe and Clitophon, S. Gaselee (transl.), 1969. London. Athenaeus, eipnosophistae C.B. Gulick (transl.), 1953. London. Bergmann, M., & Heinzelmann, M., 2004, Schedia ( K om el Gi ah om el amam epartment of eheira Report on the documentation and excavation season. 18 March – 18 April 2003. Paper presented at conference on Alexandria in Antiquity, University of Oxford, December 2004. Blue, L., & Ramses, S., 2006, Lake Mareotis Research Proj ect. Report submitted to the Egyptian Supreme Council for Antiquities on the fieldwork and results of the September 2006 field season. Blue, L., & Ramses, S., 2007, Lake Mareotis Research Proj ect. Report submitted to the Egyptian Supreme Council for Antiquities on the fieldwork and results of the May & uly/August 2007 field seasons. Casson, L., 1991, The Ancient Mariners. Princeton. De Cosson, A., 1935, Mareotis: Being an Account of the istory and ncient Monuments of the orth est Desert of Egypt and Lake Mareotis. London El-Ashmawi, F., 1998, Pottery Kiln and Wine Factory at Burg El-Arab. In .- . Empereur ed. , Commerce et artisanat dans l le andrie hell nisti ue et ro maine. ctes du collo ue d th nes organis par le C R le Laboratoire de c ramologie de Lyon et l Ecole fran aise d th nes d cembre 55-64. Athens, Paris. El-Fakharani, F., 1974, The Lighthouse of Abusir in Egypt. Harvard Studies in Classical Philology 78: 257-272. El-Fakharani, F., 1983, Recent Excavation at Marea in Egypt. In G. Grimm, H. Heinen & E. Winter (eds.), egyptiaca re ernsia II as R misch y antin ische Ägypten. Akten des internationalen Symposions 26.-30. September 1978 in Trier, 175-186. Mainz. El-Falaki, M.S., 1966, Ancient Alexandria. Alexandria: ), Dar Nashr Al-Thaquafa ( M.B. Astronome (transl.), 1872. Copenhague. El-Gindy, A., 1999, Meteorological and Hydrodynamic Conditions in the Eastern Harbour of Alexandria and its Vicinity. In . Halim ed. , Proceedings of the orkshop on the tatus of Pilot Pro ect for the ustain able Development of the Submarine Archaeological Sites at Q ayetbey Citadel and Eastern Harbour of le andria o ember Reports on ydro dynamics Geophysics Morphology and Geology, V. II: 1-82. Alexandria. El-Raey, M., Nasr, S., Frihy, O., Desouki, S., & Dewidar, Kh., 1995, Potential Impact of Accelerated Sea-Level Rise on Alexandria Governorate, Egypt. J ournal of Coastal Research 14: 190-204. El-Zouka, M.K., 1979, Irrigation reas in the estern Delta: A G eographical Study ( 8´ ), Alexandria (in Arabic). Empereur, .- ., & Picon, M., 1986, A La Recherche des ours d Amphores. In - . Empereur & . Garlan (eds.), Recherches Sur Les Amphores G recques. Actes

du colloque international organisé par le Centre national de la recherche scientifi ue l ni ersit de Rennes II et l cole fran aise d th nes th nes 10-12 Septembre 1984) : 103-126. Athens, Paris. Empereur, .- ., & Picon, M., 1998. Les Atelier d Amphores du Lac Mariout. In .- . Empereur ed. , Commerce et artisanat dans l le andrie hell nisti ue et romaine. ctes du collo ue d th nes organis par le C R le Laboratoire de c ramologie de Lyon et l Ecole fran aise d th nes d cembre 75-88. Athens, Paris. Empereur, .- ., & Picon, M., 1992, La Reconnaissance Maréotide. Extrait des Cahiers de la Céramique gyptien 3: 145-152. Empereur, .- ., 1998, Alexandria Rediscovered. New ork. Flaux, C., forthcoming, Geomorphological survey of the south-west shores of Lake Mareotis. In L. Blue & E. Khalil (eds.), A Multidisciplinary Approach to Alexandria s Economic Past the Mareotis Case tudy. Oxford. Frihy, O.E., Dewidar, Kh.M., & El-Raey, M.M., 1996, Evaluation of Coastal Problems at Alexandria, Egypt. O cean & Coastal Management 30.2-3: 281-295. Gabel, C., & Petruso, K., 1980, An Environmental and Cultural Study at Lake Maryut, Lower Egypt. Boston ni ersity frican tudies Center orking Papers No. 25. Boston. Petruso, K., & Gabel, C., 1982, Marea: A Byzantine Port in Northern Egypt. oston ni ersity frican tudies Center orking Papers No. 62. Boston. Hassan, .A., 1997, The Dynamics of Riverine Civilization: A Geoarchaeological Perspective on the Nile Valley, Egypt. orld rchaeology 29.1: 51-74. Heliodorus. Aethiopica, T. Underdowne (transl.), 1895. London. Herodotus. The Histories, A. de Sélincourt (transl.), 1971. Middlesex. Horden, P., & Purcell, N., 2002, The Corrupting Sea. Oxford. Kadous, E., 2001, Monuments of Ancient Alexandria. ( ) Alexandria (in Arabic). Kucharczyk, R., 2002, Marea 2001: Windowpanes and other Glass inds. In M. Gawlikowski & W. A. Daszewski (eds.), Polish Archaeology in the Mediterranean XIII: 65-71. Lewis, N., 1983, Life in Egypt under Roman Rule. Oxford. McGovern, P.E., 2003, ncient ine he earch for the O rigins of Viniculture. Princeton. Palladius, The Lausiac History. The Monks of Nitria. VII. I. http://www.fordham.edu/halsall/basis/palladiuslausiac.html (accessed February 2010). Peacock, D., 1992, Rome in the Desert: A Symbol of Power. An Inaugural Lecture delivered at the University of Southampton. Peacock, D., 2002, The Roman Period. In I. Shaw (ed.), The O xford History of Ancient Egypt: 422-445. Oxford. Pliny the Elder, The Natural History, J. Bostock & B.A. Riley (transl.), 1855. London. 144

E. KHALIL: MAREOTIS’ WATERFRONT le Laboratoire de c ramologie de Lyon et l Ecole fran aise d th nes d cembre 27-36. Athens, Paris. Rodziewicz, M.D., 2003, Philoxenit , Pilgrimage Harbour of Abu Mina. Bulletin de la Société Archéologi ue d le andrie 47: 27-47. Said, R., 1990, Quaternary. In R. Said (ed.), The G eology of Egypt: 487-507. Rotterdam. Strabo, G eography, H.L. Jones (transl.), 2001. London. Szymańska, H., & Babraj, K., 2008, Marea V.1: Byzantine Marea – Excavations in 2000-2003 and 2006. Krakow. Vör ös , G., 2001, Taposiris Magna: Port of Isis. Budapest. Warne, A.G., & Stanley, D.J., 1993, Late Quaternary Evolution of the Northwest Nile Delta and Adjacent Coast in the Alexandria Region, Egypt. J ournal of Coastal Research 9.1: 26-64. Wilson, P., 2007, estern elta Regional ur ey. http://www.dur.ac.uk/penelope.wilson/Delta/AbuMatamir.html (accessed February 2010).

Procopius, The Buildings, H.B. Dewing (transl.), 1940. London. Rickman, G., 1971, Roman G ranaries and Store Buildings. Cambridge. Rickman, G., 1980, The Corn Supply of Ancient Rome. Oxford. Rodziewicz, M.D., 1990, Taenia and Mareotis: Archaeological Research West of Alexandria. Acta of the First International Colloquium of the Egyptian Society of G reek and Roman Studies 1: 62-80. Cairo. Rodziewicz, M.D., 1998a, rom Alexandria to the West by Land and by Waterways. In .- . Empereur ed. , Commerce et artisanat dans l le andrie hell n isti ue et romaine. ctes du collo ue d th nes organis par le C R le Laboratoire de c ram ologie de Lyon et l Ecole fran aise d th nes 11-12 décembre 1988: 93-103. Athens, Paris. Rodziewicz, M.D., 1998b, Classification of Wineries from Mareotis. In .- . Empereur ed. , Commerce et artisanat dans l le andrie hell nisti ue et romaine. ctes du collo ue d th nes organis par le C R

145

LAKE MAREOTIS: RECONSTRUCTING THE PAST

146

S. RAMSES & A. OMAR: OUTRAGE AND DESTRUCTION

Lake Mareotis Research Project. Phases of Outrage and Destruction Sameh Ramses & Ahmed Omar

Interest and Aim of the Project The historical importance of the region of Mareotis during the Ptolemaic and Roman periods, either as an important wine production center, in support of other Egyptian industries (Strabo 17.1.7), or as a resort of the elite (Buttler 1989: 8-12), has attracted a great deal of archaeological attention, as this volume illustrates (El-Fakharani 1983; Empereur & Picon 1998; Rodziewicz 1998). Nevertheless, the region of Mareotis is threatened by modern urban development and many historical sites are subjected to destruction, not only due to its unique location in the vicinity of Alexandria, but also its moderate climate which has established the area as one of the most popular resorts in the region (Abd-Elhady 2008: 177-184). Therefore, the aim of the Lake Mareotis Research Project, conducted between the summer of 2004 until the summer of 2008, was to survey, record and map the platforms, harbours, and any other waterfront constructions that could be found along the northern and southern shores of the western part of Lake Mareotis, and to create a very detailed database for each site, as well as to determine its present state of preservation.

Phases of Outrage and Destruction According to the last survey of 2008 more than 70 archaeological sites have already been noted in the region (Fig. 2). Of course, not all of them had previously been registered, but some had been surveyed or documented.1 However, despite efforts by the Archaeological Department of Alexandria University to discover and protect some of these sites, the full extent of the area, and the variety of remains, both on land and in the water, still remains to be determined and much work needs to be done. This is particularly urgent considering the very immediate threats of modern urban development and various industrial and agricultural activities that are threatening the shores of the lake. Consequently, during the course of the survey, three main factors were identified that are believed to present a particular danger to the archaeology of the region. Fish Farms As a consequence of being inundated by water largely of agriculture runoff and drainage water, the lake water quality has deteriorated making it more difficult for a variety of fish to survive. Simultaneously, many fish-farmers began to appropriate quite large portions from the lake basin and along the shores. In the process of creating fish producing basins, a great deal of dredging and destruction of some of the archaeological sites, both offshore and under water, has taken place. Accordingly, the original shores topography has been deformed, and the dredging leads to

The survey project covers an area of about 40 km long and up to 3 km wide within the western arm of Lake Mareotis, to the west of Borg El-Arab airport, on the Sidi Kerir road, and extends towards El-Hammam city, it also includes Mareotis Island some 3.7 km long and 680 m wide (Fig. 1; see Blue, Hopkinson and Khalil this volume). Fig . 1: G eneral plan from Alexandria to Mareotis ( Lake Mareotis Research Proj ect) .

1. It is noteworthy that only about nine sites had already been registered by the Supreme Council of Antiquities (SCA), and only six are considered to be SCA properties. Atlas Almoqea Alathariaya, SCA, 2002: No. 3: El-Beheira Governorate; No. 4: The Rest of the Lower Egypt Governorates. Egypt.

147

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig. 2: Map of sites investigated during the Lake Mareotis Research Proj ect ( Lake Mareotis Research Proj ect) . Industrial Activities Although not as polluted as some parts of Lake Mareotis, the western arm determines a substantial area of the current lake littoral, and is considerably affected by a huge amount of untreated industrial waste and polluted water which flows into the lake on a daily basis. The industrial activities around the shores of the lake not only affect the water quality but in some cases have a direct impact on the

mis-interpretations of its original form. These activities particularly characterise the southern shore of the lake (Fig. 3). It is worth mentioning that some farms have appropriated quite large areas, up to nearly 4 hectares, especially in the vicinity of Mareotis Island, which embraces a great deal of archaeological remains, making access to the island increasingly difficult by cutting away the roads and destroying the routes that lead to the island. 148

S. RAMSES & A. OMAR: OUTRAGE AND DESTRUCTION

Fig.

o e amples of fish farm deformation along the southern shore of the lake Lake Mareotis Research Pro ect .

Fig. 4: Lime storing and collecting at Site 4 ( Mareotis Research Proj ect) .

Lake

Fig. 5: Lake wall affected at Site 4 ( Research Proj ect) .

archaeological sites located in the region. Site 44, located on the southwest shore of the lake, provides clear evidence for this negligence. It is directly adjacent to a cement factory, and is used to store the lime for the cement (Fig. 4). The cement mound is huge and it no doubt, amongst other things, covers the western end of a lake wall that extends along the lake shore (Fig. 5), and disregards the existence of additional archaeological features including a circular structure at the far west of the site near the water’s edge, and four square probable wine-basins, that were previously excavated by the SCA. Similar scenes are witnessed further west, for example at the site of Naga El-Mawalik (Site 109). This is an example of an endangered site that still displays aspects of its ancient harbour and enclosure walls, together with many other hills and mounds of pottery. Its location close to the cement factory, means that it is being used to supply the factory with limestone. Parts of this site are now totally inaccessible, with unrecorded archaeological features being surrounded by barbed wire.

Lake Mareotis

On the north shore of the lake, at the western extent of the survey area, Site 214, immediately opposite Naga El-Mawalik, has recently been purchased by the El-Nasr Salt Company. They have acquired two already excavated sites (Sites 214-215) which contain two wineries and a big house of the late Roman period and numerous mounds of pottery (Figs. 6-7). These sites are due to be dredged and leveled to be used as evaporations basins to produce salt. They have only been partially recorded. There are many other places in the area free of archaeological remains that could be used for such purposes. Other sites, particularly those on the northern shore of the survey area, are being used as rubbish tips totally swamped by rubbish from the city and isolated from the main road by means of barbed wire. These sites are distinctive by their bad smell. U rban Development Urban development is considered to be one of the most dangerous factors that affect the archaeological sites of the 149

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig. 6: ( above) Site 214: the partially excavated winery ( Lake Mareotis Research Proj ect) .

Fig. 7: ( right) Site 214: part of the fragmented paved oor of a Roman period aterfront building Lake Mareotis Research Proj ect) . region. Vast portions of land have been acquired, many divided into smaller units to be sold and used for housing. Both the local Bedouin and big companies that wield profit and power are responsible for destroying and leveling vast areas at a remarkably rapid rate. The Lake Mareotis Research Project has had direct experience of these actions over the last four years of survey. This can be summarised in the following four case studies:

documentation of the site, we returned to find that the three hills were totally destroyed by the locals and the Hellenistic tomb flattened ig. 12 . It is worth mentioning that this site had not been previously registered or recorded by the SCA. El-G amal ( Site 204) The situation at the site of El-Gamal was a little different as the site was quite impressive and important, to the extant that it had already been partially recorded by the SCA, and a site guard had been assigned. The destruction of this site was undertaken in stages first a soccer field was cleared to the east of the tell; the following year another one was cleared on the prohibited area of the archaeological site towards the limit of the lake shores, destroying part of the tell and site boundaries, and so the destruction continued (Fig. 13).

Site 09 A Kibotos harbour, the only well preserved example either on the sea, or on the lake. Nevertheless, the governorate has dedicated an area of about 8 hectares around the site to build a private university without leaving any access to the archaeological site (Figs. 8-9). Accordingly, over time, the site will be destroyed. Ez-Z eraa Elbahry ( Site 201) When visited in September 2004, there was a hill of about 5 m in height, which contained several mounds of pottery with traces of building plots everywhere. When we returned in May 2007, the 5 m high hill had disappeared and there were only private property signs and modern partitioning walls visible (Figs. 10-11).

Even those sites renowned for their archaeological importance are also under threat. To the southeast of Taposiris Magna, the extensive seawall and jetty associated with the site are slowly being encroached by small concrete building plot walls, purchased for urban development (see Boussac & El Amouri this volume).

Nagea Ez-Z ohorat ( Site 202) A similar pattern of destruction was noted at Nagea EzZohorat (Site 202) as was witnessed at Ez-Zeraa Elbahry (Site 201). The site comprised three hills containing amongst other remains, a cistern, a quay, and a rock-cut Hellenistic tomb. We revisited the site in August 2007 and just few days after we had commenced the recording and

Further destruction is caused at many sites by the introduction of agriculture, particularly the cultivation of figs and olives. Cattle are also allowed to graze freely around the whole region, including on established archaeological sites. Moreover, seasonal extraction of reeds in marshy lake side areas permits the access of heavy loaders and trucks to a number of the archaeological sites. 150

S. RAMSES & A. OMAR: OUTRAGE AND DESTRUCTION

Fig. 8: K ibotos harbour extending from the southern shore of the lake ( Lake Mareotis Research Proj ect) .

Fig. 9: O ne of the university properties on the land surrounding the K ibotos ( Lake Mareotis Research Proj ect) .

Fig. 10: Archeological tell divided for private possession ( Lake Mareotis Research Proj ect) .

Fig. 11: Private property sign and wall established on Site 201 ( Lake Mareotis Research Proj ect) .

Fig. 12: Flattened Hellenistic tomb at Site 202 during bulldozer work in 2007 ( Lake Mareotis Research Proj ect) . 151

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig.



o soccer fields erected at the tell boundaries around ite

C onclusion and Recommendation From the above mentioned, the slow destruction of archaeological sites in Mareotis can be specified as follows: Cultivating the boundaries of archaeological tell sites and in some cases, on the tells themselves. Setting up barbed wire around many of the sites to give the impression that they are private property, and when the authorities are notified, the signs are not removed. sing bulldozers and tractors to rapidly destroy and level the archaeological hills. Selling archaeological sites, especially the unregistered sites, to private individuals, by employing unorthodox means that appropriate the names of established people, to ensure that legally the sites cannot be retrieved, thus preventing anyone, even the governmental organizations, from taking any action.

Lake Mareotis Research Pro ect .

Generally, the aim of these suggestions would be to create a project which promotes a particular concept for both the short- and long-term management of the region, which considers its archaeological and historical importance. Such a plan should also aim to promote and preserve Mareotis for the coming generations. This is a great challenge that we all face. Bibliography Abd-Elhady, D., 2008, erasat fi sasyat l sayaha. Alexandria. Atlas Almoqea Alathariaya 3: El-Beheira Governorate, Egypt. Supreme Council of Antiquities, August 2002. Cairo. Atlas Almoqea Alathariaya 4: The Rest of Lower Egypt Governorates, Egypt. Supreme Council of Antiquities, August 2002. Cairo. Buttler, A.G., Fateh Elarab Lemesr, M.F. Abo Hadid & T. El-Massrein (transl.), 1989. Cairo. El-Fakharani, F., 1983, Recent Excavation at Marea in Egypt. In G. Grimm, H. Heinen & E. Winter (eds.), Aegyptiaca Trevernsia II, Das Römisch-Byzantinische Ägypten. Akten des internationalen Symposions 26.-30. September 1978 in Trier, 175-186. Mainz. Empereur, J.-Y., & Picon, M., 1998, Les Ateliers d’Amphores du Lac Mariout. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athè nes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole franç aise d’Athè nes, 11-12 décembre 1988: 75-88. Athens, Paris. Rodziewicz, M.D., 1998, Fom Alexandria to the West by Land and by Waterways. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athè nes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole franç aise d’Athè nes, 11-12 décembre 1988: 93-103. Athens, Paris. Strabo, G eography, H.L. Jones (transl & ed.), 1996. London, New York.

Accordingly, it has been established that the problems of the region are extensive and complex, but once identified can be resolved by the means of establishing a compromise between protecting Mareotis’ archaeological sites, whilst at the same time meeting the increasing needs of development and modernization. Therefore, the following is suggested to overcome the problems: Coordination between the different regional authorities including those that operate in the interest of archaeology, investment, irrigation, and the governorate, to establish an integrated long term plan, which recognises a satisfactory compromise that identifies the needs of all parties. nifying the mapping system of each organization, in order to get acquainted with each others properties, in order to stop the illegal acquisition of land. Support the Supreme Council of Antiquities, both to provide the facilities needed to control and protect such a vast area, and to implement archaeological legislation. By encouraging local interest in the importance of the sites both as a source of income, and as an important aspect of cultural heritage. By promoting the whole region to become an open archeological park in order not to be neglected and misused, otherwise it will soon disappear. 152

S. RAMSES & A. OMAR: OUTRAGE AND DESTRUCTION

153

LAKE MAREOTIS: RECONSTRUCTING THE PAST

154

INDEX

Select Index Places/topography Abu Mina, 41, 54, 67-73, 77-78, 81, 115, 127-130, 137 Abuqir Lake, 13, 15-16, 18 Abusir, 15, 35 Abusir Ridge, 36, 43, 142 Alexandria, 1-4, 7-8, 12-13, 15-16, 18-20, 25-28, 32-33, 35-36, 47, 49, 54-55, 57, 66-67, 69, 71-72, 75-78, 8283, 87, 89, 100, 103, 107-108, 110, 113, 115-116, 123125, 127-128, 135, 137, 141-143, 147 Beheira, 37, 107-115, 119, 122, 124-125, 147 Cairo, 13, 15, 67, 107, 123, 125, 128 Canopic Branch/Canopic Nile, 4, 7, 13, 18, 25, 36, 75, 107-108, 110, 116, 119, 122, 124-125, 135, 142-143 Chaereou, 7 Fayum, 44, 110, 127 Huwaryia/Hauwariya, 68-70, 75, 78 Kanubiye Canal, 108, 114 Kom el Giza (see also Beheira), 107-115 Lake Mareotis/Lake Mariout/Maryut Lake, 1-4, 7-8, 1119, 21-23, 25-28, 30-33, 35, 37, 44, 47-48, 57, 66-67, 72-73, 75-77, 83, 87, 95-96, 119, 121, 123-124, 128129, 135-137, 139-143, 147-152 Libya, 1, 3-4, 8, 75-77, 157 Marea/Mareia (see also Philoxenité), 1-4, 7-8, 14-15, 17, 27-28, 30-32, 35, 37, 44, 47-52, 54-55, 57-61, 66-67, 69-73, 75, 77-83, 91, 100, 128-129, 135-139, 141 Mareotic Arm, 32-33, 135-136, 138, 141-143 Mareotis Island, 28-32, 35, 37, 41-42, 44, 135-136, 139142, 147-148 Maryut Depression, 25, 35, 142 Mediterranean Sea, 11-12, 15-17, 19, 25, 27, 127-128 Naukratis Canal, 123, 125 Nile Delta, 36-37, 44, 83, 107, 110, 135, 137 Nile River, 3-4, 7-8, 11-13, 15-16, 18, 25, 27, 35-36, 39, 44, 50, 54, 66, 75, 83, 95-96, 103, 107-108, 110, 116, 125, 127, 135, 142-143 Nile Valley, 3, 27, 35, 41, 44, 54, 103, 127 Nitria/Nitraria, 8, 13-15, 143 Philoxenité/Philoxenite, 27, 35, 47, 54-55, 67, 69, 71-73, 75, 77-78, 83, 135-137, 139 Plinthine, 4, 14, 67, 87-88 Sais, 119, 124 Schedia, 7, 13, 107-110, 114-116, 125, 129, 142 Schedia Canal, 18, 142-143 Taposiris Magna/Taphosiris, 7-8, 27, 67, 71, 87-88, 92, 101, 136-137, 139, 141, 150 Western Desert, 3, 5, 123-124

El-Fakharani, F., 1-2, 27, 31, 37, 47, 49-53, 69-70, 76, 78, 129 El-Falaki, M., 1, 17-18, 37, 47, 57, 67, 75, 137 Herodotus, 3, 7, 54, 73, 75, 137 Justinian, 7-8, 76 Pliny, 4 Procopius, 7-8, 55 Sophronius, 8 St Menas, 51-52, 54, 78 Strabo, 3-4, 7-8, 18, 35-36, 39, 47, 54, 66, 75, 127, 142-143 Theodosius, 8 General amphora/amphorae, 27-29, 32, 37, 43-44, 58, 62, 72-73, 75, 78, 81-83, 87, 90, 100-103, 113, 115-116, 121, 124, 127, 135-136, 141-142 auger cores/auger survey, 30, 37-38 basilica, 51-53, 75-78, 81-83, 129 bath complex/bath facility, 28, 58, 70-72, 75, 110, 116, 128-129 bridge, 42, 89-92, 100, 107-108, 116, 137 camel, 4-5, 55 canal, 3-4, 7-8, 14-15, 18-19, 21, 27, 35-36, 50, 52, 66, 75, 91, 95-96, 107-108, 110, 114, 116, 121, 123-124, 135, 142-143 causeway, 25, 58, 60, 71, 77, 87, 89-91, 93, 95, 100, 135 cistern, 2, 28, 36-37, 42, 53, 58, 69-70, 72, 80, 136, 142, 150 coins, 51-53, 70, 79, 82, 91, 96, 100-101, 103, 107, 110, 113-115, 128 Coptic Encomium of St Menas, 54-55, 69, 72, 77 evaporation, 37, 149 finewares, 28, 121 fish farms, 19-22, 96, 100, 147, 149 fish-tank, 96-100, 103 floods/seasonal flooding, 4, 16-17, 22, 35-37, 39, 43, 54, 87, 90-91, 95, 124, 137, 143 forts, 49, 124 glass, 5, 25, 27, 52, 66, 75, 121, 142 harbours, 4, 7, 25, 27, 31, 38-39, 47-51, 53-55, 57, 66-73, 75-78, 81, 83, 87, 90-91, 93, 95-96, 98, 100-101, 103, 107-108, 110, 116, 124, 136-139, 142-143, 147, 149151 hypogea/ hypogeum, 54, 69, 71 jetty/jetties, 28-29, 32, 37, 39, 41-43, 47, 49, 53-53, 58, 60, 67, 71-72, 75, 77-78, 81, 83, 89, 91, 93-98, 100-101, 103, 136, 138-139, 142, 150 karum, 36, 43, 71 Kibotos/Kibotos harbour, 31, 47-48, 71, 75, 91, 139, 150151 kiln, 31-32, 36-37, 43-44, 70-73, 75, 77-78, 82-83, 127, 130, 136 lake port, 67, 70, 72, 75 lake wall, 30, 32, 39, 43, 139-140, 149

People Athanasios, 77 Christodorus, 8 Claudius Ptolemy/Ptolemy the Geographer, 4, 36, 47, 54 Coste, P., 87, 89-91 De Cosson, A., 1, 35, 67, 75, 87, 89

155

LAKE MAREOTIS: RECONSTRUCTING THE PAST lighthouse, 49, 58, 137 mole, 28, 31, 41, 67, 139 natron, 3, 5, 15 papyri/papyrological documents, 4-5, 8, 96, 110, 130-132 papyrus, 7, 54, 75, 142 Pharos, 3, 137 pilgrim, 54-55, 69, 71-73, 77-78, 82, 137 pilgrimage, 66, 68-72, 78 pilgrimage centre, 67, 71-72 pollution, 27-28 port, 14, 25, 27, 42, 47, 49-52, 66-67, 69-73, 75, 78, 83, 96, 123-124, 135 qadus, 28 quarry, 27, 32, 36, 43, 67, 71, 77, 91, 107, 116, 129, 143

quay, 28-29, 32, 39, 41, 43, 47, 49-50, 66-67, 72, 77, 136140, 142, 150 sakkia sakia sa iya s iyah, 28, 30, 32, 37, 41, 76, 79-81, 83, 136, 142 sebakh, 107, 110, 123 sewage, 25, 27, 77, 83 tombs, 2, 17, 32, 49, 54, 69, 71, 111, 116, 150-151 villa, 35, 43, 58, 77-78, 111, 113-114, 116, 124 vineyards, 5, 43, 75, 78, 127, 130-132 wine, 4-5, 25, 27, 43, 53-54, 70, 75, 77-78, 83, 101, 103, 121, 127-132, 142, 149 wine production, 30, 32, 70, 114, 127, 130, 132, 135-136, 142, 147 winery, 32, 43-44, 53, 69-70, 123, 127-132, 135, 149-150

156

Contents CONTENTS LIST OF FIGURES, TABLES AND MAPS INTRODUCTION CONTRIBUTORS

i ii vii viii

1. FAWZI EL-FAKHARANI: PIONEER EXCAVATOR AT MAREOTIS. Mona Haggag

1

2. THE MAREOTIC REGION IN ANCIENT SOURCES. Mohamed S. Abd-el-Ghani

3

3. A NOTE ON LAKE MAREOTIS IN BYZANTINE TIMES. Mostafa El Abbadi

7

4. A STUDY OF THE EVOLUTION OF THE MARYUT LAKE THROUGH MAPS. Ismaeel Awad

11

5. LAKE MAREOTIS RESEARCH PROJECT. Lucy Blue

25

6. THE RESULTS OF A PRELIMINARY SURVEY AT MAREOTIS ISLAND. Dylan Hopkinson

35

7. THE CITY OF MAREA/PHILOXENITÉ. REFLECTIONS ON THE ALEXANDRIA

47

UNIVERSITY EXCAVATIONS, 1977-1981. Mona Haggag 8. MAREA PENINSULA: OCCUPATION AND WORKSHOP ACTIVITIES ON THE

57

SHORES OF LAKE MARIOUT IN THE WORK OF THE CENTER D’ÉTUDES ALEXANDRINES (CEAlex, CNRS USR 3134). Valérie Pichot 9. ON INTERPRETATIONS OF ARCHAEOLOGICAL EVIDENCE CONCERNING

67

MAREA AND PHILOXENITE. Mieczyslaw D. Rodziewicz 10. MAREA OR PHILOXENITE? POLISH EXCAVATIONS IN THE MAREOTIC REGION 2000-2007.

75

Krzysztof Babraj & Hanna Szymańska

11. THE LAKE STRUCTURES AT TAPOSIRIS. Marie-Françoise Boussac & Mourad El Amouri

87

12. SCHEDIA, ALEXANDRIA’S HARBOUR ON THE CANOPIC NILE. INTERIM REPORT ON

107

THE GERMAN MISSION AT KOM EL GIZA/BEHEIRA (2003-2008).

Marianne Bergmann, Michael Heinzelmann & Archer Martin

13. RECENT SURVEY WORK IN THE SOUTHERN MAREOTIS AREA. Penelope Wilson

119

14. WINERIES OF THE MAREOTIC REGION. Dorota Dzierzbicka

127

15. WATERFRONT INSTALLATIONS AND MARITIME ACTIVITIES IN THE

135

MAREOTIC REGION. Emad Khalil 16. LAKE MAREOTIS RESEARCH PROJECT. PHASES OF OUTRAGE AND DESTRUCTION.

147

Sameh Ramses & Ahmed Omar

SELECT INDEX

155 i

LAKE MAREOTIS: RECONSTRUCTING THE PAST

List of Figures, Tables and Maps Figures Chapter 4. Ismaeel Awad 1: Maryut Lake region. 2: 1548, Vray portraict de la ville d’Alexandrie en Egypte, Pierre Belon du Mans. 3: 1570, detail from Aegyptus antiqua, including a focused map of the region of Alexandria, Abraham Ortelius. 4: 1588, detail from Africae Tabula VII, Livio Sanuto. 5: 1655, detail from Royaume et désert de Barca et l’Egypte, Nicolas Sanson D’Abbeville. 6: 1717, detail from Carte de la Basse Egypte et du cours du Nil, Paul Lucas. 7: 1722, detail from Carte de l’Egypte Ancienne, Claude Sicard. 8: 1743, detail from Carte d’Egypte, Richard Pococke. 9: 1753, detail from Carte de l’Egypte ancienne et moderne, Robert de Vaugondy. 10: 1764, detail from Carte des Embouchures du Nil,e t partie de son cours, Jacques Nicolas Bellin. 11: 1801, Carte d’Alexandrie, French campaign. 12: 1801, Carte d’Alexandrie, French campaign, original scale 1:100,000. 13: 1866, Carte des Environs d’Alexandrie, Mahmoud Bey El-Falaki. 14: 1866, Carte des Environs d’Alexandrie, Mahmoud Bey El-Falaki, original scale 1:200,000. 15: 1911 (1914), Atlas of Egypt, Vol.I ,L ower Egypt,Sur vey of Egypt (G. Daressy), original scale 1:50,000. 16: 1949, map series 1:100.000. Survey of Egypt, original scale 1:100,000. 17: 1970s, map series 1:25.000, Egyptian General Survey Authority, original scale 1:25,000. 18: 1991, map series 1:50.000, Egyptian General Survey Authority. 19: 2001, Landsat satellite image, resolution 15m/pxl. 20: Evolution diagram of Maryut Lake’s surface areas (1801-2001). 21: Comparison between Maryut Lake 1801, French campaign map series 1:100,000 and 0 m level in the Mareotic region. 22: Comparison between Maryut Lake 1801-2001. Chapter 5. Lucy Blue 1: Location of Alexandria and Lake Mareotis. 2: An approximation of the ancient and present limits of Lake Mareotis. 3: Sites located along the northern shoreline, the southern shoreline and on Mareotis Island. 4: Topographic, ceramic and auger survey plan of Site 44 on Mareotis Island. 5: An example of a ‘lake wall’ feature (Site 44). 6: The Kibotos site (Site 09). 7: The ‘complex building’ (Site 13).

11 12 12 13 13 14 14 14 14 15 16 16 17 17 18 18 19 20 20 21 22 23 26 27 29 30 30 31 31

Chapter 6. Dylan Hopkinson 1: The location of Mareotis Island and the sites mentioned in text, modified from De Cosson 1935. 2: Schematic of the main features of Mareotis Island. 3: The littoral features mentioned in this paper and topographic survey of the eastern settlement of the island. 4: The littoral features mentioned in this paper and topographic survey of the western settlement of the island. 5: Longitudinal and transverse schematic sections through the ‘water tower’ on Mareotis Island.

42

Chapter 7. Mona Haggag 1: Kibotos harbour structure on the south shore of Lake Mareotis (Site 9 –L ake Mareotis Research Project). 2: Kibotos harbour structure, mooring ring. 3: Rock-cut Kibotos harbour structure at Ikingi Mariut. 4: Byzantine harbour of Marea. 5: Plan of the eastern harbour of Marea, after El-Fakharani 1977. 6: Decumanus between the eastern and middle ports of Marea. 7: Plan of shops off the main decumanus. 8: Basilica, Marea.

47 48 48 50 51 51 51 52

ii

35 36 38 40

9: Enigma Buliding, Marea. 10: Winery excavated by El-Fakharani 1983.

52 53

Chapter 8. Valérie Pichot 1: Map of Egypt: location of Lake Mariout and the site of Marea. 2: General plan of the site identified as Marea. 3: The peninsula of Marea. 4: A) pier to the north of Marea peninsula, view from the north-west. B) ancient causeway connecting the peninsula to the mainland, view from the west. 5: Ancient causeway under excavation of Sector 1, view from the east. 6: Ancient causeway under excavation of Sector 2, view from the north. 7: Landmark and jetty to the south of the peninsula, view from the south-west. 8: Results of the geophysical survey undertaken on the Marea peninsula. 9: Sector 4 - Hearths in FS4024, under excavation. 10: Overall plan of the western zone of Sector 3. 11: Overall plan of the eastern zone of Sector 3. 12: Eastern zone of Sector 3 under excavation. 13: Sector 3 - One of the two limestone sphinxes discovered in the tower building ST300. 14: Sector 3 - Bronze candelabra after restoration in the CEAlex laboratory. 15: Sector 3 - Hoe chalk burner. Chapter 9. Mieczyslaw D. Rodziewicz 1: Central part of the plan of Mareotis drawn by Mül ler-Wiener in 1966, with the pilgrimage road from the lake harbour to Abu Mina and settlements, among them settlement “M” in Huwaryia village. 2: Abu Mina with surrounding settlements/karms, and pilgrimage road towards the lake harbour. 3: Byzantine House/Hospitium at the pilgrimage road from the lake harbour to Abu Mina. 4: Byzantine pilgrimage harbour/Philoxenite (Nos. 5-17) and western/coastal rural remains (Nos. 1-4). On the eastern side, structured causeway (No. 18). Chapter 10. Krzysztof Babraj & Hanna Szymańska 1. Remains of structures along the southern shore of Lake Mareotis. 2. Marea. Byzantine baths. 3. Marea. Baths and western courtyard. View from the south-west. 4. Marea. Well operated with sāqiyah. View from the south. 5. Marea. Funerary chapel. View from the west. 6. Marea. Plan of the basilica. 7. Marea. Apse of the basilica. View from the west. ‘A’ and ‘b’ locate the two burials. 8. Marea. Basilica. Amphora kiln. View from the west. Chapter 11. Marie-Françoise Boussac & Mourad El Amouri 1. General map of the site and location of areas under study at Taposiris Magna and Plinthine. 2. Drawing by Pascal Coste, made in 1820. 3. Map of the remains of the ancient city of Taposiris by A. De Cosson, 1935. 4. Location of the different sectors of the site of Taposiris that were studied along the lake shore. 5. View of the Taposiris bridge from north. 6. Plan and elevation of the north pile of the bridge, November 2004. 7. Model and reconstruction of the bridge in Taposiris Magna. 8. Overall plan and detailed outline of the eastern jetty. 9. The southern end of the eastern jetty. 10. Jetty at Gamal. View from north-west. 11. Jetty at Quseir. View from south-east. 12. Plan and elevation of the inner south face of the north flush and buttress. 13. View of the north flush during excavation, from east. 14. Lay out, elevations and sections of north and south walls of the water supply channel through the eastern jetty. 15. Water supply channel through the eastern jetty opening into the hoop-shaped basin. 16. Grooves and notches in the water supply chanel. 17. View of the hoop-shaped basin, the water supply channel and the fish-tank. iii

57 58 59 60 60 60 60 61 62 63 64 65 65 65 66 68 68 70 71

76 78 79 80 80 81 82 82 88 89 89 90 92 92 92 93 93 94 94 94 95 97 98 98 99

LAKE MAREOTIS: RECONSTRUCTING THE PAST 18. 19. 20. 21.

The pavement in the hoop-shaped basin. The opening between the hoop-shaped basin and the fish-tank. Main types of ceramics found in Taposiris Magna (Sector 9) during the Late Roman era. Ceramics from the filling of the hoop-shaped basin.

Chapter 12. Marianne Bergmann, Michael Heinzelmann & Archer Martin 1: Alexandria and its hinterland in 1925, Survey of Egypt, sheet 4. 2: Schedia. Topographical map based on quickbird satellite image (2005) indicating excavated areas (A) and corings (H). 3: Schedia. Map with reconstruction of the ancient topography. 4. Schedia. Plan of the southeastern settlement (Kom el Giza, Kom el Hamam) with excavation areas 1980-92 and 2003-06. 5: Kom el Giza, Area 3. Plan of Hellenistic baths. 6: Kom el Giza, Area 1. Plan of Roman villa, tombs and later basins. 7: Kom el Hamam. Plan with excavations areas 2003-2006. German Mission at Kom el Giza/Beheira. 8: Kom el Hamam. Plan of Area 6 with Roman storage building and later phases. 9: Kom el Hamam. Reconstruction of the Roman storage building. 10: Kom el Hamam, Area 9. Roman enclosure wall. 11: Kom el Hamam. Plan of Area 8 with basins for wine-production.

99 99 101 102 108 109 109 110 111 111 112 112 113 114 115

Chapter 13. Penelope Wilson 1: Ancient sites in the Delta, highlighting the Mareotis region. 2: Map of the sites surveyed, with 0 metre sea-level contour area highlighted in grey. 3: The main mound at Kom el-Mahar. 4: Late Ptolemaic to Early Roman pottery sample from Kom el-Mahar. 5: Red brick structures at Kom Trugi, excavated below foundation level. 6: Section through the cemetery mound at Kom el-Qadi.

119 120 121 122 123 123

Chapter 14. Dorota Dzierzbicka 1: A map of the Mareotic region with marked locations of wineries. 2: A schematic drawing of a Type 4 winery with names of its elements used in Greek papyri from Egypt.

127 130

Chapter 15. Emad Khalil 1: The approximate ancient and present limits of Lake Mareotis. 2: The Lake Mareotis Research Project survey area along the shores of the western Mareotic Arm. 3: Taposiris Magna tower. 4: The middle quay of the harbour of Marea. 5: The quay of Marea harbour was made of large regular limestone blocks without evidence of the use of mortar. 6: The box-shaped harbour on the southern shore of Lake Mareotis. 7: One of the quays extending into the lake at the northern shore of Mareotis western arm. 8: A lake wall extending parallel to the southern shore of the lake. Evidence for red mortar (opus signinum) can still be seen between the blocks. 9: Some of the multi-room square buildings located at Site 118 along the northern shore of the Mareotis Island. The buildings could have been used for storage purposes. Chapter 16. Sameh Ramses & Ahmed Omar 1: General plan from Alexandria to Mareotis. 2: Map of sites investigated during the Lake Mareotis Research Project. 3: Two examples of fish farm deformation along the southern shore of the lake. 4: Lime storing and collecting at Site 44. 5: Lake wall affected at Site 44. 6: Site 214: the partially excavated winery. 7: Site 214: part of the fragmented paved floor of a Roman period waterfront building. 8: Kibotos harbour extending from the southern shore of the lake. 9: One of the university properties on the land surrounding the Kibotos. 10: Archeological tell divided for private possession. 11: Private property sign and wall established on Site 201. 12: Flattened Hellenistic tomb at Site 202 during bulldozer work in 2007. 13: Two soccer fields erected at the tell boundaries around Site 204. iv

135 136 137 138 138 139 140 140 141

147 148 149 149 149 150 150 151 151 151 151 151 152

Tables Chapter 4. Ismaeel Awad 1: Comparison table between different periods of Maryut Lake.

21

Chapter 13. Penelope Wilson 1: Summary table of sites surveyed by the ‘Sais and its Hinterland Project’ in the region of southern Lake Mareotis.

124

Chapter 14. Dorota Dzierzbicka 1: Installations forming part of a wine-making complex mentioned in Greek papyri containing references to immovable property found in vineyards or to the appurtenances of a winery.

131

Maps Chapter 7. Mona Haggag 1: Marea in Lake Mareotis. 2: Marea’s public buildings on the lake shore.

48 50

v

LAKE MAREOTIS: RECONSTRUCTING THE PAST

vi

Introduction The following collection of papers represents the final synthesis of a conference entitled The International Conference on the Archaeology of the Mareotic Region. Lake Mareotis: Reconstructing the Past hosted by the University of Alexandria, Egypt between 5th & 6th April 2008. The objective of the conference was to bring together scholars that work in the Lake Mareotis region of Alexandria to a forum where they could present and discuss their ongoing research and field projects. The two-day conference was a great success as it provided an opportunity to share data and tackle themes and issues of common concern. It was also an opportunity for the Lake Mareotis Research Project, a collaboration between the Centre for Maritime Archaeology, at the University of Southampton, and the Department of Underwater Antiquities (DUA) of the Supreme Council of Antiquities (SCA), Alexandria, in conjunction with the Centre for Maritime Archaeology and Underwater Cultural Heritage, University of Alexandria, to show case the Lake Mareotis survey project and present recent results to colleagues. Lake Mareotis is a shallow body of brackish water currently about 90 km2 in area that extends south of Alexandria for about 12 km and to the west in the form of an arm some 40 km and is 3 km wide. Lake Mareotis was much larger in antiquity having been subject to siltation and reclamation over the last two thousand years. Due to its location and characteristics, Lake Mareotis it is known to have contributed significantly to the ancient economy of Alexandria and to the internal and international transport system of Greco-Roman Egypt. The significance of this region has been acknowledged since at least the 1930’s (De Cosson 1935), and subsequently highlighted by the pioneering work of scholars such as ElFakharani (1983). This volume is published in honour of these early pioneers, in particular El-Fakharani who not only initiated scholarly investigation of a number of sites in the Lake Mareotis region but who also engaged for the first time, future Egyptian scholars by including and encouraging students of Archaeology from the University of Alexandria to participate in excavations at Marea (see Haggag this volume). Thus, the conference and this publication attempts to follow in the footsteps of El-Fakharani by seeking to encourage an inclusive dialogue between all scholars currently active in the region. The following papers represent the most recent attempt to contextualise and interpret Lake Mareotis. Current research, not only that generated as a product of the Lake Mareotis Research Project (Blue & Khalil, forthcoming), but also exciting new discoveries from other sites around the shores of the lake, are beginning to provide a more comprehensive understanding of the breadth and scale of activities conducted around the shores of Lake Mareotis and highlight the substantial contribution the region made to the ancient economy of Alexandria. There is still much to be discovered but with continued dialogue we can begin to tackle unanswered questions and resolve sometimes contradictory, interpretations. This volume therefore presents a collection of thoughts and perspectives that are often variable in their focus and approach. A degree of editorial consistency has been employed but in order to retain the particular and diverse perspectives of the individual authors, specific nomenclature has been maintained. To that end, a variety of spellings for Lake Mareotis (Maryut, Mariout) and sites in the Mareotic region such as Marea, Mareia etc., are supported. It is hoped that this volume will encourage future discussion and publications that continue to highlight the importance of the Lake Mareotis region in antiquity. Lucy Blue Acknowledgements The editor would like to acknowledge both the British Academy and the Leverhulme Trust who provided financial support to undertake the Lake Mareotis Research Project of which this volume is just one product. In addition, the support of Dr Emad Khalil and the University of Alexandria is acknowledged for hosting and organising the conference. I would also like to thank the anonymous reviewer of this volume for their helpful comments. Bibliography El-Fakharani, F.A., 1983, Recent Excavation at Marea in Egypt. In G. Grimm, H. Heinen & E. Winter (eds.), Aegyptiaca Trevernsia II, Das Röm isch-Byzantinische Ägypten. Akten des internationalen Symposions 26.- 30. September 1978i n Trier, 175-186. Mainz. Blue, L., & Khalil, E. (eds.), forthcoming, A multidisciplinary approach to Alexandria’s economic past: The Mareotis case study. Southampton Monograph Series. Oxford De Cosson, A., 1935, Mareotis: Being an Account of the History and Ancient Monuments of the North-West Desert of Egypt and Lake Mareotis. London vii

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Contributors Marie-Françoise Boussac Professor in Greek History Paris Ouest Nanterre University 92000 Nanterre, France [email protected]

Mostafa El Abbadi Professor Emeritus of Greco-Roman Studies Faculty of Arts University of Alexandria Hussein Sobhi St. El-Shatbi 21526 Alexandria, Egypt [email protected]

Delphine Dixn euf Institut Français d’Archéologie Orientale 37, rue al-Cheikh Aly Youssef B.P. Qasr al-Ayni 11562 11441 Cairo, Egypt [email protected]

Mohamed S. Abd-el-Ghani Professor of Ancient History Dept. of Archaeology and Greco-Roman Stiudies Faculty of Arts University of Alexandria Hussein Sobhi St. El-Shatbi 21526 Alexandria, Egypt

Dorota Dzierzbicka Department of Papyrology Institute of Archaeology University of Warsaw Krakowskie Przedmieście 26/28 00-927 Warsaw 64 Poland [email protected]

Mourad El Amouri Ipso Facto Bureau d’Étude et de Recherche, Archéologie et Océanographie 4 rue de Tilsit 13006 Marseille, France [email protected]

Thomas Faucher ANR-Nomisma Université Paris-Sorbonne-Paris IV 1, rue Chevalier de la Barre 75018 Paris, France [email protected]

Ismaeel Awad Topography Department Centre d’Études Alexandrines 50, rue Soliman Yousri 21131 Alexandria, Egpyt ismaeel_a [email protected]

Mona Haggag Bibliotheca Alexandrina P.O. Box 138 El-Shatbi 21526 Alexandria, Egypt [email protected]

Krzysztof Babraj Archaeological Museum in Kraków ul. Senacka 3 31-002 Kraków , Poland [email protected]

Michael Heinzelmann University of Cologne Institute of Archaeology Albertus Magnus Platz D-50923 Cologne [email protected]

Marianne Bergmann Georg-August-University Goettingen Archaeological Institute Nikolausberger Weg 15 D-37073 Goettingen [email protected]

Dylan Hopkinson Centre for Maritime Archaeology University of Southampton Highfield Southampton, England SO17 1BJ [email protected]

Lucy Blue Director, Centre for Maritime Archaeology University of Southampton Highfield Southampton, England SO17 1BJ [email protected]

viii

Sameh Ramses Supervisor of Egyptian Excavations & Surveying Team Department of Underwater Archaeology Supreme Council of Antiquities (SCA) 19 Abdel Moety el Gazaly, Stanley Alexandria, Egypt [email protected]

Emad Khalil Centre for Maritime Archaeology & Underwater Cultural Heritage Faculty of Arts University of Alexandria Hussein Sobhi St. El-Shatbi 21526 Alexandria, Egypt [email protected]

Mieczyslaw D.R odziewicz Ul. Literacka 25 m.10 01-864 Warsaw, Poland [email protected]

Archer Martin American Academy in Rome Via Angelo Masina, 5 I-00153 Roma [email protected]

Hanna Szymańska Research Centre for Mediterranean Archaeology University of Warsaw 11, Sharia Mahalla, Heliopolis Cairo, Egypt [email protected]

Ahmed Omar Inspector, Department of Underwater Archaeology Supreme Council of Antiquities (SCA) 19 Abdel Moety el Gazaly, Stanley Alexandria, Egypt

Penelope Wilson Department of Archaeology Durham University South Road Durham, England DH1 3LE [email protected]

Valérie Pichot Centre d’Études Alexandrines 50, rue Soliman Yousri 21131 Alexandria, Egpyt [email protected]

ix

LAKE MAREOTIS: RECONSTRUCTING THE PAST

x

Fawzi El-Fakharani: Pioneer Excavator at Mareotis Mona Haggag

This volume is dedicated to the memory of Fawzi El-Fakharani who was, in many ways, a pioneer investigator of the area of Mareotis.

certain or clear that it was a temple dedicated to Hercules (Kanellopoulos 1994: 60-65). In 1970 Fakharani was delegated to Libya, where he established a department of Classical Archaeology. In Libya, Fakharani excavated the area of the Byzantine palace of the City of Tocra, ancient Taucheira. The patriarch, who presented the city at the council of Nicea in 325 CE, must have resided in this palace. Fakharani uncovered parts of the Eastern Church to which the palace was attached. He also unearthed the church s confirmatorium and baptistery, as well as two beautiful mosaic panels (pers. comm.).

Fawzi El-Fakharani was born in Alexandria in 1921. He graduated from the Faculty of Arts, University of Alexandria in 1946 and directly after his graduation he went to the United Kingdom for higher studies. In 1953 he obtained a Masters degree from the University of Liverpool. His undergraduate dissertation was entitled “The origins of stucco reliefs in Classical Art”. Afterwards he moved to University College London where he acquired a doctoral degree in 1957 on “Stucco Reliefs in Roman Art”.

In 1973, back in Alexandria, Fakharani began his studies of the city of Marea, thus realising his life’s dream, which was to discover and document the ruins of this city. After a long and exhaustive period of preparation, he finally obtained the permission to excavate the site identified previously as the city of Marea, to the south of the western extremity of Lake Mareotis, some 45 km distance from Alexandria. With the exception of previous identifications of the site as that of Marea made by Mahmoud Pasha ElFalaki (1872: 85, 86, 91, 96, 97, 101), and the valuable monograph on Mareotis by De Cosson (1935: 131-135), the academic community of archaeologists and researchers had paid no particular interest in the area of Mareotis. Fakharani’s determined enthusiasm and efforts to identify the city’s area proved successful, as the results of his excavations shed light on the various stages of life within the city and on the importance of the area as a whole (El-Fakharani 1983: 175-204). Some of the articles of this volume are devoted to Marea (see Haggag, Pichot, Rodziewicz, Babraj and Szymańska through which the reader can figure out that the discovery of Marea was and is, of great importance to researchers, as well as to the community at large.

Fakharani returned to Egypt and was appointed lecturer in the University of Alexandria in 1958. In 1964 he was promoted to an assistant professor, and won the chair of Professor of Classical Archaeology in 1974. He was appointed Chair of the Department of Classical Civilization between 1976 and 1980, and the director of the Institute for Mediterranean Studies between 1986 and 1989. In 1990 he became Chair of the Department of Archaeology at the University of Tanta, Egypt. In 1966 Prof. Fakharani was delegated to the University of Amman in ordan. There, he began his first endeavours in archaeological fieldwork. His excavation target was to continue uncovering the grand Roman theatre of Amman. He published the results of his investigations in the J ahrbuch Des Deutschen Archäol ogischen Instituts Archäol ogischen Anzeiger, in 1975 (El-Fakharani 1975a: 377-403). In 1967, his interests shifted towards the site of Qal’at AlJabal on which stands the so-called Temple of Hercules (El-Fakharani 1975b: 533-554). There he reinvestigated the remains and uncovered most of the building’s foundations and annexed colonnades. Scholars differed with regard to which deity the temple was dedicated, yet Fakharani’s hypothesis was that the building had characteristic features unknown to temples of the East Roman Provinces in the time of Marcus Aurelius, although common in libraries of the same epoch in the neighbouring countries (El-Fakharani 1975b: 554). Fakharani reached the conclusion that the building was not a temple but the library built by Marcus Aurelius at Philadelphia, the city that he described as Alexandria’s daughter city. Despite his reasonable arguments and study published in the periodical of the University of Rostock, the building was still termed as the “Temple of Hercules”. This remained the case until a study conducted in 1994 of an inscription carved on the architrave of the building showed that it was not at all

With only ten thousand Egyptian pounds as a budget, Fakharani was given the daunting task of locating a suitable area to undertake a bold and pioneering step in Egyptian archaeology, his goal being to accommodate both male and female Egyptian students in the field. This was something that had never been undertaken before and had to be managed with great care and respect for social traditions. Not only did he have to secure separate sleeping and study areas but he also had to supply electricity to the camp, food and sustenance for the students, and transportation to the desert area of Mareotis. All of this he accomplished on a meagre budget. Besides the difficulties of obtaining permission from the university to take the students to the field, he had also to convince their parents, especially those of the female students, that their daughters would be safe and 1

LAKE MAREOTIS: RECONSTRUCTING THE PAST secure. It was not easy for Egyptian families during the 1970s to send their daughters to the desert for field training, but Fakharani managed to allay any of the families fears and worries.

German team. They all emphasised that the remains from the Ptolemaic era lay only in the area of the nursery which then belonged to the Faculty of Agriculture, the University of Alexandria. Again, it took determination and patience to acquire the permission to excavate this site. During this time, Fakharani excavated a few trial trenches in the area around the remains of an alabaster tomb. Although the trenches yielded nothing of conspicuous importance, a Byzantine well just next to the tomb shed new light on the use of this area during this period. As the level of the monolithic pavement of the early Ptolemaic tomb was higher than that of the Byzantine rim of the uncovered well, it became evident that the tomb was not in situ as Adriani had assumed (Adriani 1940: 15-23), and that it had been moved to its present location during the Byzantine or early Islamic periods (author’s interpretation).

Soon after the ancient buildings emerged from the desert sands, articles and interviews in local and international media started to appear. These articles helped win the support of the families of the students who had been allowed to camp in the desert with Fakharani on this pioneering expedition. The student families soon realised from these articles and media reports just how important the work was that their young daughters and sons were accomplishing. Thus, Prof. Fakharani overcame all obstacles and all the students, including myself, who participated in the excavations, were, for the first time in the history of the department, to become the lucky generation of students who had the opportunity to be trained in the field. We are extremely grateful to Prof. Fakharani for that experience. The directors of the Lake Mareotis Research Project, Lucy Blue, Sameh Ramses and Emad Khalil (to whom this volume owes its inspiration), are now reminding us of those wonderful days by involving students in their current fieldwork surveying the shores around Lake Mareotis, the first time this has happened since the pioneering days of Fakharani.

At the same time that Prof. Fakharani received the permission for his excavation of the nursery, Parkinson’s disease had already begun digging its deadly roots into his aging body. Accordingly, he entrusted the project to another great scholar who has also made considerable contributions to the archaeology of Alexandria and its environs, namely Jean-Yves Empereur. In March 2004, Fakharani passed away but he left behind a generation of empowered, highly motivated, committed and devoted students who strive to continue their work to the standards and example set by Fawzi El-Fakharani.

Prof. Fakharani never once lost sight of the idea that Marea was the “Pompey of the East”, to use his own expression (El-Fakharani 1977: 5). It deserves more care and diligence as it represents a unique archaeological site that gives us a complete picture of the everyday life of ordinary people, in contrast to other archaeological projects in Egypt which have a tendency to focus on discovering how our once noble Pharaonic families lived and prepared for the afterlife. Marea, together with the results of survey and excavations from other settlements around the shores of the Lake (see Blue, Hopkinson, Boussac & El Amouri, Bergmann, et al., Wilson, Dzierzbicka, Khalil, and Ramses & Omar this volume), illustrate how people in a postPharaonic era lived, what their beliefs were, what their baths looked like and how they were used, the shape and layout of their shops, cisterns, and how they buried their dead. Marea necessitates more preservation, conservation, restoration and preparation efforts to emphasise the site’s potential, in order to amongst other things, place it firmly on the tourist map of Egypt.

Bibliography Adriani, A., 1940, Fouilles et decouvertes. Alexandrie. I. Tombeau en alabaster du cimetiere latin. Annuaire 3, 1935193 : 15-23. Alexandria. De Cosson, A., 1935, Mareotis: Being an Account of the History and Ancient Monuments of the North-West Desert of Egypt and Lake Mareotis. London El-Fakharani, F.A., 1975a, Das Theater von Amman in Jordanien. J ahrbuch des Deutschen Archäol ogischen Instituts Archaologischer Anzeiger 3: 377-403. El-Fakharani, F.A., 1975b, The Library of Philadelphia (?) Or The So-Called Temple on the Citadel Hill in Amman. Wissenchaftlische Z eitschrift der U niversität Rostok XXIV.6: 533-554. El-Fakharani, F. A., 1977, A Preliminary Report on the 1s t Season of Excavation at Marea. Excavation Report presented to the Antiquities Department, Egypt. El-Fakharani, F.A., 1983, Recent Excavation at Marea in Egypt. In G. Grimm, H. Heinen & E. Winter (eds.), Aegyptiaca Trevernsia II, Das Röm isch-Byzantinische Ägypten. Akten des internationalen Symposions 26.- 30.Se ptember 1978i n Trier, 175-186. Mainz. El-Falaki, Mahmoud Bey, 1872, Memoire sur l’antique Alexandrie,s es faubourgs,s es environs. Copenhagen. Kanellopoulos, C., 1994, The Great Temple of Amman, Vol. I: Architecture. American Center of O riental Research – Amman: i-xii.

inally, in 1999, akharani again took us to the field. This time akharani s field was in Alexandria itself. His investigations were of the Latin cemetery which he believed to have been the site of the royal Ptolemaic burial ground. A geophysical survey was conducted in three successive phases: firstly, by colleagues from Alexandria niversity, secondly, in collaboration with Mrs. Calliope Lemniou Pappacosta, a researcher from Greece in consultation with experts from the University of Patras, and thirdly, by a

2

M. ABD-EL-GHANI: ANCIENT SOURCES

The Mareotic Region in Ancient Sources Mohamed S. Abd-el-Ghani

and on the south Lake Marea, also called Mareotis: “This is filled by many canals from the ile, both from above and on the sides …” (Strabo 17.1.7).

Before addressing the Mareotic region in antiquity it would be appropriate to explore how Egyptian frontiers are discussed in the classical sources. There is a crucial and decisive paragraph in Strabo (17.1.5) to this effect – one that makes particular reference to the western frontiers of Egypt – the primary focus of this paper – and is quoted “verbatim” as follows:

What is particularly interesting to note about the inhabitants of the Mareotic region, and the Western Desert of Egypt in general, is the ethnic distinction that characterises them apart from the rest of the Egyptians, even after their identity as Egyptians was settled from an administrative viewpoint. It is a phenomenon that is observed throughout the ancient texts through to the present day. This contrast between the Mareotes and the Egyptians is explicitly expressed in a document dated from the middle of the 4th century AD about the smuggling of natron.1

“Now the early writers gave the name Egypt to only the part of the country that was inhabited and watered by the Nile, beginning at the region of Syene and extending to the sea but the later writers down to the present time have added on the eastern side approximately all the parts between the Nile and the Arabian Gulf (i.e. the present Red Sea), and on the western side the parts extending as far as the oases, and on the sea-coast the parts extending from the Canobic mouth to Catabathmus (the present el-Salloum) and the domain of the Cyrenaeans. For the kings after Ptolemy became so powerful that they took possession of Cyrenaea itself and even united Cyprus with Egypt. The Romans, who succeeded the Ptolemies, separated their three dominions and have kept Egypt within its formers limits.”

Marea is stated by Herodotus as being located at the key position with regard to the Egyptian frontiers to the west. He relates that king Psammetichus I of the 26th Dynasty posted garrisons at Elephantine on the side of Ethiopia, at Daphnae of Pelusium on the side of Arabia and Assyria, and at Marea on the side of Libya (Herodotus II.30). Marea was also mentioned by other classical authors as the battlefield of some pitched battles in antiquity before the conquest of Egypt by Alexander the Great. In his reference to the second Egyptian revolt against the Persian rule under King Artaxerxes (c. 460 BC), Thucydides pointed out that Inaros son of Psammctichus, “setting out from Maria, the city just north of Pharos, caused the greater part of Egypt to revolt from king Artaxerxe” (Thucydides I.104).

Among the early writers referred to above by Strabo, it was Herodotus (II.17) who adopted the ancient theory that Egypt was the land inhabited by Egyptians and was watered by the Nile. Nevertheless, he considered that the Egyptians were not only the inhabitants of the Nile Valley and the Delta proper, particularly the area that was directly watered by the Nile River, but also those places to which the Nile waters extended. In order to illustrate this, he set out to write a significant account about the inhabitants of the cities of Marea and Apis. His account, in this concern, goes as follows:

Diodorus of Sicily reported that the decisive pitched battle between King Apries, the legitimate king (588-566 BC) of the Saite Dynasty, and Amasis II, one of the king’s generals and claimant to the throne who became a renowned king, took place near the village of Maria (Diodorus Siculus I.68),2 after the former’s defeat by the Greeks of Cyrenê and Barcê .

“The men of the cities of Marea and Apis, in the part of Egypt bordering on Libya, thinking themselves to be not Egyptians but Libyans, and misliking the observance of the religious law which forbade them to eat cow s flesh, sent to Ammon saying that they had no part or lot with Egypt: for they dwelt outside the Delta and did not consent to the ways of its people, and they wished to be allowed to eat of all food. But the god forbade them: all the land, he said, watered be the Nile in its course was Egypt and all who dwelt lower than the city Elephantine and drank of that river’s water were Egyptians.” (Herodotus II.18)

The first specific and detailed account about Lake Marea and the Mareotic region was that of Strabo. After his statement that Lake Mareotis was filled by many canals from

1. P. Lond. II.231, p. 285 - Hunt and Edgar, Select papyri II. 428 = P. Abinn. I.9, ll. 4-5: ὥστε ὅσα νίτρα καταλαμβάνεις εἴτε δια Μαρεωτῶν εἴτε δια Αἰγυπτείων κατερχόμενα. 2. Cf. Herodotus II.169 where he identified the site of that battle to be at the town of Momemphis “ἐς Μώμεμφιν πόλιν” which is roughly identified as Kom Abu Billu by Ball 1942: 18 oyotte & Charvet 1997: 118, n. 249 identified it as aujourd hui K m el-Hisn in their commentary on Strabo 17.1.22

That they drank of that river’s water is well attested in Strabo s account of Alexander s first visit to the place of Alexandria, the advantages of the site, and his determination to found a city bearing his name on the site. Among the advantages of the (new) city’s site was its location between two (seas), on the north the Egyptian sea, as it is called, 3

LAKE MAREOTIS: RECONSTRUCTING THE PAST the Nile, both from above and on the sides, he goes on to complete the picture of the whole region as follows:

sites of the “Libyan coastal nome” from immediately before Catabathmus (es-Salloum) to Glaucus Promontory (El-Imâ yid) (V. 5. 4 - 7), he goes on to report the sites of the “coastal Mareotic Nome” or “Νομοῦ Μαρεώτου παράλιος”, i.e., namely, the village of “Cheimw” (El-Bordâ n) and Plinthine (ruins 6 km NE of Sidi Kireir) (Ptolemaios IV.5.8). Finally, he comes further east to Chersonesos the Smaller (El-Dekheila) and then to Alexandria, the capital of the whole of Egypt, to quote Ptolemy (Ptolemaios IV.5.9).

“...and through these canals the imports are much larger than those from the sea, so that the harbour on the lake was in fact richer than that on the sea, and here the exports from Alexandria also are larger than the imports and anyone might judge, if he were at either Alexandria or Dicaearchia (Puteoli in Italy) and saw the merchant vessels both at their arrival and at their departure, how much heavier or lighter they sailed thither or therefrom. And in addition to the great value of the things brought down from both directions, both into the harbour on the sea and into that on the lake, the salubrity of the air is also worthy of remark. And this likewise results from the fact that the land is washed by water on both sides and because of the timeliness of the Nile’s risings … at Alexandria, at the beginning of the summer, the ile, being full, fills the lake also and leaves no marshy matter to corrupt the rising vapours. At that time, also, the Etesian winds blow from the north and from a vast sea, so that the Alexandrians pass their time most pleasantly in summer.” (Strabo 17.1.7.)

In another passage of Ptolemy (13-32) he gives a list of the villages inland of the Lybian ome from west to east the last village to the east being the one called “Mareotis”. In another passage he provides another list of the “cities and villages of the Mareotic Nome” (the part inland), further to the east on the confines of the previous Lybian ome: the farthest village to the east being that called the “village of ancient Marea” or Παλαιμάρεια κώμη (Ptolemaios IV.5.34). According to Ptolemy’s calculations the village of Mareotis, the farthest point to the east of the Libyan Nome, was situated on longitude 58 (27º .25’) and latitude 28º .20’, however, the village of ancient Marea is located at longitude 60 (29º .25’) and latitude 30º .10’. Thus, the village of “Mareotis” was situated to the south-west of the village of “Ancient Marea”, at a location yet to be identified with a specific ancient site. Ball 1942: 114 could not identify the present location of either site.

From the above paragraph of Strabo the advantages of Lake Mareotis with its vivid harbour, as well as that of the Mareotic region in general, is quite obvious: intensive commercial activities of imports and exports from the interior of Egypt as well as Mediterranean cargoes an excellent and healthy climate which was also a tourist attraction in itself. Another passage Strabo (17.1.14) sheds more light on the size of the lake and the other activities of the inhabitants of its surrounding shores. He describes the situation as follows:

Some of these aspects referred to in the works of the classical authors concerning the Mareotic region are reflected in papyrological documents, scarce though they are. In the light of such documents the region of Mareotis constituted an administrative “nome” during the 1st (OGIS 669 = S.B. V.8444 AD 68 , l. 48 P. Iand. IV.53, col. 3, l. 1 AD 88 , 2nd P. Ryl. II.78, l.8 AD 157 P. slo III.84 AD 138161 ll. 12-14 P. Haun. II.22, l. 3. and 3rd (P. Flor. III.338, l.6 BG I.13, AD 289, l.2. centuries AD.

“Lake Mareia, which extends even as far as this [ the Chersonesus/ Marabit according to oyotte & Charvet 1997 or el-’Dekheila’ according to Ball 1942] , has a breadth of more than one hundred and fifty stadia and a length of less than three hundred. It contains eight islands and all the shores round it are well inhabited and the vintages in this region are so good that the Mareotic wine is racked off with a view to ageing it.”

About a century after Strabo’s visit to Egypt, Pliny the Elder in his Historia Naturalis reported about Lake Mareotis as follows:

The second important aspect highlighted in these few documents is the close relationship between the Mareotic Nome and Alexandria. Some individuals in these documents would come to Alexandria, through the Mareotic ome, for business or official or judicial purposes. This is clear in some private or business letters P. Haun. II.22 P. Flor. III.338). Sometimes they hint at some degree of co-ordination between the high officials of Mareotis and Alexandria when necessity requires. In a poorly preserved document, the essence of which could be understood in its general lines owing to the existence of several key words, two persons domiciled in the Mareotis region are mentioned in a petition to the strategos of Alexandria regarding the theft of two camels belonging to them. According to the publisher s reading and filling of the lacunae, which seems almost plausible, they ask the Alexandrian strategos to write to his Mareotic counterpart to summon some suspected person to present himself before him for investigation (P. Oslo III.84). Why was the Alexandrian strategos asked to interfere instead of resorting directly to the Mareotic strategos? Does this further illustrate a degree of collaboration between the two strategoi?

“Lake Mareotis, which lies on the south side of the city, carries traffic from the interior by means of a canal from the Canopic mouth of the ile it also includes a considerable number of islands being 30 miles across and 250 miles in circumference, according to Claudius Caesar.” (Pliny V.XI.63)

Strabo and Pliny also note the variations in measurement of the lake particularly during the inundation season in the summer when the ile flooded. rom their accounts it is clear that the lake was thus much more extensive in antiquity than today oyotte & Charvet 1997: 104, n. 179 Strabo 17.1.14). Last but not least among the classical works concerning the description of the Mareotic region, is the Geography of Claudius Ptolemaios in the 2nd century AD. In his fourth book (Chapter 5 about Egypt) Ptolemy begins his description of the geography of Egypt from west to east stating the longitude and latitude of each given site. After covering the 4

M. ABD-EL-GHANI: ANCIENT SOURCES Bibliography

The theft of two camels in this document also leads us to comment on the repeated reference to camels in the Mareotic documents. This is not unexpected from an area situated on the edge of the Western Desert where the camel was the chief means of traffic and transport. In one of the documents relating to business affairs, a sum of 900 drachmae is reported as a rent for some camels (P. Haun. II.22, ll. 9-10 and note on l.9.). In another document dated AD 289, two Mareotes sell an Arabian camel belonging to them to one of the chief officers primpilaris of the Prefect of Egypt for 16½ talents of silver (BGU I.13, II. 1-6).

Papyri BGU = Aegyptische U rkunden aus K öni glichen ( later Staatlichen) Museen zu Berlin, G riechische U rkunden (13). H. Satzinger (ed.), 1904. Berlin. P. Abinn. = The Abinnaeus Archive: Papers of a Roman Officer in the reign of Constantius II (1-9). H.I. Bell, V. Martin, E.G. Turner & D. Van Berchem eds. , 1962. Oxford. P. Fay. = Fayum Towns and their Papyri (134). B.P. Grenfell, A.S. Hunt, D.G. Hogarth & .G. Milne eds. , 1900. London. P. Flor. = Papiri greco-egizii, Papiri Fiorentini III. Documenti e testi letterari dell‘ età romana e bizantina (III, 338). G. Vitelli (ed.), 1915. Milan. P. Fouad = Les Papyrus Fouad I (11). A. Bataille (ed.), 1939. Cairo. P. Hamb. = G riechische Papyrusurkunden der Hamburger staats- und U niversität sbibliothek (I, 39). P.M. Meyer (ed.), 1924. Leipzig. P. Haun. = Papyri G raecae Haunienses (II, 53). Letters and mummy labels from Roman Egypt. A. B low- acobsen. transl. & ed. , 1981. Bonn. P. Iand. = Papyri Iandanae IV, 53 . K. Kalbfleisch, E.H. Schaefer, G. Rosenberger, D. Curschmann, L. Eisner, . Sprey, G. Spiess, L. Spohr & . Hummel eds. , 1912. Leipzig. P. Lond. = G reek Papyri in the British Museum (II, 231). F. Kenyon & H. Bell eds. , 1973. Milan. P. Oslo = Papyri O sloenses III, 84 . S. Eitrem & L. Amundsen (eds.), 1925. Oslo. P. Ryl. = Catalogue of the Greek and Latin Papyri in the John Rylands Library. Manchester (II, 98, 227). C.H. Roberts & E.G. Turner eds. , 1952. Manchester. S.B. = Sammelbuch griechischer U rkunden aus Aegypten V, 8444 . . Bilabel & E. Kiessling eds. , 1952. Wiesbaden.

Some of the Mareotic products are also mentioned in the few available documents. In spite of the excellence and distinction of the Mareotic vineyards and wine, there appears to be limited documentary sources that comment on their merits. A single document, a private letter dating to the 2nd century AD, in which the sender tells his friend and addressee that he dispatched him a variety of gifts, among which we find: ... and a number of bunches of grapes, - 6 small clusters: 2 of them white, 2 of the ‘royal’ variety, the Mareotic as it is called, and likewise 2 smoke coloured, - in a small basket through Harklides” (P. Fouad, 77, ll. 15-20, l. 17). In other documents dating to the 3rd century AD, there are occasional references to “strips of linen”, ταινίδιον: some of them white, some “Mareotic” (P. Ryl. II. 227, ll. 25-33). If we notice that the use of ταινίδιον is a diminutive of ταινεία which means “a headband worn especially as a sign of victory and if we bear in mind that a strip of land near Lake Mareotis was named ταινεία, mentioned in Athenaeus (I.33e) and C. Ptolemaios (4.5.14), could one infer some connection between this place and the production of “linen headbands” whence “it might have derived its name? There is another mention of “ a glassware ... of a fine Mareotic quality” from the 4th century AD (P. Fay. 134, ll. 4, 6). Another 4th century document makes a clear reference to the illegal smuggling of natron by Mareotes or other Egyptians across the desert. In this correspondence between two high officials in charge of the natron monopoly (AD 342-51), strict orders are issued to detain the camel drivers and their beasts, whether in Arsinoe or elsewhere, in order to safeguard the natron of the Fiscus (P. Abinn. I.9 = P. Lond. II.231).

O ther Sources Athenaeus, Deipnosophists, C. Burton (transl.), 1927. London. Ball, ., 1942, Egypt in the Classical Geographers. Cairo Diodorus Siculus, Diodorus Siculus. C.H. Oldfather (transl.), 1935. Cambridge, MA. Herodotus, The Histories. A. de Sélincourt (transl.), 1971. Middlesex. Pliny the Elder, Natural Histories. . Bostock & H.T. Riley (eds.), Perseus Digital Library: http://old.perseus.tufts.edu/cgi-bin/ptext?doc= Perseus:text:1999.02.0138:toc accessed anuary, 2010 . Ptolemaios (Claudius Ptolemy), The G eography. E.L. Ste venson transl. & ed. , 1932. ew ork. Strabo, G eography. H.L. ones transl. , 2001. London. Thucydides, The Peloponnesian War. .M. Dent transl. , 1910. ew ork, London. oyotte, ., & Charvet, P., 1997, Strabon, Le voyage en Egypt. Paris.

Finally, there are some references to army troops stationed in the Mareotic region, together with the payments and extra payments to them, as well as the divisions of the army to which they were assigned (P. Iand. IV. 53 (AD 88), col. 3 P. Hamb. I.39 AD 179 63 BB . Although the image of the Mareotic region, especially in the papyri, is somewhat fragmentary, mosaic-like, and far from complete, it does shed, however, some light on the region. It is clear that the material evidence of future archaeological excavations will add a great deal to our knowledge of this region.

5

LAKE MAREOTIS: RECONSTRUCTING THE PAST

6

M. EL ABBADI: LAKE MAREOTIS IN BYZANTINE TIMES

A Note on Lake Mareotis in Byzantine Times Mostafa El Abbadi

This paper will present readings of selected passages in the ancient sources that have a bearing on Lake Mareotis and its region. As will be seen, the name Mareotis was often exchanged for Mareia, the name of its capital city and harbour, which in Pharaonic times served also as a garrison city protecting the north-western approaches to Egypt, as briefly stated in Herodotus II.18 & 30 and Thucydides I.104.1 .

Emperor ustinian built many constructions and in particular the residences of the magistrates and baths.”

Of special interest is his use of the Greek form Taphosiris, and not the Latin corruption Taposiris, which was used by Strabo. Also the construction work of ustinian reveals the continued importance of the town more than five centuries after the time of Strabo, despite the total change of religion from paganism to Christianity.

The first detailed description of the Lake Mareotis region is found in Strabo 17.1.71 in his well known description of Alexandria where reference is also made to the southern region of Lake Mareotis. In a further significant passage, Strabo 17.1.15 comments on an important aspect of the economic value of the lake:

As regards the southern confines of Alexandria, Procopius reports VI.1. 1-5 : The ile River does not flow all the way to Alexandria, but after flowing to the town which is named Chaereou, it then turns to the left, leaving aside the confines of Alexandria. Consequently the men of former times, in order that the city might not be entirely cut off from the river, dug a very deep canal from Chaereou and thus by means of a short branch made the river accessible to it. There also, as it chances, are the mouth of certain streams flowing in from Lake Maria.

The byblus i.e. papyrus plant grows in the Egyptian marshes and lakes, as also the Egyptian cyamus believed to be the lotus plant Germer 1985: 39 ff. from which was made the ciborium a kind of vessel for drinking it has stalks approximately equal in height, about 10 feet. But though the byblus is a bare stalk with a tuft on top, the cyamus produces leaves and flowers in many parts, and also a fruit like our cyamus, differing only in size and taste. Accordingly the cyamus-fields afford a pleasing sight, and also enjoyment to those who wish to hold feasts therein. They hold feasts in cabin-boats, in which they enter the thick of the cyami and the shade of the leaves for the leaves are so very large that they are used both for drinking-cups and for bowls, for these even have a kind of concavity suited to this purpose and in fact Alexandria is full of these in the work-shops, where they are used as vessels and the farms have also this as one source of their revenues, I mean the revenue from the leaves.”

In this canal, it is by no means possible for large vessels to sail, so at Chaereou they transfer the Egyptian grain to boats lemboi which they are wont to call diaremata, and thus convey it to the city, which they are enabled to reach by way of the canal-route and they deposit it in the quarter of the city, which the Alexandrians call Phialè . But since it often came about that the grain was destroyed in that place by the people rising in sedition, the Emperor ustinian surrounded this district with a wall … .”

In this passage, the following points are of special interest: a The freshwater canal from Chaereou received additional supplies from certain streams flowing in from Lake Maria”. It is worthy of note that Procopius states that the freshwater canal to Alexandria branches off at a location he called Chaereou, whereas Strabo uses the name of Schedia. Surprisingly, other sources of the 6th century and later, use the name of Chaereou only. Significantly we can mention the Mosaic Map of Madaba which is also believed to date from the 6th century Avi- onah 1954: 16-18 . Another source is Stephanus Byzantius 677:18 also in the 6th century, who speaks of Chaereou as a city in Egypt. To all appearances, Schedia and Chaereou were two different locations close to one another, Schedia north of the canal and located on the Canopic Branch, whereas Chaereou is situated immediately south of the canal Haas 1997: 25, top of map 2 and seems to have gained in importance at the time of ustinian when Stephanus called it a polis.

Furthermore, in connection with feasting, Strabo makes mention of Taposiris at the western end of Lake Mareotis and adds 17.1.14 that there they also: hold a great public festival …. and near it there is a rocky place on the sea where likewise crowds of people in the prime of life assemble during every season of the year.” With this classical description in mind, we move on to some Byzantine sources in order to identify any similarities or differences in the descriptions attributed to the lake in later periods. An interesting passage is to be found in Procopius work on the Buildings of ustinian in the 6th century VI.1.12 . It deals with the regions to the west and south of Alexandria. He states that to the west: In that territory, is a city one day s journey distant from Alexandria, Taphosiris by name, where they say the god of the Egyptians, Osiris, was buried. In this city the

7

LAKE MAREOTIS: REC

STR CTI G THE PAST

b In that canal the large ile vessels could not sail so their cargo of grain had to be transferred to diaremata light transport boats . This word was known only in Egypt hence Procopius explained it by the Greek word lembos”. The same word with the change of a vowel dieremata” has been found in papyri cf. ertel 1917: 130 . These dieremata were probably also used to navigate across Lake Mareotis.

The above discussed passages provide evidence that with the founding of Alexandria, Lake Mareotis acquired a lively and vital significance that continued into late antiquity. With its variety of flowering plants, the lake provided a pleasurable recreation resort for the Alexandrians who were, according to Strabo, in the habit of holding feasts in cabin boats. Even as far west as Taposiris, Strabo adds, a great public festival was held, obviously in connection with the cult of siris. More than five centuries later, Procopius reports that ustinian showed interest in that town Taphosiris by adding new construction work which is an indication of its continued importance.

c The grain was eventually carried by light transport boats and deposited in the quarter of Alexandria called Phialè . This Phialè is not attested anywhere else and we have the impression that it was close to a landing-place near the canal. However, Procopius speaks of it as already existing and how it used to be plundered in times of trouble hence ustinian surrounded it with a wall for protection.

It is of interest to know that the freshwater canal that links to Alexandria received additional supplies from certain streams flowing in from Lake Maria . This passing mention by Procopius provides additional evidence regarding the intricate hydraulic system of the freshwater canal. Furthermore, this canal provided the final facility for transferring grain from large Nile vessels – which could not sail in the canal - to light transport boats dieremata .

inally, we come to a text by Sophronius early 7th century in his account of the Miracles of Saints Cyrus and J ohannes. The miracle happened to a certain Christodorus, deacon of the church of the two saints at the time of ohn the Almsgiver, then Bishop of Alexandria 610-619 , whom Sophronius describes as the beloved of the poor . In his account, Sophronius has this to say miracle no.8 :

Finally, the lake and the regions surrounding it were of vital economic value to Alexandria, since the waters of the lake teemed with fish and water-fowl and the surrounding fertile lands flourished with large plantations until the early 7th century. Accordingly, Alexandria depended upon this neighbouring hinterland for the regular supply of its daily needs of meat and vegetables.

Lake Mareia: ou Alexandrians beloved of Christ, you all without exception know it well i.e. the miracle , so also do a good number of Egyptians and most of the Libyans and as well as you, who are curious to know about Alexandria. Christodorus once sailed across Mareotis with the intention of inspecting the domains belonging to the church of the saints. A storm suddenly blew violently the time was winter , agitating the waters of the lake and raising great waves. Christodorus was in great danger and thought he was going to perish. Nevertheless, he prayed to the saints who listened to his prayer … . And the wind at once dropped and he was saved.”

Bibliography Avi- ona, M., 1954, The Madaba Mosaic Map. erusalem. Calderini, A., 1935-1987, Dizionario. Cairo, Madrid. C.Theod. = Codex Theodosianus, T. Mommsen & P.M. Meyer eds. , 1904-5, 2 Vols. Berlin. Donner, H., 1992, The Mosaic Map of Madaba. Kampen. http://198.62.75.1/www1/ofm/mad/legends/legends156.html. Accessed ctober 2009. Courtesy Prof. M. Haggag . Gascou, . transl. & ed. , 2006, Sophrone de J érusalem, Miracles de Saints Cyr et J ean. Paris. Germer, R., 1985, Flora des Pharaoischen Aegypten, DAI. Sonderschrift 14. Kairo with thanks to Prof. Mirvat Seif . Haas, C., 1997, Alexandria in Late Antiquity. Baltimore, London. Herodotus, Historiae, C. Hude transl. , 1966. xford. Leontius eapoleos, Life of ohn the Almsgiver. In E. Dawes & .R. Baynes transl. & eds. , Three Byzantine Saints: Contemporary Biographies of St. Daniel the Stylite, St. Theodore of Sykeon, and St. J ohn the Almsgiver 1948. xford. ertel, ., 1917, Liturgie. Leipzig. Palladius, The Lausiac History. C. Butler transl. , 1904. Cambridge. Palladius, Historia Lausiaca. Migne, Patrologia Graecia 34. 1860, reprint 1998. Gottingen.

For our purpose however, the main point of interest in this passage is the mention of the domains of the church . In the Byzantine period, just as important temples of earlier times disposed of great wealth and property, so now the Church of Alexandria possessed considerable property. In 415 Emperor Theodosius, in a law on patronage, guaranteed the property of the Church of Alexandria which had been acquired before 397 C.Theod. I.23.6 . The material wealth of the Church of Alexandria is also asserted by the fact that it owned a merchant fleet in the Mediterranean. Early in the 7th century the same ohn the Almsgiver, used that fleet to import corn from Sicily during a famine at Alexandria Leontius eapoleos, 13 . However, throughout antiquity, the lake s waters teemed with fish and water-fowl P.Tebt. 3.867 Leontius eapoleos, 8 Rufinus, 27.10 . Alexandria depended on its neighbourhood for the supply of its daily needs of meat and vegetables. We may also recall that there were some 2,000 monks near Alexandria at Nitria, south of Lake Mareotis and that however frugal their diet might have been, they were amply sustained by products of the region Palladius VII, 1-2 Sozomen, VI, 29 . 8

M. EL ABBADI: LAKE MAREOTIS IN BYZANTINE TIMES Procopius, Buildings. H.B. Dewing transl. , 1961. London. P.Tebt. = Tebtunis Papyri, vol. III.2, A.S. Hunt & .G. Smyly transl. , 1933. London Rufinus, Historia Monachorum. Migne, Patrologia Latina 21, 1849. Gottingen. Sophrone de rusalem, Miracles de Saints Cyr et J ean, . Gascou transl. , 2006. Paris.

Sozomen, Historia Ecclesiastica. Migne, Patrologia Graecia 67. 1864, reprint 2002. Gottingen. Stephanus Byzantius, O pera, 4 vols., K.W. Dindorf ed. , 1825. Leipzig. Strabo,G eography, H.L. ones transl. , 1917. London. Thucydides, Historiae, H. Stuart ones transl. , 1897. Oxford.

9

LAKE MAREOTIS: REC

STR CTI G THE PAST

10

I. AWAD: MARYUT LAKE THROUGH MAPS

A Study of the Evolution of the Maryut Lake through Maps Ismaeel Awad

This study concerns the evolution of the shape of Maryut Lake and its connections with the Nile River and the Mediterranean Sea from antiquity until the modern day by studying and observing maps of the Maryut Lake region, classified chronologically from the oldest to the most recent.

The aim of this study is to use maps to measure the evolution of the Maryut Lake, to test the value of using the map as a tool in the study, to evaluate the type of results obtained, and lastly, to try to forecast the shape and state of the Maryut Lake in the future. The Maryut Lake region consists geographically of five different parts (Fig. 1). In the north, a sandy strip extends along the Mediterranean coastline. To the south there is a calcareous ridge known as Abusir-Mexx that has an average altitude of 30 m high. The center of the Maryut region is divided into two sub-divisions, the first is the Maryut Basin which is situated towards the east with an average altitude attaining sea level; the second sub-division is the Maryut Valley situated towards the west of the region. And finally, the Maryut Ridge lies towards the south-west and has an average altitude of 50 m high.

In the following pages, two different types of available maps are considered. The first type of map represents edited illustrations of Maryut Lake by different historians and ancient cartographers who built up their illustrations through visual observation or by drawing. The study of these illustrations using the superimposition method, essentially overlapping one map on another, cannot be applied, as any attempt to do so would result in the production of inaccurate results based on the fact that these illustrations display approximate distances and inexact locations. The second type of map utilised topographic maps, which when edited using accurate topographic methods, can produce more accurate results. The application of topographic methods began during the French campaign at the end of the 18th century in which levelling and goniograph tools are used to construct maps in Egypt. Topographic methods have developed over time through to the current use of satellite images in mapping. The data extracted from these maps are accurate enough to be simply compared and overlapped.

About 20 maps and illustrations are used in this study, of which the oldest map was edited at the end of the 16th century and the most recent was extracted from a Landsat satellite image produced in 2001.There are two different ways of studying these different types of maps, as will be revealed during the course of this paper. Fig.1: Maryut Lake region.D igitised by I. Awad. Archives CEAlex.

11

LAKE MAREOTIS: RECONSTRUCTING THE PAST Illustrations The illustrations utilised in this study date from the end of the 16th century until the end of the 18th century. Some of them were constructed ‘indoors’ without ever stepping foot in Egypt. They were done by using other maps edited by other publishers, used as base maps for adding new information. or this reason it is difficult to use these illustrations to derive any form of accurate reference or in application to an overlapping study.

the evolution of names and locations of settlements such as towns or villages existing all around the Mareotic region. Illustrations edited before the end of the 16th century used to ignore the Maryut Lake as it was of little importance at that time, certainly when compared with the town of Alexandria. Ancient map makers rarely mentioned Maryut Lake, locating it at the margin of the map and illustrating its location in small letters. They portrayed an ignorance of the lake’s shape, such as for example the illustration edited in 1548 by Pierre Belon du Mans (Fig. 2).

However, some useful information could be extracted from examining these illustrations, such as: the evolution of the connections between Maryut Lake and the Nile River, Nile branches and the Mediterranean Sea,

The illustrations used in this study are classified chronologically from the oldest to the most recent: Fig.2: 1548, Vray portraict de la ville d’Alexandrie en Egypte,P ierre Belon du Mans. J ondet G .,192,pl .I I. Archives CEAlex.

Fig.3: 1570,de tail from Aegyptus antiqua, including a focused map of the region of Alexandria, Abraham O rtelius. J ondet G .,192. Archives CEAlex.

12

I. AWAD: MARYUT LAKE THROUGH MAPS 1 .01: 1570, Abraham O rtelius (Fig. 3)2 It can be observed from the general map that the Maria, et Mareotis lacus qui et Ara potes (Maryut Lake’s) shape took a direction from north-east to south-west incorporating eight islands as shown in the figure.

River), which was also connected to the Nile River at its south-eastern end. The boats depicted in the Moeris Swamp show that there was extensive navigation during this period between Moeris Swamp, Maryut Lake and the Nile River.

Maryut Lake had two connections, the first was the connection with the Myris quod et Moeris stagnum (Moeris Swamp) to the south through the Lycus u ius (Lycus

The second connection related directly to the Nile River in three places:3 the first one to the north related to the Canopica fossa4 (Canopic Branch), the second connected the lake with the Canopic Branch at the north-east near the town of Schedia, and finally, the third connected the lake directly with the Nile River at its eastern side. .02:1 158,L ivio Sanuto (Fig. 4)5 In this map, the Bycheira lacus (Maryut Lake) has deformed in its shape forming a triangle enclosing a large number of islands (21 islands). Just one connection linked the lake to the Canopic Branch at its eastern end while no connection was illustrated with the Moeri dis lacus (Moeris Lake). .03:1 165,N icolas Sanson D’Abbeville (Fig. 5)6 The map of D’Abbeville has similar limitations to the previous map. The shape of Maryut Lake is similar although more rectangular and adopting more of a northwest to southeast orientation. However, most of the islands that had been previously observed have disappeared. .04:1 17,P aul Lucas (Fig. 6)7 No connections are shown with the Lac de la Mareote (Maryut Lake) in this map while some islands reappeared (nine in total). The Moeris Lake has totally disappeared, and at the same time the Lac du Madie (Madieh Lake/Abuqir Lake is illustrated for the first time.

Fig.4: 158, detail from Africae Tabula VII, Livio Sanuto. Archives Biblioteca Nazionale Centrale Di Roma.

.05:1 172,C laude Sicard (Fig. 7)8 In Sicard’s illustration the Mareotis Palus (Maryut Swamp’s) form has changed again to resemble the shape of a bird or a duck. Some connections are apparent between Maryut Lake and the Nile to the east that joined together before flowing into Maryut Swamp. A small lake named Nitraria Palus9 itraria Swamp was identified to

Fig.5: 165, detail from Royaume et désert de Barca et l’Egypte, Nicolas Sanson D’Abbeville.C ommercial reproduction.

1. There is another illustration of Ortelius dated from 1595, found in a late commercial reproduction from the beginning of the 21th century. 2. Aegyptus antiqua, map of the Antique Egypt. 3. There is a fourth connection shown in the focused map of Alexandria and not in the general map. It is situated between the first and the second connections at the north passing near the village of Chabrie pagus. 4. Mentioned in the focused map of Alexandria. 5. Africae Tabula VII, map of Africa, sheet no. 7. 6. Royaume et désert de Barca et l’Egypte, Kingdom of the desert of Barca and Egypt. 7. Carte de la Basse Egypte et du cours du Nil, map of Lower Egypt and the Nile River. 8. Carte de l’Egypte Ancienne, map of Ancient Egypt. The original map edited by Claude Sicard was not found. This is a synthesis of two copies extracted from the original map done by D’Anville and Danlisle and the place name’s list of Sicard. Sauneron & Martin 1982. 9. There are two lakes named Nitraria in the map: one near the Maryut Lake which is mentioned in the study and the other located in Wadi Natrun between Alexandria and Cairo.

13

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig.6: 17, detail from Carte de la Basse Egypte et du cours du Nil,P aul Lucas. Archives IFAO .

Fig.7: 172,de tail from Carte de l’Egypte Ancienne, Claude Sicard.Saune ron & Martin 1982, Archives IFAO .

Fig.8: 1743, detail from Carte d’Egypte,R ichard Pococke. Archives IFAO .

Fig.9: 1753,de tail from Carte de l’Egypte ancienne et moderne,R obert de Vaugondy. Archives J .- Y .E mpereur.

the south-east of Maryut Lake, connected with it through Acheron Canalis (Acheron Canal). The town of Marea is illustrated, but strangely to the north-west of Maryut Swamp and not on the southern shores of the lake as located nowadays.

that used to connect Moeris Lake with Maryut Lake. The towns of Plinthine.S. and Taposiris.P .11 were indicated but very far from the lake towards the west. .07:1 ca 1753,R obert de Vaugondy (Fig. 9)12 th This 18 century map indicates yet a further change to the shape of Birk Mariou (Maryut Lake) now depicted as being aligned from north-east to south-west. There is still

.06:1 1743,R ichard Pococke (Fig. 8)10 The shape of Birk Mariou.s ./ Lacus Mareia.f ./ Mariotis.S. (Maryut Lake) changed again in this map taking a triangular form. A lake is mentioned at the place of the Nitraria Lake but located further south than before, and the Moeris Lake appeared as Birk El K aroon/ Lacus Moeris feu Moerius (Karoun Lake).

10. Carte d’Egypte, map of Egypt. 11. Known as ancient navigation ports situated actually at the western arm of the Maryut Lake. 12. Carte de l’Egypte ancienne et moderne, map of Ancient and Modern Egypt. The base map is inspired from Claude Sicard’s map.

Bahr Bellomah i.e Mare seu Lacus Sine Aqua (lake or sea without water) was mentioned at the place of the old canal 14

I. AWAD: MARYUT LAKE THROUGH MAPS one easterly connection between Maryut Lake and the Nile River. A lake is mentioned and located at the same place of the Nitraria Lake as in the previous map. .08:1 1762,R igobert Bonne13 The Bonne map has many similarities with the previous map, with just one change which is the appearance of the town of Abusir located at the middle of the northern shoreline of the Maryut Lake. .09:1 1764,J acques Nicolas Bellin (Fig. 10)14 A connection between Lac Mareotis (Maryut Lake) and the Mediterranean Sea is shown for the first time to the west of Alexandria, with the disappearance of Lake Nitraria. Two connections appeared between Maryut Lake and the Nile (both meet at the same point, southeast of the lake). One of them approached from the east and the other from the Bahr-el-Helame ou Fleuve seché (canal or sea without water) from the south-east, and passed through a Lac de Sel (Salty Lake)15. .10: 179,E dme Mentelle & Pierre-G regoire Chanlaire16 In this map of 1799 Sebaga ou Birk Marioul17 (Maryut Lake) reappeared in the form of a triangle with what appears to be a sea connection to the west of Alexandria. The connection with the Nile River appeared through the Canal de Bahiré Bahirey Canal flowing into the Maryut Lake, and Lac de Natron (Natron Lake) appeared at the place of the ex. Nitraria Lake southeast of Maryut Lake. The Bahhr-bé-lâ- mâ ou Fleuve sans eau (canal without water) was also shown approaching from the Birket-il K erun (Karoun Lake) which is located to the south in the Feï um region, extending towards the southwest of Maryut Lake but without any intervening connections.

Fig.10: 1764,de tail from Carte des Embouchures du Nil, et partie de son cours, J acques Nicolas Bellin. Archives J .- Y .E mpereur. geo-referencing the maps in the same coordinate system, digitising each map in AutoCAD program, overlapping each digitised map one on the other, analysing the digitised features in ArcGIS program.

Towns are named around the Maryut Lake such as the town of Mariout (Marea) located on the northern shoreline of the lake and Tour des Arabes (town of Abusir) near to the Mediterranean Sea.

.01:2 180,F rench campaign,s cale: 1: 10. (Figs. 11 & 12)18 This is the first accurate topographic map concerning the Mareotic region. It was edited by the engineers of the French campaign who came to Egypt between 1798 and 1801. The map was constructed by accurate topographic instruments including a level and a goniograph.

Topographic Maps In order to extract precise locations and establish an accurate surface area, four steps were undertaken when studying the second type of map:

In this map, a cut is shown in the dike separating the Bohéï réh Marî oût ou lac Mareotis (Maryut Lake) from the Bohéï réh Maadî éh, lac Madié ou d’Aboûqî r (Madieh or Abuqir Lake). The dike was breached by the English army in April 180119 during the war between the French

13. Carte de l’Egypte ancienne et moderne, map of Ancient and Modern Egypt. 14. Carte des Embouchures du Nil,e t partie de son cours, map of the Nile river mouths, and a part of its stream. 15. The Salty Lake is at the place of the Nitraria Lake situated in Wadi Natrun. 16. Focused map Carte particuliè re et détaillée du Delta (Detailed map of the Delta) in the Carte physique et politique de l’Egypte (Physical and political map of Egypt). There is another similar map Carte d’Egypte (map of Egypt) edited by Jean-Baptiste Bourguignon D’Anville in 1765. Charles-Roux 1910. 17. Sebaga is near to Sabkha in Arabic language which means marsh, and Birk comes from Birka in Arabic, too, which means pond.

18. Carte d’Alexandrie, sheet no. 37 extracted from the Atlas of the French Expedition Carte topographique de l’Egypte et de plusieurs parties des pays limitrophes. 19. The dike was breached 19 April, 1801, by the English army under the orders of General Sir John Hely-Hutchinson, in order to cut off the freshwater supply brought by the Alexandria canal from the French garrison located in Alexandria, and to cut the road relating them to the main French army situated in Cairo.

15

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig.1 1: 180 Fig.12:

180,

, Carte d’Alexandrie, French campaign. Archives J .- Y .E mpereur. Carte d’Alexandrie,F rench campaign,or iginal scale 1: 10,.D

and English. As a result, the water in Abuqir Lake (which was situated at a higher altitude than the Maryut Lake and was directly connected with the Mediterranean Sea), was redirected into Maryut Lake, causing flooding in its eastern (125 km² ) and western (90 km² ) extremities.

igitised by I. Awad. Archives CEAlex.

The depth of Maryut Lake was measured in units of feet using depth soundings all around the lake. We can observe from the depths shown on the map that the deepest zone in the Lake was the north-eastern zone that was an average depth of about 4.5 m. Maryut Lake becomes shallower

Two lake limits are represented on the map, the first referred to the limit of the flood on 9 August, 1801,20 and the second showed the limit of the lake after the flood had returned to its preceding limit on 10 January, 1802.21

20. The flood limits were obtained from M. Tawiland s map scale: 1:128.000), engineer in the English army. Jacotin 1809-1828: 83. 21. Due to the effect of the northerly wind and to the reduction of water in the Nile River branches. 22. Maybe it was the connection mentioned before in the ancient illustrations between the Mediterranean Sea and the Maryut Lake at the west of Alexandria.

Another dried up old cut22 is illustrated linking Maryut Lake with the Mediterranean Sea to the east of the Anse de Marabou (Gulf of Marabou), west of Alexandria. 16

I. AWAD: MARYUT LAKE THROUGH MAPS

Fig.13:

186

, Carte des Environs d’Alexandrie,M ahmoud Bey El-Falaki. Archives CEAlex.

Fig. 14 : 1866, Carte des Environs d’Alexandrie, Mahmoud Bey El-Falaki, original scale 1:200,000. Digitised by I. Awad. Archives CEAlex.

time before the flood which was about 890 km 211.300 feddans). At this point, the lake’s maximum width (from north to south) was about 45 km, and its maximum length (from east to west) reached about 56 km.

from the north-eastern zone towards the south (2 m in average) and the west (1.5 m in average) until the lake attends 0 m at its limits. In addition, two levelling profiles23 were undertaken across Maryut Lake, one at the Maryut Valley24 and the other along the dried old cut mentioned previously.

23. Le Pè re 1809-1828: Pl. 19. 24. Between the ruins of the town of Marea and the Mediterranean Sea passing through the saint tomb of Aboû e l K eï r.

Thanks to digitising, we are able to ascertain with sufficient accuracy the precise surface area of the lake at that 17

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig. Atlas of Egypt, Vol. I, Lower Egypt, Survey of Egypt (G. Daressy), original scale 1: 50,. Digitised by I. Awad. Archives CEAlex. Fig. 16:

194

, map series 1: 10.

Survey of Egypt, original scale 1: 10,.

.02: 186,M ahmoud Bey El-Falaki,s cale: 1: 20. (Figs. 13 & 14)25 This map is interesting because El-Falaki (1872) mentioned that Lac Maréotis (Maryut Lake’s) shape at that time (1866) was similar to the lake’s shape when the town of Alexandria was founded in antiquity, according to Strabo’s notes.26 The author also defined the ancient branches of the Nile as the Canopic Branch and Schedia Canal.

Digitised by I. Awad. Archives CEAlex.

Lake was also reduced in area to 450 km² (which is about 50% of its area as indicated in 1801). This reduction in area occurred in just 65 years, and was especially apparent in the eastern part where its maximum length was about 24 km.

25. Carte des Environs d’Alexandrie, map of the surroundings of Alexandria 26. El-Falaki 1872: 99.

From this map, it can be observed that the Lac d’Aboukir (Abuqir Lake) has reduced in area. Proportionally Maryut 18

I. AWAD: MARYUT LAKE THROUGH MAPS

Fig.17: 1970 s,m ap series 1: 25.0,E Archives CEAlex.

gyptian G eneral Survey Authority,or iginal scale 1: 25,0.D

igitised by I. Awad.

.03:2 19 1,Sur vey of Egypt,s cale: 1: 50. (Fig. 15) 27 The area of Maryut Lake continued to reduce particularly on its eastern margins. The Egyptian government began to have some interest in this zone by managing a system of irrigation canals and drains (for example: El-Omum Drain, whose waters flowed from the south-eastern side of the Maryut Lake) in order to cultivate the region around Maryut Lake. The area of the lake was reduced to 340 km² (38% of 1801s area). However, the dimensions of the lake from east to west appear to have extended compared to previously (about 75 km). The ancient cut between the Mediterranean Sea and Maryut Lake to the west of Alexandria is again mentioned.

north-eastern shores, particularly with the construction of Nozha30 and the seaplane base. By now most of the western part of the lake had dried up, perhaps partly due to the fact that this map was surveyed in the dry season.31

28 .04:2 1930,S urvey of Egypt,s cale: 1: 10. The area of Maryut Lake continued to reduce in size particularly towards the south-east as the lake became increasingly exploited for agriculture purposes. An artificial salt marsh appeared for the first time in the middle of the lake in order to produce salt.

A great deal of development had also occurred in the lake:

.06:2 1970s ,E gyptian G eneral Survey Authority, scale: 1: 25.0 (Fig. 17) 32 By the 1970s a big development had occurred at the eastern and southern parts of Maryut Lake with an expansion of irrigation canals and drains. The Nubareya Canal was completed in 1976 channeling waste water runoff from agriculture lands into the lake, and then into the Mediterranean Sea via the Mexx pump station.

In the middle of the lake, the salt marshes became much larger (it has expanded ten fold in just 20 years). Big changes also took place in the north-eastern part of the lake. The seaplane base was transformed into a big fish farm, and Nozha airport became a part of the land in order to separate the new fish farm from Maryut Lake.

.05:2 194,S urvey of Egypt,s cale: 1: 10. (Fig. 16) 29 The lake was further reduced on its eastern shores. This map also indicates early interest in expansion along its

Maryut Lake’s surface area has reduced further to just 210 km² (51.000 feddans) (which is about 1/4 of its area in 1801).

27. The maps of this series have been collected from: Survey of Egypt 1914. Eight maps have been used in digitising: sheet no. 009 Ramleh, sheet no. 010 Abuqir, sheet no. 020 Mallaha, sheet no. 021 Alexandrie, sheet no. 022 K afr El Dawar, sheet no. 046 Hammam, sheet no. 047 Bahig and sheet no. 048 K ing Mariout. 28. German edition, four maps have been used in digitising: 88/42 Hammam, 88/48 G hayata, 92/48 Alexandria and 92/54 Damanhur. 29. Arabic edition, four maps have been used in digitising: 0° 31’/29° 40’ Abu El-Matamir, 31° /29° K rair, 31° /29° 40’ Alexandria and 31° /30° 20’ Damanhur.

. 30. Known at that time as Fouad the First airport: 31. The western zone s water level is directly influenced by dry and wet seasons. 32. Fourteen maps have been used in digitising: 90/450 Hammam, 90/465 Mahatet El-G harbaneyat, 91/450 K om Makhoura, 91/465 Burg El-Arab, 92/465 Burg El-Arab, 91/480 Mahatet Bahig, 91/495 Aeroport of Amreya, 92/480 K reir, 92/495 Amreya, 93/480 El-Deir, 93/495 Mexx, 93/510 K om Echo, 94/510 Ramleh and 94/525 Munshaat Buline.

19

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig.18:

19

,m ap series 1: 50.,E

gyptian G eneral Survey Authority.D igitised by I. Awad. Archives CEAlex.

Fig.19:

201

,L andsat satellite image,r esolution 15m

/ pxl.D igitised by I. Awad. Archives CEAlex.

.07:2 19,E gyptian G eneral Survey Authority, scale: 1: 50. 0 (Fig. 18) 33 A huge development occurred in the southern and eastern parts of the Mareotic region in the early 1990s which enabled the entire lake to be part of the same irrigation system, and large areas of the middle zone of the lake were transformed into fish farms.

.08:2 201,L ANDSAT satellite image,r esolution: 15m / pixel (Fig. 19) 34 Satellite images were used to study more recent changes in the lake morphology due to the unavailability of more

These were the last map series produced concerning the Mareotic region, and edited by the Egyptian General Survey Authority.

33. Six maps have been used in digitising: NH35-L5d Hammam, NH35L6c Burj al-’Arab, NH35-L6d K ing Mariout, NH35-P3a Sidi K irayr, NH35-P3b Alexandria and NH36-M1a K afr El-Dawar. 34. Downloaded from the University of Maryland site (www.umd.edu).

20

I. AWAD: MARYUT LAKE THROUGH MAPS Table 1: Comparison table between different periods of Maryut Lake.

Fig.20: Evolution diagram of Maryut Lake’s surface areas .

contemporary maps since the last map series was produced in the 1990s.

as the construction of fish farms and artificial salt marshes, all of which require stable quantities of water.

A topographic map was extracted from a LANDSAT satellite image of a medium resolution of 15 m/pixel, which is enough to digitise the limits of the lake, marshes, fish farms, wide canals and drains, but not enough to define the other small features such as buildings, narrow canals, secondary drains or roads.

n the other hand, natural influences to the shape of Maryut Lake have a limited impact nowadays (such as rain and ground water). However, any loss in the lake’s surface area in the future depends on the rate and types of uses of the lake and the implemented management plan. Forecasting the Shape of Maryut Lake Digitising the topographic maps does provide an opportunity to forecast the future shape of the Maryut Lake region, if left to evolve naturally without any external interference. A forecasting study is undertaken by extracting and digitising different contour levels of the Maryut Lake region. The maps utilised in this digitising exercise were the map series of scale 1:25,000 edited in 1940s, and the same map series edited in 1970s; these maps series were chosen because they are the most accurately contoured maps that covered the whole Mareotic region. The accuracy of the results of such an exercise is after all dependent upon the precision of the maps used, and on the data extracted.

The map showed that the area of Maryut Lake’s is still about 210 km² (51,000 feddans). Its maximum dimensions attended 66 km in length and 11 km in width. From the table (Table 1) and the diagram (Fig. 20) showing the evolution in surface areas of Maryut Lake from 1801 until 2001, we can notice that the lake has lost 75% of its surface area in just 200 years. The greatest loss occurred between the period from 1801 to 1866, in which the lake lost half of its area due to drying and land fill. Maryut Lake became more stable with regard to its rate of surface area lose from the 1970s until now. During this 30 year period it lost 0.1% of its surface area. The main contributor to the reduced rate of water loss was irrigation and the increase in activities around the shores of the lake, such

The 0 m contour line was the first digitised, which refers to the limits of the Maryut Basin, which is the same as sea level (0 m). By comparing this contour map with Maryut 21

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig. Comparison bet een Maryut Lake French campaign map series in the Mareotic region.D igitised by I. Awad. Archives CEAlex.

top and m le el bottom

Results (Fig. 22) The surface area of Maryut Lake reduced by less than a quarter (24% ) over a period of 200 years (1801-2001). This does not bode well for the future of Maryut Lake, particularly when you consider the intensive construction projects in and around the lake especially at its eastern end, its northern and at its north-western parts having been continuously divided up, dried out and filled to construct tourist villages and settlements.

Lake s limits in 1801 without flood , it can be seen ig. 21) that they are approximately identical. That means that the lake’s limits in 1801 had reached the 0 m level. Using this kind of study, the Maryut Lake’s limits can be compared to any topographic map with the different digitised contour levels such as the levels of 0 m, -1 m, -2 m and -3 m, in order to study the lake water’s level in any period. We chose these precise levels because the deepest altitude of the Mareotic region is about -4 m which is limited to a small area in the north-eastern part of the lake and the highest level that the lake’s water reaches did not exceed 1 m.

The use of Maryut Lake itself such as fish farms or salt marshes has increased from 1 km² (0.3% of Maryut Lake’s area) in 1930 into 39.6 km² (18.6% of Maryut’s Lake area) 22

I. AWAD: MARYUT LAKE THROUGH MAPS

Fig.2:

Comparison between Maryut Lake 180-

201.D

igitised by I. Awad. Archives CEAlex. Charles-Roux, F., 1910, Les origines de l’Expédition d’Egypte. Paris. Carte d’Alexandrie, 1801, French campaign. Carte topographique de l’Egypte et de plusieurs parties des pays limitrophes [ MAP] . El-Falaki, Mahmoud Bey, 1872, Memoire sur l’antique Alexandrie,s es faubourgs,s es environs. Copenhagen. Jacotin, M., 1809-1828, Mémoire sur la construction de la carte de l’Egypte. Description de l’Egypte, Etat Moderne, Tome 2, 2è me Partie. Paris. Jondet, G., 1921, Atlas historique de la ville et des ports d’Alexandrie. Cairo. Le Pè re, G., 1809-1828, Planche 19, Vallée du Nil et Lac Maréotis. Description de l’Egypte, Antiquités, A. Vol. 5. Paris. Livio Sanuto, 1588, Africae Tabula VII. Biblioteca Nazionale Centrale di Roma [ MAP] . Lucas, P., 1664-1737 (1724), Voyage du Sieur Paul Lucas, fait en MDCCX IV & par ordre de Louis X IV, dans la Turquie, l’Asie, Sourie, Palestine, Haute et Basse Egypte. Paris. Pococke, R., 1745, Observations on the Islands of the Archipelago, Asia Minor, Thrace, Greece and some other parts of Europe. A Description of the East and some other countries. Vol. II. Part II. London. Sauneron, S., & Martin, M., 1982, O euvres III - Parallè le géographique de l’ancienne Egypte et de l’Egypte Moderne. Cairo. Survey of Egypt, 1914, Atlas of Egypt, Vol. I, Lower Egypt, Cairo. de Vaugondy, R., 1753, Carte de l’Egypte ancienne et moderne [ MAP] .

in 2001. That means that the lake’s use has increased about 40 times in just 70 years for salt production. C onclusion The map can be a very useful tool in a study of temporal evolution of any geographic feature, such as the case of Maryut Lake. When assessing the results of such studies, the following conditions must be considered to increase the reliability of the product: Data must be accurate to all the region of study the more accurate the data the better the results). Data maps and satellite images must be accessible. This is a particular problem when investigating the Maryut Lake region where some maps of some of the series were not found due to negligence of archiving or due to security reasons.35 Data must be updated regularly to make the study of the region reliable and more accurate. In the case of the present study, the last series of maps produced concerning the whole region was in 1991. This means that for the last 17 years, no other map series concerning the Mareotic region have been produced, despite continued development and huge observable changes having occurred in the Mareotic region since this time. Bibliography D’Abbeville, N.S., 1655, Royaume et désert de Barca et l’Egypte [ MAP] . Bellin, J.N., 1764, Carte des Embouchures du Nil, et partie de son cours [ MAP] .

35. Several military zones exist around the Mareotic region causing the exclusion of some maps which include these.

23

LAKE MAREOTIS: RECONSTRUCTING THE PAST

24

L. BLUE: LAKE MAREOTIS RESEARCH PROJECT

Lake Mareotis Research Project Lucy Blue

Alexandria was by any standards one of the great cities of the Mediterranean. Since its foundation in 331 BC and for almost a millennium to follow, it was the political, economic and cultural capital of Egypt, and one of the most significant emporia and complex ports in the Hellenistic and Roman worlds. As such, Greco-Roman Alexandria has been the subject of much historical and archaeological research (Fraser 1972; Haas 1997; Empereur 1998; Goddio, et al. 1998) that has revealed the wealth of the city, much of which was generated by trade through its important and monumental harbour. However, the complexity of the Alexandria harbours system, which included not only harbours on the sea, but also on Lake Mareotis, has never been thoroughly understood. The significance of Lake Mareotis in the history of the Greco-Roman port-city of Alexandria, is undeniable (Fig. 1). Settlements and industrial units located along the shores of Lake Mareotis, located to the south and west of the city, are known to have produced amongst other things, glass, textiles, pottery and wine (Empereur & Picon 1998), all of which were transported to the city. Yet the dynamics of this productive hinterland have not been fully appreciated in terms of the important contribution the region made to the economy of Alexandria and thus to Egypt as a whole.

The principal objective of the Lake Mareotis Research Project was to undertake a detailed systematic survey of the western extension of the lake in order to record, quantify and assess the archaeological resources of the area in a comprehensive manner. As a result, we hoped to determine a better understanding of the nature and extent of economic and maritime activities in the Mareotic region and to explore the relationship between the component parts of this complex system and how they developed and changed over time, in order to determine a more comprehensive understanding of the economic functions of ancient Alexandria.

In this context, the Lake Mareotis Research Project set out to investigate and determine the role of the lake, the function of the sites around its shores and the part they played in supporting the metropolis of Alexandria in antiquity.

In ancient times, Lake Mareotis was connected to and fed by the Canopic Branch of the River Nile (Strabo 17.1.7). By the 12th century the lake had silted up but prior to its silting it was much larger and deeper (Warne & Stanley 1993: 53-8; Said 2002). The main body of Lake Mareotis currently covers an area of about 90 km2 and is defined to the north by a chain of lithified carbonate ridges extending parallel to the northern shoreline of the lake and the Mediterranean coast (Fig. 2). It is nearly rectangular in shape and extends south of Alexandria for about 12 km, merging along its southern and eastern shores into the Western Deltaic region. It would once have extended considerably further south, quite possibly a navigable distance of about 50 km. It is separated from the lake’s main body by causeways and shallows and its relative size varies according to the season.

Envi ronmental and Political C ontext of Lake Mareotis Lake Mareotis represents one of the most distinctive geomorphological features along the north-west coast of Egypt (Warne & Stanley 1993). It is unique compared to other lakes on the north coast of Egypt, being the only one with no direct connection to the Mediterranean Sea, but originally with direct access to the River Nile and hence to the whole of Egypt. Therefore, it is believed that the location and characteristics of Lake Mareotis gave the city of Alexandria one of its major advantages (Strabo 17.1.7).

C ontext of the Lake Mareotis Research Project The Lake Mareotis Research Project is a collaboration between the Centre for Maritime Archaeology, at the University of Southampton, and Department of Underwater Antiquities (DUA) of the Supreme Council of Antiquities SCA , Alexandria. The project first started in 2004 when a British Academy Small Grant was awarded to support a pilot survey season in the western arm of the lake, to the west of Alexandria. The survey covered an area of approximately 40 km east-west by 3 km wide, along the shores of the western extension of Lake Mareotis, from Sidi Kerir to El-Hammam. During this first season over 60 sites where identified, many equating to new discoveries (Blue & Ramses 2005). The potential wealth of the sites in the area was thus realised and subsequently further funding was sought and awarded from both the British Academy and the Leverhulme Trust to conduct a more detailed survey on both shores of Lake Mareotis and on an island that is located in the eastern region of the survey area.

Thus at present, Lake Mareotis is a body of shallow brackish water that occupies about 13% of an extensive subsea-level depression that is sometimes referred to as the Maryut Depression. The remaining 87% of the depression has been drained and is now used primarily for agriculture. The main body of the lake is heavily polluted by untreated sewage and industrial waste. It has also been divided into many sub-basins which are used for industrial and agricul25

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig.1: Location of Alexandria and Lake Mareotis ( Lake Mareotis Research Proj ect) .

26

L. BLUE: LAKE MAREOTIS RESEARCH PROJECT tural purposes. Moreover, the shores of the lake are subject to continuous irrigation, drainage and reclamation, which means that it is constantly changing and unstable (Warne & Stanley 1993: 29-30 rihy, et al. 1996: 282 . The average depth of the remaining lake is less than 1 m, although it is assumed to have been greater in antiquity to allow boats to sail safely.

and Boussac & El Amouri (this volume) on the Hellenistic port of Taposiris Magna. In addition, rescue work has been undertaken by the Supreme Council of Antiquities (SCA) at a number of smaller sites around the shores of the lake largely in advance of development. The majority of known sites have not been systematically studied however, and the pilot survey conducted in 2004 revealed that only a fraction of the available evidence is currently recognised. In addition, as indicated previously, the area is massively under threat from development and pollution. Large sections have been quarried, particularly of the northern limestone ridge of the study area, while extensive sections of the remaining lake are being exploited for aquaculture and its shores cultivated for agriculture. Much of the land has been sold for development, and along the edges of the lake harmful polluting industries such as salt factories, geochemical and petrochemical processing plants, as well as numerous sewage plants, all present a real threat to this lacustrine environment – the landscape is constantly changing and with it the archaeological record is being destroyed (see Ramses & Omar this volume). For all these reasons it was determined that a comprehensive survey of the area was essential in order to record what remains before the information is lost forever.

Due to its location and characteristics, Lake Mareotis played a significant role in the internal and international transport system of Greco-Roman Egypt (Rodziewicz 1998), particularly after it had been connected to the sea by a navigation canal at Alexandria (Strabo 17.1.7: 31). It became an important conduit of communication between the River Nile and the Mediterranean Sea, and river boats would have transported trade goods from the Nile Valley to the harbours of Mareotis and onwards to Alexandria and beyond. As indicated, Mareotis also supported economic activities around its shores such as the cultivation of vines and grain, and the manufacture of ceramics, wine and glass (Empereur & Picon 1998). However, despite the important role of Lake Mareotis, previous research conducted in the region has been largely limited to specific areas and topics, such as the work of El-Fakharani (1983) amongst others, on the Byzantine port of Marea (Philoxenite) (see now Haggag, Babraj & Szymańska, Rodziewicz and Pichot this volume), Empereur & Picon (1998) on the amphorae and wine production in Mareotis (also see Dzierzbicka this volume),

Objective s of the Lake Mareotis Research Project The objectives of the collaborative Lake Mareotis Research project were to:

Fig.2: An approximation of the ancient and present limits of Lake Mareotis ( E.K halil) . 27

LAKE MAREOTIS: RECONSTRUCTING THE PAST Systematically survey the western arm of Lake Mareotis (shores 40 x 3 km wide, and Mareotis Island,1 3.7 km in length and up to 680 m wide , bearing in mind the complex palaeogeographical history of the lake. Record, quantify and assess the archaeological resource. Determine the extent, chronology and, where possible, function, of each site. Determine the nature and the extent of the economic and maritime activities of the region. Determine the significance of the Mareotic region in relation to Alexandria, particularly during the Hellenistic, Roman and Byzantine periods when the lake is believed to have been most active. Identify the degree of threat to the archaeological resource in relation to each site before it is lost to development and pollution.

A luxgate Radiometer Magnetometer geophysical survey was conducted to establish the layout of structures beneath the surface at specific sites.2 Ceramic survey was undertaken to determine the chronology, nature and function of each site. Sedimentological survey to determine changing lake levels and identify geomorphological changes was conducted at selected sites. Limited site-clearing to articulate walls and identify more detailed building phases, was undertaken. A photographic archive was compiled of each site and features within the site. Analysis of the five survey seasons conducted in the western arm of Lake Mareotis between 2004 and 2008 is almost complete, and brief preliminary observations are presented in this paper. The final results will be published in a single monograph detailing the work undertaken and an interpretation of the final results Blue & Khalil forthcoming . In total, over 70 sites were documented, ranging in date from the Hellenistic period to the 7th century AD (Fig. 3). A clear distinction was noted between the nature of the sites on the northern shores compared to those in the south and on Mareotis Island. The north shore sites were clearly defined, but quite heavily eroded, tell sites. The sites to the south appear to represent more dense urban concentrations with associated agricultural and industrial areas, particularly towards the east where the island ridge may have once been an extension of the southern shore and the ridge associated with the site of Marea. Many of the sites, particularly on the southern shore, were partially buried under aeolian and alluvial sediments deposited by the eroding hillside and wind-blown material from the north.

Methodology As a result of the pilot survey conducted in 2004 a total of over 60 sites were identified, as previously mentioned. Some of these sites were already documented in the archaeological record, some having been partially excavated by the Western Nile Delta section of the SCA and foreign missions, or noted but not recorded in detail in previous publications. However, at least half of the sites identified are believed to be new discoveries. The sites were prioritised on the basis of their maritime significance, particularly where maritime installations such as moles, jetties, quays, mooring rings etc., had been identified. GPS coordinates were taken to fix the position of each site located, each site was photographed and basic data pertaining to its nature were recorded. Since 2004 four further seasons of survey have been conducted between autumn 2006 and summer 2008, with the objective of recording in detail the maritime installations and to place the sites in their environmental and economic context. In order to achieve these objectives, the following methodology was implemented: A desk-based assessment of the region was undertaken with the aid of high resolution Quickbird satellite images, to establish the nature and extent of the sites already identified. This information along with all the data acquired as part of the survey project, is archived in a GIS-based database. Each site was surveyed using a GPS-based Real Time Kinematic (RTK) satellite navigation system to produce detailed topographic maps of each site. Specific buildings and/or features were planned using a Leica TCR705 Total Station, and the data was downloaded in the field into AutoCAD via TheoLt software.

N orth Shore Sites (Fig. 3) A total of ten sites were identified on the northern shore. They date from the Hellenistic to Late Roman periods with a concentration of activity in the Roman period. They were primarily tell sites and their size ranged from c. 7000 to 180,000 m2. The ceramic assemblages at the sites included a mixture of amphorae, finewares, tablewares and cooking wares. Many of the finewares and amphorae were imports. Many of the sites on the northern shore contain wells and cisterns as well as evidence for mortar-lined basins, probably from bath complexes, which would suggest a domestic context. Limited evidence of irrigation exists in the form of water wheels or sakkias, and qadus pots associated with water wheels or sakkias were found at some Early and Late Roman sites. The evidence suggests that these sites were largely domestic in nature and each site represents a fairly contained and distinct unit with a clear association with

2. Preliminary investigations in the 2007 season were not as insightful as we had hoped, due to the fact that the fluxgate radiometer M36 we had been using had limitations. Thus, the 2008 geophysical survey was conducted with a Bartington Instruments Grad 601-2 dual sensor fluxgate gradiometer. The dual sensors survey at twice the rate of a single sensored instrument and with more refined results.

1. The Lake Mareotis Research Project refers to this ridge, largely surrounded by water located in the south-eastern region of the survey area some distance from the southern shore of the lake, as Mareotis Island. However, its island status in antiquity has yet to be determined and it is also acknowledged that it is not, and was not in antiquity, the only island in the lake.

28

L. BLUE: LAKE MAREOTIS RESEARCH PROJECT

Fig.3: Sites located along the northern shoreline,t he southern shoreline and on Mareotis Island ( Lake Mareotis Research Proj ect) . sites extend across the lake towards each other. Thorough investigation of the ceramic samples collected from these sites will shed more light on their possible nature and relationship.

the waterfront. A number of the settlements featured jetties extending into the lake on their southern shores. The survey also revealed possible spatial and functional relationships between a number of sites on the northern shore and specific sites on the island and along the southern shore. For example, it was noticed that at least one of the largest sites on the northern shore of the lake (AlGamal, Site 204) is located immediately opposite a settlement located at the western end of the island (Site 23; see Hopkinson this volume). Quays and jetties from these

South Shore Sites (Fig. 3) A total of 26 sites were identified on the southern shore. The sites ranged in date from the Hellenistic to the Late Roman periods, but generally exhibited limited activity in the Early Roman period. The earlier period displays some imports, largely amphorae from the Aegean region, but 29

LAKE MAREOTIS: RECONSTRUCTING THE PAST

is predominantly composed of locally produced material, while the Late Roman material reveals a greater amount of imported pottery from all over the eastern Mediterranean.

Fig.4: Topographic,c eramic and auger survey plan of Site 4 on M areotis Island ( image Lake Mareotis Research Proj ect) .

The sites towards the western extremity of the southern shores formed distinct and in some cases (Site 44 and Site 109) quite substantial units, and are spaced well apart. The archaeology at these sites was largely concentrated at the foot of the limestone ridge (Fig. 4). The sites extended north towards the shore of the lake where many exhibited a series of linear features that extend parallel to the shoreline down to the water’s edge and were open to the E-NE. They appeared to support low-lying areas of marshy ground between the main occupation area and the shoreline ridge, or in many cases a boundary wall to the north. This type of linear feature, known as a ‘lake wall,’ essentially contained a body of water and could have been used for agricultural purposes, perhaps for feeding animals or harvesting reeds, as is still practiced in the region today. This was noticed at a number of sites and also towards the western end of the island (see below) (Fig. 5).

Fig.5: An example of a ‘ lake wall’ feature ( Site 4) ( photo Lake Mareotis Research Proj ect) .

To the east, the sites were located on a limestone ridge that is effectively an extension of the ridge upon which both Mareotis Island and the site of Marea are situated. The nature of these sites appears to reflect similar dates and activities to that of the eastern end of the island. A number of sites have been identified that contained wine production facilities and/or sakkias designed to facilitate the shifting of water for example Site 100, and Sites 01 and 106 . Generally, however, these sites are less substantial than 30

L. BLUE: LAKE MAREOTIS RESEARCH PROJECT of a few ceramics previously identified by El- akharani El- akharani 1983: 176-178 . Therefore, these buildings might contribute towards resolving the mystery of the so far elusive Pharaonic harbour of Marea raser 1972: 146 El- akharani 1991: 28, 1983: 176 .

those on the island or by comparison with Marea, and they tend to cluster opposite the western end of the island, or along the ridge that extends between Marea and the island. Further investigation is needed to identify the relationship between the eastern end of the island, the south-easterly sites and the western extent of Marea.

Mareotis Island (Fig. 3) Mareotis Island is a limestone ridge joined to the southern shore at its western end. It is possibly an adjunct of the southern shore that has extended out into the lake from the Marea Ridge. It is aligned roughly northeast – southwest and is 3.7 km in length and 680 m at its widest point. Some 40 sites were initially identified primarily concentrated at the eastern and western ends of the island, with a few sites in its central section that were largely industrial-scale Early Roman kiln sites (see Hopkinson this volume).

Two sites stand out as exceptions to the others noted along the southern shores; the Kibotos (the square-shaped harbour) (Site 09), and the ‘complex building’ (Site 13) (Figs. 6 & 7 . The Kibotos Site 09 is essentially a harbour area constructed of large limestone blocks (0.8 x 1 m), defined by a series of moles that enclose an area some 60 m long to S and 36 m E to W wide see Khalil this volume). A second site, known as the ‘complex building’ (Site 13), is located a short distance to the southwest of the Kibotos and is composed of a continuous wall constructed of huge blocks (1.3 x 1.1 m), with a smooth outer surface that slopes up and inwards from the foot of the building. It has been proposed that both these buildings date to the Pharaonic period (Haggag 1984: 277-280). No earlier Pharaonic material has yet been identified at any of the sites recorded in the survey area, with the exception

The majority of the material is Hellenistic in date, particularly towards the eastern end of the island, with a flurry of activity again in the Late Roman period, particularly towards the western end of the island. The ceramic assemblage indicates some imported material especially for the earlier Hellenistic period, but in general it is locally

Fig.6: The K ibotos site ( Site 09) ( photo Lake Mareotis Research Proj ect) .

Fig.7: The ‘ complex building’ ( Site 13) ( photo Lake Mareotis Research Proj ect) .

31

LAKE MAREOTIS: RECONSTRUCTING THE PAST produced, particularly the production of amphorae. The nature of the structures, with the exception of the centrally located industrial-scale amphorae production kiln sites, is essentially urban in character. The scale of the buildings, particularly those located on the ridge towards the east, is impressive, and many are associated with numerous waterfront structures located along the northern shore on the edge of the coastal plain. At the western end of the island, the settlement extends from the ridge to the north over a flat plain. Whereas the ridge supports further urban structures, the plain to the north appears to be more agriculturally orientated, with features including possible ‘lake walls’ and sakkias similar to those identified along the south-western shores of the lake (see above).

the lake’s Mareotic Arm (see Khalil this volume). The frequency and abundance of these installations along the shores of the Mareotic Arm reflect the extent of maritime activity that was taking place in this region. Besides the jetties and quays which are positioned perpendicular to the shore, another type of maritime installation has also been identified, positioned parallel to the shore, and described as a lake wall. The so-called lake walls define the shores and protect them from the effects of silting and sedimentation. As many as five lake walls were discovered along the southern shore of the lake and the northern shore of Mareotis Island, areas which are more subject to silting as a result of the prevailing northwest wind. The lake walls could have been utilised as docking facilities for merchant vessels, but were more probably associated with agricultural activities, perhaps the cultivation of reeds as is still witnessed in the region today.

There is also evidence for quarrying on the island, as well as rock-cut tombs. Towards the eastern end of the island, a possible town wall has been identified together with a series of jetties and platforms associated with the waterfront, possible warehouses and a small inlet embayment. All these are discussed in more detail in Hopkinson (this volume).

The Lake Mareotis Research Project has also recorded numerous archaeological sites of industrial and agricultural nature such as amphorae production sites, wineries and the remains of several water wheels (sakkia) which were used for lifting water for irrigation and for use in baths and other domestic purposes. With one exception, all the amphorae and wine production sites are located on the southern shore of the lake or on Mareotis Island. In antiquity, the southern shores of the lake were characterised by a fertile plain which supported and enabled agricultural activities to flourish. In addition, the abundance of calcareous clay, particularly suitable for amphorae production, resulted in a thriving large-scale amphorae industry. The study of the ceramic assemblages collected from the surveyed sites indicate that the amphorae production centres were active from the Ptolemaic to the Late Roman period.

What is worthy of note is the fact that the majority of sites on the island are concentrated along the ridge and the northern shores, and that several were clearly built in a very specific relationship with sites on the lake s north shore (see above). The northern concentration of sites has also raised questions about the island’s status. Does the fact that so few sites have been found along the island’s southern shores suggest that it may have been linked to the southern shore of the lake in antiquity? Recent geomorphological work suggests that in fact the area to the south of the island has always been waterlogged to some degree, but that the waters were shallow and access by boat was problematic, and as such activities in this area were restricted.

Since no lake-side archaeological sites were recorded on the southern shore of the lake opposite Mareotis Island, questions have been raised about the nature of the island and whether or not it was indeed an island in antiquity. However, sedimentological analysis carried out during the final survey season revealed that Mareotis Island was probably an extension of a ridge that extended some 10 km to the west from the site of Marea. In addition, it was ascertained that the area between the island and the southern shore of the lake was either waterlogged or marshland in antiquity (Flaux forthcoming).

Preliminary C onclusions The Lake Mareotis Research Project has identified and systematically surveyed over 70 sites dating from the Hellenistic period through to the 7th century, in what still remains of the western arm of the once much larger Lake Mareotis. The sites vary extensively in nature, size and function, but demonstrate the importance this area had in relation to Alexandria and the mechanisms by which Mareotic products were produced and arrived at the city. Local pottery production is indicated by the numerous kiln sites discovered in the region, particularly on the island. There is also considerable evidence for imported pottery from as early as the late 4th century BC and as late as the 7th century AD. The nature and size of the sites identified varied from rural small holdings to large urban settlements, with associated production sites, water storage facilities and agricultural and industrial complexes.

Previous research conducted in the Mareotic region has been largely limited to specific sites however, the Lake Mareotis Research Project has systematically investigated all the archaeological remains along the shores of the lake’s western arm. The project also assessed the results of previous archaeological work undertaken in the Western Deltaic region. Based on this, it is reasonable to suggest that the Mareotic Arm was in fact the most active area in Lake Mareotis in antiquity, and that most of the Mareotic products arriving in Alexandria were in fact coming from the western Mareotic Arm, rather than from settlements associated with the lake’s main body (see Khalil this vol-

Different types of waterfront installations were recorded, including more than ten anchorage facilities, which form the majority of maritime installation along the shores of 32

L. BLUE: LAKE MAREOTIS RESEARCH PROJECT que d’Athè nes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole franç aise d’Athè nes, 11- 12dé cembre 198. Athens, Paris. Empereur. J.-Y. (ed.), 2002, Alexandrina 2. Cairo. Empereur, J.-Y., & Picon, M., 1998, Les ateliers d’amphores du Lac Mariout. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athè nes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole franç aise d’Athè nes, 11- 21 décembre 198: 75-88. Athens, Paris. Flaux, C., forthcoming, Geomorphological survey of the south-west shores of Lake Mareotis. In L. Blue & E. Khalil (eds.), A Multidisciplinary Approach to Alexandria’s Economic Past: the Mareotis Case Study. Oxford. Fraser, P.M., 1972, Ptolemaic Alexandria, Vol. I. Oxford. rihy, .E., 1996, Some Proposals for Coastal Management of the Nile Delta Coast. O cean & Coastal Management 30.1: 43-59. Goddio, F., Bernard, A., Bernard, E., Darwish, I., Kiss, Z., & Yoyotte, J., 1998, Alexandria: The Submerged Royal Q uarters. Oxford. Haas, C., 1997, Alexandria in Late Antiquity: Topography and ocial Con ict. Baltimore. Haggag, M., 1984, Ports in Ancient Egypt till the Arab Conquest. Unpub. MA diss., Alexandria University. Hassan, F.A., 1997, The Dynamics of a Riverine Civilization: A Geoarchaeological Perspective on the Nile Valley, Egypt. World Archaeology 29.1: 51-74. Rodziewicz, M.D., 1998, Fom Alexandria to the West by Land and by Waterways. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athè nes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole franç aise d’Athè nes, 11- 21 décembre 198: 93-103. Athens, Paris. Said, R., 2002, Did Nile Flooding Sink Two Ancient Cities? Nature 415: 37-38 Strabo, G eography, H.L. Jones (transl.), 1949. London. Warne, A.G., & Stanley, D.J., 1993, Late Quaternary Evolution of the Northwest Nile Delta and Adjacent Coast in the Alexandria Region, Egypt. J ournal of Coastal Research 9.1: 26-64.

ume). Finally, the Lake Mareotis Research Project has outlined the true scale of activities in the Mareotic Arm with respect to industrial, agricultural and urban sites, being much more extensive than previously suggested, thus supporting the significant importance of the area and highlighting the role it played in the economy of Hellenistic and Roman Alexandria. Acknowledgments The author would like to thank the Supreme Council of Antiquities for permission to work in the Mareotis region, particularly colleagues from the Underwater Department who were instrumental to the success of this project, in particular the support and help of the project co-director Dr Sameh Ramses and project manager Dr Emad Khalil. The project would not have been possible without the financial backing of both the British Academy and the Leverhulme Trust particularly for funding Dr Khalil’s two-year post doctoral fellowship. Finally, thanks is due to all students of Southampton and Alexandria Universities and Egyptian and UK colleagues who worked on the project without whom the survey would not have happened. Bibliography El-Fakharani, F.A., 1983, Recent Excavation at Marea in Egypt. In G. Grimm, H. Heinen & E. Winter (eds.), Aegyptiaca Trevernsia II, Das Röm isch-Byzantinische Ägypten. Akten des internationalen Symposions 26.- 30. September 1978 in Trier: 175-186. Mainz. El-Fakharani, F., 1991, The Kibotos of Alexandria. Studi Miscellanei 28: 21-28. Blue, L., & Ramses, S., 2005, Lake Mareotis Research Project – Preliminary Report from Pilot Survey, August 2004. Newsletter of the Archaeological Society of Alexandria 15 ( February) : 5-16. Blue, L., & Khalil, E. (eds.), forthcoming, A multidisciplinary approach to Alexandria’s economic past: The Mareotis case study. Oxford. De Cosson, A., 1935, Mareotis: Being a Short Account of the History and Ancient Monuments of the NorthWestern Desert of Egypt and Lake Mareotis. London. Empereur, J.-Y. (ed.), 1998, Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du collo-

33

LAKE MAREOTIS: RECONSTRUCTING THE PAST

34

D. HOPKINSON: MAREOTIS ISLAND SURVEY

The Results of a Preliminary Survey at Mareotis Island Dylan Hopkinson

Introduction

northern and southern limits of the western arm of Lake Mareotis and form a natural flooded depression known as the Mallahet Maryut Depression (Fig. 1). The island measures 3.7 km long by an average of 400 m (maximum 680 m) wide and is connected to the mainland by an area of low-lying marshy ground to the southwest at the foot of the Gebel Mariout Ridge (Fig. 2). The lands around Lake Mareotis formed an important agricultural production centre in antiquity known to have been richly populated with farming villages, market-towns and villa estates (Strabo G eography 17.1), and would have been dependent on seasonal rainfall and ile floodwaters for their irrigation and fresh drinking water.

Location The site known as Mareotis Island lies 65 km southwest of Alexandria within the western arm of Lake Mareotis, closely associated with its southern shore. It lies 4 km to the west of the site of Marea/Philoxenité, and directly north of the modern town of Bahig (Fig. 1). The island was subject to systematic survey in 2007 and 2008 as part of the broader Lake Mareotis Research Project (see Blue this volume). A large number of archaeological features are clearly visible on the ground which date to the Greco-Roman periods largely indicated by the substantial quantities of ceramics present in surface scatters. This paper presents the overall character of the archaeological features observed, and some preliminary interpretation with specific reference to the littoral zone on the northern coastal plains where ‘maritime’ and lakeside activities were largely focused.

Environmental Background In antiquity Lake Mareotis provided a transport connection between Alexandria and its agricultural and industrial hinterland in the Mareotic region; and via the Nile waterway to the Egyptian interior along the Nile Valley. Strabo states that the lake is filled by many canals from the ile, both from above and on the sides” (G eography 17.1.7). It was these canals that provided the fresh water which came

The island is formed from the remnants of a small limestone ridge which lies between the two much larger Abusir and Gebel Mariout Ridges. These two ridges define the

Fig. he location of Mareotis Island and the sites mentioned in te t modified from e Cosson

35

.

LAKE MAREOTIS: RECONSTRUCTING THE PAST Fig.2: Schematic of the main features of Mareotis Island ( Hopkinson) .

from the Canopic Branch of the Nile apparently keeping the lake water sweet, and filling the cisterns of Alexandria. The direct link to the ile during the seasonal floods is also likely to have had a significant impact on lake levels. The exact location and number of these connections is not firmly understood, but it is likely that the ancient courses would have mirrored those of the more recent canals (Warne & Stanley 1993: 15-16).

of sites located on the southern shore of the lake, where Roman foundations are observed to have been buried under 3 m of deposits (Khalil 2005: 39). Holocene sedimentation rates between 0.02 and 0.07 cm per year are indicated by Smithsonian boreholes either side of the island (Warne & Stanley 1993: 50). This sedimentation is likely to have obscured archaeological features over time, and to have raised the level of the lake bed while consequently reducing the depth of the lake waters (Flaux 2008: 9-10, III.2).

There is sufficient environmental evidence to show that the climate of the region has not changed significantly in the past two thousand years (Ball 1942: 20; Brooks 1949: 333). Evidence for this includes an analysis of a weather diary kept in Alexandria by Ptolemy in the 2nd century, which concludes that “the number of rainy days per annum was much the same as it is nowadays, though the rainfall was more evenly spread out over the year” (Murray 1951). So we can conclude that the climate was semi arid as it is today. Today rainfall in the Western Coastal Desert region is greatest along the margin by the sea and is typified by its great variability from location to location and from year to year, which suggests that in antiquity even this fertile and productive region would have had a precarious existence were it not for the fresh waters of the lake and the reliable annual ile floodwaters.

Modifications to the Lake since the G reco-Roman Period The water regime in the area has been greatly modified since antiquity, most notably by the silting of the connections between the lake and the Nile; when the Canopic Branch of the ile finally silted up around the 12th century the lake also began to dry up and shrink in size. Strabo also tells us that the extent of the lake must have been much greater in antiquity than it is today, and extended around 40 km to El Bordan to the southeast and to Kom Al Trouga 70 km to the southwest of Alexandria (De Cosson 1935: 26) (see Fig. 1). Warne & Stanley also support these dimensions which amount to around 700 km2 of land lying below the 0 m sea-level datum (Warne & Stanley 1993: 53).

Light north-westerly winds are generally observed in the region and these transport large quantities of clayey loess sediments from the degrading limestone ridges and silts and sand from the coastal plain. These silts are deposited in colluvial sheets at the base of the ridges and on north facing shores. This phenomenon has been noted previously (Warne & Stanley 1993: 53), and at a number

It is very likely that even without a direct connection to the lake the effects of the ile flood would still have been noticeable in the water levels due to groundwater seepage (A.G.,Brown, Univ. of Southampton, pers. com.). When the Aswan Dam was constructed in 1964 there was therefore a significant change to the water level in the ile 36

D. HOPKINSON: MAREOTIS ISLAND SURVEY Delta aquifers and the lake. As a result the hydrodynamic environment today is very different from antiquity and we have no direct way of understanding the range of water levels that would have been experienced during the floods. As Ramses & Omar (in this volume) indicate, modern infrastructure, fisheries and agricultural drainage ditches have further modified the lake, splitting it into several artificial basins and remodelling large stretches of coastline. These basins have water regimes that are artificially managed for modern needs, and large amounts of water are pumped into the sea to keep the waters artificially low for land reclamation projects.

this sakia to the cisterns of the town. To the south of this are the remains of two very interesting pottery kilns with a large heap of broken pottery thrown there as it was ‘scrapped’ from them. North of the sakia there is a long jetty running into the Lake from which ferry-boats once plied.” (De Cosson 1935: 130)

The reference here to waterwheels (‘sakia or ‘saqiya’), cisterns, ceramic mounds from pottery production and lakeside features characterise the island in general terms, but the description bears little direct relation to the position of the remains observed today. Consequently the archaeology of Mareotis Island represents a significant untapped resource that can enhance our understanding of the region.

Today the lake is fed only by water seepage from the sea, precipitation, and by irrigation ditches that channel agricultural waste-water from the Beheirah Governorate. As a result of all of these changes, it has been calculated that the current lake level is only 13% of the ancient lake surface area (Warne & Stanley 1993). Observations made during the 1970s show that the western arm of the lake was totally dry at least from Marea/Philoxinite and to the west. However, a recent geoarchaeological survey as a component of the Lake Mareotis Research Project indicates that the depositional environment close to the modern shoreline had always been lacustrine except during the recent centuries (Flaux forthcoming).

Fieldwork In August 2007 under the auspices of the Lake Mareotis Research Project, a six-week survey was undertaken to plot the visible archaeological remains on the island using RTK GPS and Total Station surveys. In order to provide an understanding of the dating of features a controlled surface ceramic collection strategy was adopted. An RTK GPS topographic model was also completed with 5 m transects across the island. In addition to this, a geoarchaeological auger survey was undertaken in 2008 to help contextualise the results and to begin to understand how site formation processes can inform our understanding of the site. Of particular interest here was the relationship between the ancient water levels and the archaeological features, and the impact of the changing lake levels on the development of the sites. A more detailed account of the Lake Mareotis Research Project and its methodologies is presented in this volume (see Blue this volume, see also Khalil and Ramses & Omar this volume)

Deltaic subsidence and tectonic shifts may also have changed the topography of the region as it is known to have done within the Nile Delta region further to the east; however it has not proved possible to use existing geoarchaeological datasets to understand the significance of these events. Consequently as a result of the various geographical factors described, we must accept that the current lake level cannot be used as a reference datum from which to reconstruct ancient lake levels and to assess the impact of the annual floods and the summer evaporation on the shoreline. However, understanding these relationships is a fundamental factor to our interpretation of the archaeological structures lying near the shore.

The majority of archaeological structures are located on high ground, occupying the central limestone ridge that forms a series of hills that extends as a discontinuous spine along the length of the island. Two main settlements were identified concentrated at the eastern and western extremes of the island, with further activities evident on the north facing coastal plains below the central ridge. The central section of the island is some 2 km long and is much less densely occupied than the western and eastern extremities. The activities here represent industrial amphorae production and possible agricultural activities.

Previous Research There has been a significant increase in our understanding of the archaeology of the Mareotic region in recent times brought about by the work of El-Fakharani (1974, 1983, 1984), Rodziewicz (1983, 1990, 1998a, 1998b, 2002), and Empereur (1986, 1998; Empereur & Picon 1986, 1998, 1992), however previous research on the island itself appears to have been negligible. There are numerous cartographic representations of the region, two of the most informative being the one on Sheet 37 of the Atlas G éographique in the escription de l gypte from 1809, and on Carte des en irons d le andrie produced in 1866 by Mahmoud Bey El-Falaki; however, the detail pertaining to the island in both cases is limited and difficult to reliably geo-reference. A further cryptic description of the archaeology on the island is given by De Cosson (1935):

C hronology The ceramics recovered from the island indicate that the settlements was well established some time during the Hellenistic period and that occupation potentially continued through to the 7th century AD. The western settlement appears to have been continually occupied as indicated by ceramic forms recovered, whereas the eastern settlement seems to have suffered a decline in activity in the Early Roman period which reached its nadir some time in the 4th to 5th centuries with a late recovery in the 6th century. The results of the ceramic survey are not commented on in detail here, but will be published shortly (Tomber & Thomas forthcoming).

“Many buildings can be traced, and at the eastern end the circular stone platform and oblong well of an ancient sakia will be found. Stone channels are traceable leading from

37

LAKE MAREOTIS: RECONSTRUCTING THE PAST Archaeology of the Eastern Settlement The eastern settlement (Fig. 3) is concentrated on the high ground of the limestone ridge which measures approximately 650 m by 350 m; the majority of this area is covered with building remains of various construction styles and materials. The layout and construction styles of the visible building remains indicate some degree of social organisation; such as possible administrative and commercial areas, some are grouped around what appear to be large, open areas at a number of locations on top of the ridge. A long boundary wall is located winding along the top of the ridge in the north-western limit of the settlement; this wall measures 240 x 1.20 m wide, and may have served a defensive function protecting the settlement from the west. To the west of the settlement there is a small low-lying depression that could well have formed a natural inlet prior to silting; around its northern and southern margins the remains of buildings that may have been possible wharf structures have been identified. This inlet appears to have served as an inland harbour which would have provided sheltered mooring. Auger cores taken during the 2008 season confirm that the sediments in this depression were indeed deposited in lacustrine conditions in antiquity (Flaux forthcoming).

are difficult to interpret in detail because there are several phases of activity. The relative dating of these features has not yet been possible. These structures appear to fall into several categories of linear features, walls and lakeside buildings. Linear Features ( G roup A) The first category of structures is a series of five linear features located on the low-lying northern coastal plain that extends between the present northern shoreline of the island and the foot of the limestone ridge (Fig. 3, Group A). These features range between 65 to 120 m in length, and appear as low, roughly built earthen ridges up to 4 m wide which are spaced roughly between 30 to 100 m apart. At the waters edge, additional features can be seen that extend along the current shoreline. As yet, these water front features have an undefined chronological relationship to the linear features (Group A) and represent primary and secondary activities over a period of time. The landscape between the base of the limestone ridge and the lakeside features appears to be largely devoid of additional structures and the linear features are relatively isolated with only a few abutting or adjacent features. It has already been established that little is known about the absolute water-levels of the lake in antiquity, and unfortunately the 2008 geoarchaeological survey did not target the area around the linear features therefore our interpretations

To the north of the eastern settlement, along the northern coastal plain of the island, between the lake and the foot of the ridge, are a series of structures and buildings which

Fig. he littoral features mentioned in this paper and topographic sur ey of the eastern settlement of the island ( Hopkinson) .

38

D. HOPKINSON: MAREOTIS ISLAND SURVEY of the exact nature and function of these features remains speculative, however two possibilities present themselves. Firstly, they may represent crude earthen jetties projecting from the ridge into the lake during periods of higher water-levels. The other possibility is that the coastal plain was dry in antiquity and that the earthen ridges may have divided the coastal land into units, potentially for cultivation as is suggested by the interpretation of similar features identified in the western settlement. We know from ancient sources that the Nile had a dramatic effect on the Delta region; Strabo states “at the rising of the Nile the whole country is under water and becomes a lake, except the settlements; and these are situated on natural hills or on artificial mounds, and contain cities of considerable size and villages, which, when viewed from afar, resemble islands” (G eography 17.1.4). It is therefore possible that the annual variation in the level of the lake level was large enough for both scenarios to be valid.

‘quay’ wall to demarcate the edge of the lake either as protection from inundation or to form a working platform, perhaps for the mooring and loading of boats. In some instances the construction of these sections of wall seem to be related to the linear earthen ridges of Group A, and it would appear that in some cases they were built to connect or at least respect, the linear features lakeside termini. The elevations recorded from various sections of these walls indicate that the surviving top course of the walls are not all at the same level but in fact exhibit a difference in height of over 1.83 m. This may indicate that either the top course of some of the sections is missing, or that these walls were built at different times and at different mean lake-levels rather than as a single planned ‘quay’ building effort. Coastal uildings Associated with the first group of linear features Group A) are a large number of walls indicating the presence of a number of buildings that had been constructed along the shoreline in this area (see Fig. 3). Since the relative dating of structures has yet to be determined it is not possible to discuss the phasing of the structures. However, what is apparent is that if the linear features were jetties and if they were built at the same period as the coastal buildings, then these buildings would have been susceptible to serious flooding. It is vital that the relationships and phasing of these features is investigated as it will have important implications with respect to interpretation of the landscape and the features located within it.

Linear Walls ( G roup B) A second group of linear walls were observed to the west of Group A along the north-western shores of the lake in the eastern settlement area. This group is located to the west of the wall that winds along the top of the western side of the limestone ridge, and down onto the coastline (see Fig. 3). These linear features were observed as low rubble walls 0.50 m thick, their northern lakeside ends are close to the current waterline and appeared to extend inland for distances of between 25 to 60 m. None of these walls reached the foot of the limestone ridge and in most cases extended less than half way across the plain. Three of these walls were spaced approximately 35 m apart with a further group 30 m to the east that was composed of three roughly parallel walls in very close proximity. The coastal ends of the first three walls had associated low rubble linear structures or platforms which extended for a short distance along the coastline.

Amongst these coastal buildings are a group of structures that deserve particular attention. Three isolated groups of buildings of similar construction were identified at three different areas of the plain (see Fig. 3; Group D). Two of these are located on the north coast and a third on the eastern extreme of the island with a south-easterly coastal aspect. The buildings appear to be arranged in symmetrical blocks facing each other and around 13 m apart, each measuring approximately 10 m wide and between 30 and 55 m in length, and were aligned perpendicular to the coast. The building pairs consist of single rows of cellular rooms measuring roughly 10 m by 5 m. Their location and alignment relative to the coast suggests that their function was somehow related to the lake and that they could have been warehouses, shops or boathouses.

The alignment of the main walls of Group B is interesting since they were constructed on the same orientation as the linear features discussed above (Group A). However, the Group B linear features are not perpendicular to the shoreline since this section of the coast turns to the south and is west facing. As a result these features are on an oblique alignment relative to the current shoreline. The stone construction, relative close spacing, alignment to the coast, and their location to the west of the long winding ‘boundary’ wall, suggest a distinct function as compared to the Group A linear features. One possibility is that these features may have been concerned with land reclamation or stabilisation, serving perhaps as a series of ‘groins’.

Summary Overall there appears to be a great deal of established activity in the eastern settlement. Buildings on the ridge settlement hint at developed social organisation and those on the northern coastline show a strong association with the lake. Features interpreted as an inland harbour, lakeside walls, and possible jetties, all suggest an established interaction with the lake. There are a far greater number of buildings on the northern coastline as compared to the south coast in this part of the island, which may suggest that the southern coast was possibly dry or marshy and therefore inaccessible to boats in antiquity.

shlar Lake alls Group C Large sections of dressed ashlar blocks are observed forming stretches of wall along the northern edge of the plain (see Fig 3). These walls closely respect the natural topography of the current shoreline and appear to have originally been laid as stretchers extending parallel to the shore. This might suggest that the walls were built as a sort of 39

LAKE MAREOTIS: RECONSTRUCTING THE PAST Archaeology of the Western Settlement The western settlement occupies a similar location to the eastern settlement, situated on the high ground of the limestone ridge which measures 820 by 200 m (Fig. 4). The area exhibits the remains of various buildings, with a mixed function suggested by the construction materials and surface finds. verall the preservation of these structures is not good; a large area in the extreme west has been heavily damaged by earthmoving machines, and in other instances only patchy areas of construction and dislocated wall lines survive. As a result, it is not possible to identify any zoning of functionality or social organisation, although a few significant groups of structures were observed. As in the eastern settlement the northern coastal plain appears to be the focus of activities. Once again it is the lakeside features that are the focus of interest in this paper although some of the structures on the ridge will be discussed.

Ashlar Shoreline Wall Alignment ( Feature E) The area along the waterline is densely overgrown with water reeds and has silted up making the water level very shallow and the ground marshy. Amongst the reeds can be seen the upper surface of sections of a long straight wall (see Fig 4). All the shoreline features in the western settlement appear to respect this northern boundary. The wall is built from limestone ashlars laid as stretchers with a flush north-facing aspect, suggesting that it was laid in a specific relation to the lake. The various sections of this feature appear to suggest a continuous wall with overall dimensions 0.40 m by as much as 245 m in length. No excavations of this feature were undertaken so it is not possible to indicate the number of courses or the height of the wall. The location and construction of this wall suggest that it has a significant relationship to the shoreline if water levels similar to those seen today are assumed. It appears to be a lakeside wall protecting the land behind or creating a working platform from which to conduct activities related to the lake. Some of the other features close to this wall appear to butt against this wall suggesting that it may also form part of a larger pattern of structures.

The features of the coastal plain are sparsely distributed and appear to fall into three clear categories: a section of ashlar wall aligned along the shoreline; long, thin linear walls forming low ridges perpendicular to the coastline; and localised masonry features typified by curvelinear walls and associated ashlar lined pits, there is an overall lack of secondary activities such as the coastal buildings on the eastern coastal plain, which makes interpretation somewhat easier.

Linear Walls ( G roup F) The second group of features located on the coastal plain of the western settlement, is a series of roughly paral-

Fig.4: The littoral features mentioned in this paper and topographic survey of the western settlement of the island ( Hopkinson) .

40

D. HOPKINSON: MAREOTIS ISLAND SURVEY Schi ler identifies 14 different types of saqiya from Roman, Islamic, and modern contexts (Schiøl er 1973: 13), including wheels with qudus pots attached directly to the rim, and wheels which powered a ‘pot-garland’ or bucket chain. In the case of the features associated with the western settlement it seems that the curved walls represent part of the animals’ circular track, while the ashlar lined pit is likely to be associated with either the wheel or the gear mechanism.

lel linear features aligned perpendicularly to the current coastline; they are constructed from faced ashlar blocks that in some cases have degraded to stone rubble and in some areas only survive as low earthen ridges. Seven potential features were observed in this group (see Fig. 4), all of which have some connection with the current coastline and appear to respect the alignment of the possible lakeside wall. These linear features extend back from the coast, south towards the main ridge; their lengths vary between 25 and 100 m but on average only reach halfway across the plain and appear to show no direct relationship with the ridge itself. Two of these walls are connected by a linking wall at their southern ends.

Many sherds of qudus pots were observed along the northern flank of the limestone ridge adjacent to the plain but were not recorded anywhere else on the island; the location directly adjacent to the water is also appropriate for this interpretation of water management, and places the northern coastal plain in an agricultural setting.

These walls have much in common with similar features at the eastern end of the island, they have the same orientation relative to the coastline and occupy the ground between the coast and the ridge, with their coastal terminus closely associated with sections of ashlar wall aligned along the coast. This suggests that the walls could have potentially functioned as jetties when lake waters were high or as boundary markers potentially prescribing plots of land along the coast that butt against the ashlar wall.

In the light of this interpretation it is possible to reinterpret the long linear walls between the ridge and the quay wall; these may potentially be seen as boundaries demarking areas of farmed land irrigated by saqiyas. Oleson concludes that rudimentary saqiyas from modern contexts could usefully irrigate similar portions of land to those possibly represented by the area of land contained within the linear walls (Oleson 1984: 369).

Cur elinear alls and shlar Lined Pits Features G The land between the linear walls had a sparse distribution of ashlar constructed features that form the third group of structures. They are represented by sections of curved wall in two locations where additional associated wall lines were observed. These curved walls are incomplete but appear to represent rounded features with projected diameters of about 10 m (see Fig. 4).

Linear Mole Extending into the Lake towards a Small Island At the north-eastern extent of the western settlement in the marshy ground of the shoreline plain is a site that is unique on the island see ig. 2 . This is significant because it is the only ancient feature on Mareotis Island that currently extends into the lake, and as such it represents a useful indicator of lake levels in antiquity. The feature is a 250 m long linear mole construction that extends out into the lake to join with a small island that measures roughly 30 by 15 m. This small island supports a rectangular building measuring 20 x 10 m that was built from substantial masonry blocks. Three jetties extended from the north of this building into the lake measuring between 7 and 15 m in length. The feature extends towards a promontory settlement on the opposite side of the lake where there is a similarly built reciprocal linear construction that extends for a short distance into the lake on the same alignment towards the small island. These features imply a functional relationship with the lake perhaps for controlling the movement of transport along the lake as well as between the two settlements.

The interior structure of one of these rounded features is preserved and takes the form of ashlar built components lining a rectangular slot which measured 4.80 by 0.90 m. This feature, although unexcavated and partially observed, bears a significant similarity to ancient water-lifting devices known as saqiya which have been observed archaeologically from broadly contemporary Greco-Roman sites in the region such as Abu Mina located some 15 km south of the island (Schiøl er 1973: 129-136; Oleson 2000: 263; see Fig. 1), and Tuna Al-Gabal near Hermopolis in the Nile Valley (Schiøl er 1973: 141-148). There are further saqiya features in the Mareotic region which have not been published in detail; examples in the immediate vicinity of the island are described as having a “circular stone platform and oblong well” (De Cosson 1935: 130).

Tower-like Feature ( Feature H) A final feature in the western settlement is worthy of mention, this is a small artificial mound of collapsed building material some 4 m in height which lies just northeast of the main ridge on which the settlement is located. After cleaning the flat top of the mound, masonry elements were observed that formed a structure similar to the ashlar lined pits mentioned above. Rather than having a single pit slot this feature had a pit that was divided into two separate apertures with overall dimensions very close to the saqiya discussed above.

Waterwheels were in common use in Egypt from Hellenistic through to medieval times, and are simple devices using a gear mechanism to convert the horizontal movement of an animal walking round a circular track into vertical movement for lifting water. The water was held in pots called ‘ udus which are connected to the wheel. Saqiya are used predominantly in agricultural contexts (Oleson 1984: 126) but have also been recorded in the bath-house buildings, as at Abu Mina (see Fig. 1). Waterwheels are also known to occur in many configurations 41

LAKE MAREOTIS: RECONSTRUCTING THE PAST Fig. Longitudinal and transverse schematic sections through the ater to er on Mareotis Island ( Hopkinson,af ter McCann et al. .

top of the tower and a further 1.5 m to the vat on top of ridge, an overall height in the order of some 5.5 m, the additional height of about 1.5 m being accounted for by the assumed superstructure which held the qudus pot-garland above the surviving tower.

It is possible that this feature is the top of a tower used to lift water from a well known to exist close to the bottom of the mound using a pot-garland passing through each aperture. An elevated ‘bridge’ constructed from earth and masonry blocks connects the tower to the main ridge at a point where a plaster lined vat is located as part of a larger building complex that could potentially represent a public bathhouse or cistern and water management feature (Fig. 5). An example of such a tower from the late 1st century AD is known from the port of Cosa in Italy (Oleson 2000: 258). The so called ‘spring house’ at Cosa initially lifted water to a height of 3 m but it burnt down and was replaced with a structure that lifted water over a height of 10 m to an elevated conduit. If this hypothesis is correct the structure on the mound would have lifted water to about 4 m to the

Summary The western settlement is generally poorly preserved, however there is an apparent lack of secondary activities on the northern plain which makes it easier to interpret the features that survive. The presence of water lifting devices and possible plots of agricultural land suggest that the majority of the coastal plain was above the mean waterlevel in antiquity. As such an alternative interpretation of the linear walls as possible jetties is not so well supported, 42

D. HOPKINSON: MAREOTIS ISLAND SURVEY although they could have served this function during the flood season. The absence of further buildings on the plain may suggest that this community did not have a strong association with the lake for their subsistence; however the presence of the promontory towards the small island may suggest that at some point the settlement had a significant role in administering activities on the lake.

plications for the interpretation of the other ‘coastal’ walls on the island. To the south of the central ridge of the island, there are two adjacent ‘plots’ of land which are enclosed by low walls of faced stone and rubble (see Fig. 2). These features resemble ‘karum or vineyard plots; features that are usually found inland in a 15 km wide band south of the Gebel Mariout Ridge, as indicated on Survey of Egypt maps from 1927. K arum are features that are poorly researched but the author believes they were used to raise the depth of topsoil on agricultural plots with the advantage of increasing the space for healthy root growth and for storing rainwater in the soil reservoir (Hopkinson 2007). There seems then to have been some small scale industry and subsistence agriculture in the middle of the island that is not directly related to either of the main settlements.

Archaeology of the C entral Island The land between the eastern and western settlements on the island is roughly 2 km long and appears to have been sparsely utilised in antiquity for industrial and agricultural activities. There are four sites of large scale pottery manufacture (see Fig. 2; Kiln “A” & Kiln “B”), each with extensive associated storage rooms, and large mounds of ceramic sherds wasted during the firing process. It seems that these potteries, like many identified near the coast on the southern mainland, were exclusively producing amphorae forms known to have been used for making and packaging the large quantities of wine for which the region is renowned (Empereur & Piccon 1986). Two of the kilns originate in the Hellenistic period (see Fig. 2; Kiln “A”), one close to both the eastern and western settlements. These sites continued in use through to the mid-1st century AD, when amphorae production switched to two adjacent sites (see Fig. 2; Kiln “B”). Production of amphorae at the new sites continued from the late 1st to the 5th centuries AD.

Water Leve ls and C onnection to the Mainland One important consideration for the interpretation of the features on the island is the relationship between the island and the mainland to the south, and the local hydrological environment throughout its occupation. The island is currently largely surrounded by water but it should not be assumed that this was the situation in antiquity because modern irrigation and dyke formation have modified the movement of water. A number of the features that have been identified on the island may suggest a greater connectivity to the mainland to the south. All of the potential maritime focused features such as the possible jetties, quays and lake walls are located on the north facing shores of the island, and the absence of these features on the southern shores suggests the absence of water in this area that could be usefully exploited. Moreover the features that are observed along the current southern shore appear to be associated with dry land activities. The ‘karum plots are agricultural features thought to be concerned with collecting and maximizing the use of precipitation and as such suggest that there is no alternative readily available water resource in close proximity. The second feature that appears to preclude the existence of a significant southerly body of water in antiquity is the large kiln structure observed in the west of the island (see Fig. 4; ‘kiln site’). These kilns are known to have significant subterranean fire chambers up to 4 m deep where the fire was lit, with a perforated firing plate above to allow the hot air to circulate (El-Fakharani 1983 . However, the firing plate of this kiln is currently only a short height above the lake waters to the south allowing little room for the firing chamber and indicating that it could only have been used if the land to the south between the island and the mainland had been dry. The inference that there was connectivity between the island and the southern shores of the lake, is supported by what we know about the organisation of wineries of the region, consisting of three essential elements: the kiln, the winery, and the estate villa (Empereur & Picon 1992). At least three winery structures are known in the immediate vicinity of the island on the southern mainland, and no structures suitable for the production of wine have been identified on the island itself. It is possible to suggest therefore that

Further evidence of light industry on the island is indicated by two small quarrying sites on the limestone ridge (see Fig. 2), although their capacity is very small compared to the quantity of stone used in the settlements and it is more likely that the majority of the building stone was transported across the lake from the large quarry sites known to exist in antiquity on the Abusir Ridge to the north (Oliver & De Cosson 1938: 169-170). In the middle of the island, associated with the kiln sites, there are a group of structures that occupy the top of the ridge that could have been workshops or accommodation for the workmen involved in these industries. A number of features from these structures extend down towards the northern coastline, although on the plain below there is only one visible structure. This is a 75 m long section of ashlars laid as stretchers aligned with the current coastline (see Fig. 2). The wall lies 70 m inland from the current coastline, which is surprising since the flush face of the wall at first suggests a coastal function similar to the sections of lake wall or quay found associated with the eastern and western settlements. However, if this was its function then the land directly north of the wall equates to a large area of siltation and aeolian soil deposition. Small scale excavations either side of the wall in 2008, however, indicated that there was only one course of stones surviving and that the soils surrounding the feature appeared to display similar sediment regimes and were not deposited underwater. It therefore seems that this section of wall had no direct association with the lake, and its purpose is unclear. This finding is also significant since the construction of a flush face apparently respecting the coastline has im43

LAKE MAREOTIS: RECONSTRUCTING THE PAST the kilns on the island were producing amphorae for the wineries on the southern mainland. However, the recent geomorphological enquiry undertaken as part of the Lake Mareotis Research Project has confirmed that the body of water that currently extends from the southern shores of the island was in fact waterlogged in antiquity (Flaux forthcoming). The island ridge was an extension of the Marea Ridge and thus enclosed a body of shallow water to the south. How continually waterlogged this area was in antiquity has yet to be determined but the geomorphological survey has indicated that there was continuous lake sedimentation between the island and the southern shores since antiquity. Thus, it is assumed that access from this area into the main body of the lake was restricted by the ridge, thus deterring settlement on the southern shores of the island, as witnessed by the lack of archaeological sites in this region. It is clear however that the situation is very complicated and that our understanding of the archaeology of the island is incomplete.

El-Fakharani, F.A., 1983, Recent Excavation at Marea in Egypt. In G. Grimm, H. Heinen & E. Winter (eds.), egyptiaca re ernsia II as R misch y an tinische Ägypten. Akten des internationalen Symposions . . eptember in rier, 175-186. Mainz. El-Fakharani, F., 1984, The Kibotos of Alexandria. Studi Miscellanei 28: 23-28. El-Falaki, Mahmoud Bey, 1872, Memoire sur l anti ue le andrie ses faubourgs ses en irons. Copenhagen. Empereur, .- ., 1986, n atelier de dressel 2-4 en gypt au IIIe siécle’ de notre ére’. In J.-Y. Empereur (ed.), Commerce et artisanat dans l le andrie hell n isti ue et romaine. ctes du collo ue d th nes organis par le C R le Laboratoire de c ram ologie de Lyon et l Ecole fran aise d th nes d cembre 599-608. Athens, Paris. Empereur, J.-Y., 1998, Alexandria Rediscovered. New York. Empereur, J.-Y., & Picon, M., 1986, A La Recherche des Fours d’Amphores. In J-Y. Empereur & Y. Garlan (eds.), Recherches Sur Les Amphores G recques. Actes du collo ue international organis par le Centre national de la recherche scientifi ue l ni ersit de Rennes II et l cole fran aise d th nes th nes eptembre 103-126. Athens, Paris. Empereur, J.-Y., & Picon, M., 1998, Les atelier d’amphores du Lac Mariout. In J.-Y. Empereur (ed.), Com merce et artisanat dans l le andrie hell nisti ue et romaine. ctes du collo ue d th nes organis par le C R le Laboratoire de c ramologie de Lyon et l Ecole fran aise d th nes d cembre 75-88. Athens, Paris. Empereur, J.-Y., & Picon, M., 1992, La Reconnaissance des Productions des Ateliers Cérmiques: L’ Exemple de La Maréotide. E trait des Cahiers de la C rami ue gyptienn 3: 145-152. Flaux, C., forthcoming, Geomorphological survey of the south-west shores of Lake Mareotis. In L. Blue & E. Khalil (eds.), A Multidisciplinary Approach to Alexandria s Economic Past the Mareotis Case tudy. Oxford. Hopkinson, D.J., 2007, An Assessment of G reco-Roman Lakeside Agriculture and Trade in the Mareotic Region of Egypt. Unpub. MA diss., University of Southampton. Khalil, E.K.H., 2005, Egypt and the Roman maritime trade: a focus on Alexandria. Unpub. PhD thesis, University of Southampton. McCann, A.M., Bourgeois, J., Gazda, E.K., Oleson, J.P., & Will, E.L., 1987, The Roman Port and Fishery of Cosa a Center of ncient rade. Princeton. Murray, G.W., 1951, The Egyptian Climate: An Historical Outline. The G eographical J ournal 117.4: 422-434. Oleson, J.P., 1984, G reek and Roman Mechanical WaterLifting e ices he istory of a echnology. London. Oleson, J.P., 2000, Water-Lifting. In Ö . Wikander (ed.), andbook of ancient ater technology, 217-302. Leiden

C onclusion The archaeology of Mareotis Island consists of two settlements which appear to be concerned with agricultural, industrial and water management activities as well as lacustrine focused features which may indicate that fishing activities or lake transportation services were conducted from the island. The island represents a largely unexplored archaeological asset that has the potential to give insights into daily life in a non elite community; when looked at in conjunction with other known sites in the area it may provide an opportunity to begin to address settlement distributions and social organisation and make comparison between the lake region and those communities on the Delta, and perhaps archaeological findings from the ayum or Nile Valley, where the most studied examples of agricultural settlements are known. Such comparisons with sites from the Egyptian interior may prove to be less than representative of ordinary Egyptian practices where the majority of the population was located on the Nile Delta (Bagnall 2007: 227). There is a great potential for future research on Mareotis Island and the broader Mareotis region and these provisional findings only begin to scratch the surface of what the island and Mareotic region in general has to offer. Bibliography Bagnall, R.S., 2001, Archaeological Work on Hellenistic and Roman Egypt, 1995-2000. American J ournal of Archaeology 105. 2: 227-243. Ball, J., 1942, Egypt in the Classical Geographers. Cairo. Brooks, C.E.P., 1949, Climate through the ages. New York. De Cosson, A., 1935, Mareotis: Being a Short Account of the History and Ancient Monuments of the Northestern esert of Egypt and of Lake Mareotis. London. El-Fakharani, F., 1974, The “Lighthouse” of Abusir in Egypt. ar ard tudies in Classical Philology 78: 257272. 44

D. HOPKINSON: MAREOTIS ISLAND SURVEY le Laboratoire de c ramologie de Lyon et l Ecole fran aise d th nes d cembre 27-36. Athens, Paris. Rodziewicz, M.D., 2002, Mareotic Harbours. In C. Déco bert (ed.), Alexandrie Médiévale 2. Etudes Alexandrines 8: 1-22. Schiøl er, T., 1973, Roman and Islamic Water-lifting Wheels. Denmark. Strabo, G eography, H.L. Jones (transl.), 1949. London. Survey of Egypt (27/135) 1:100,000 Map; Sheet 88/48 “El Ghayata”. Survey of Egypt/Ministry of Finances, Cairo. Tomber, R., & Thomas, R.I., forthcoming, Pottery from the Lake Mareotis Research Project. In L. Blue & E. Khalil (eds.), A Multidisciplinary Approach to Alexandria s Economic Past the Mareotis Case tudy. Oxford. Warne, A.G., & Stanley, D.J., 1993, Late Quaternary Evolution of the Northwest Nile Delta and Adjacent Coast in the Alexandria Region, Egypt. Journal of Coastal Research 9.1: 26-64.

Oliver, F.W., & De Cosson, A., 1938, Note on the Taenia Ridge – with especial reference to quarries, sites and an ancient road between Alexandria and Abu Sir. ulletin de la oci t Royale d rch ologie d le andrie 32: 163-175. Rodziewicz, M.D., 1983, Alexandria and District of Mareotis. Graeco rabica 2: 199-216. Rodziewicz, M.D., 1990, Taenia and Mareotis: Archaeological Research West of Alexandria. Annual of the Egyptian Society of G reek and Roman Studies 1: 62-78. Rodziewicz, M.D., 1998a, From Alexandria to the West by Land and by Waterways. In J.-Y. Empereur (ed.), Commerce et artisanat dans l le andrie hell nisti ue et romaine. ctes du collo ue d th nes organis par le C R le Laboratoire de c ramologie de Lyon et l Ecole fran aise d th nes d cem bre 93-103. Athens, Paris. Rodziewicz, M.D., 1998b, Classification of Wineries from Mareotis. In J.-Y. Empereur (ed.), Commerce et artisanat dans l le andrie hell nisti ue et romaine. ctes du collo ue d th nes organis par le C R

45

LAKE MAREOTIS: RECONSTRUCTING THE PAST

46

M. HAGGAG: EXCAVATIONS AT MAREA/PHILOXENITÉ

The City of Marea/Philoxenit . Reflections on the Alexandria University Excavations, 1977-1981 Mona Haggag

In 1977 Alexandria University patronized an expedition under the direction of the late Prof. Fawzi El-Fakharani to search for the ancient city of Marea, mentioned by different classical authors as the capital of the Mareotic Nome during the late Pharaonic period (Thucydides I.68; Herodotus II.30; III.12; 15; IV.161; Diodorus I.68.).

Sadek concluded that all the remains identified were not earlier than the 4th-5th centuries AD (Sadek 1978: 67-80). All trial trenches proved to have remains of stone buildings dating to the Byzantine period. Pot sherds were entirely of Byzantine manufacture (Shahin 1983). The same is true for the architectural material, construction methods and techniques. These results raised a major question about the location of the pre-Byzantine town mentioned in classical literature (Thucydides I.68; Herodotus II.30; III.12; 15; IV. 161; Diodorus I.68). . The Pre-Byzantine Remains (Map 1) By exploring the area some 5 km west of the Byzantine remains, earlier remains of Marea proved to exist beyond the main perimeter of the Byzantine site. A group of harbour installations were found on an island to the west.1

Mahmoud Bey El-Falaki located the site of Marea some 45 km west of Alexandria on the southern shores of the western extremity of Lake Mareotis near El-Hawaria (ElFalaki 1872: 85, 86, 91, 96, 97, 101; see other Haggag chapter in this volume . El- alaki based his identification of the site according to its location on the map of Ptolemy the Geographer IV,5. 16.17 . This location is confirmed by Ball (1942: 25-30), Breccia (1922: 337), De Cosson (1935: 131-135), Rowe (1954: 128-145) and Fraser (1972: 143-146) due to four visible jetties which still project into the lake today. The jetties are associated with a broad ridge of hard limestone extending from east to west across the town parallel to the southern coast of the lake and bordered, from both north and south, by calcarenite land.

The ‘ Pr’ -Shaped Harbour (Fig. 1, Map 1) The main harbour of the island is a three sided structure projecting into the lake basin about 30.5 m wide and 57 m long. It is built of undressed blocks of limestone of Cyclopean size. A number of the upper blocks in this structure have drilled cylindrical holes which were either bollard holes or mooring rings (Fig. 2). According to the traditions of ancient Egyptian harbour architecture, ports of rectangular plan in the shape of the sign “pr” were designed for harbours facing the open seas. Earlier examples existed in the seaport of Alexandria (Jondet 1916: 13-20; El-Fakharani 1991: 25), the most conspicuous of which is the one inside the Western Harbour mentioned by Strabo as the Kibotos (G eography, Xvii, I, 10), another one is associated with the Island of Antirrhodes which lies inside the Eastern Harbour of Alexandria discovered by Goddio (Goddio, et al. 1998: 21-29, pl. I). This feature consists of two rows

The Surve y As part of the 1977 campaign and in collaboration with the University of Guelph, Canada, Mohamed Sadek conducted a proto-magnetometer survey at the site, concentrating on the shoreline area from which the four jetties projected. Data was collected in lines parallel to the lakeshore from west to east in a grid of 800 m long and 100 m wide at intervals of 4 m. Soundings were taken in the area west of the first quay, and some trial trenches were investigated in three locations on the waterfront near the jetties. The trenches went deep into the bedrock which was found at intervals of 4.6-6.00 m below the present ground surface. Fig.1: K ibotos harbour structure on the south shore of Lake Mareotis ( Site 9– L ake Mareotis Research Proj ect) ( photo M.H aggag) .

1. It is worthy of notice here that Gauthier, referring to some hieroglyphic text, stated that the ancient city of Marea was located on an island in the lake (Gauthier 1925-1931: vol. III, 53-54).

47

LAKE MAREOTIS: RECONSTRUCTING THE PAST Map 1: Marea in Lake Mareotis ( V. Atef) .

Fig.2: K ibotos harbour structure,m ooring ring ( photo M.H aggag) .

Fig.3: Rock-cut K ibotos harbour structure at Ikingi Mariut ( photo M. Haggag) .

48

M. HAGGAG: EXCAVATIONS AT MAREA/PHILOXENITÉ of stakes or piles approximately parallel, running southwest to north-east, along the axis of the main branch of the island, that forms a rectangular shaped-harbour against the island’s shoreline. A third one, hewn in the rock at Ikingi Mariut, about 20 km west of Alexandria (Fig. 3), was identified by El- akharani2 (2002: 203-208).

to a dromos. Another similar dromos on the opposite side leads to a funerary chamber. The ceiling is carved to imitate wooden beams in the manner common to Pharaonic cemeteries. False doors were also carved in the walls with a shaft in front of each one leading to the burial chamber below. All tombs have been looted. The only remains were a few local pointed bottom jars which are likely to be Late Pharaonic in date (El-Fakharani 1983: 176-178).

The Fort (Map 1) To the south-west of the pr-shaped port remains of a substantial structure built on a monolithic platform about 21 m wide and 24 m long was uncovered. The structure has an L-shaped plan. The walls seem to have consisted of rough core masonry with outer casing walls. The blocks of the outer casing are carefully cut and fitted together in oblique joints without using any mortar except for a foliage layer of non reddish mortar used in some of the courses to facilitate the sliding of such gigantic sized blocks. These external walls are slanting inward at an angle of about 40 degrees, and are based on larger foundation blocks with squared edges. The building is approached from the south by means of a ramp of dressed and carefully fitted stone blocks (Haggag 1984: 277-280). The use of rough masonry for the core while the huge blocks of the facing are very carefully polished, the oblique joints of the blocks, the use of such a thin layer of mortar, the inward slanting of the outer walls as well as the general plan of the building, are clear indications to a Late Pharaonic date. The building is interpreted as the fort of Psammaticus III (525 BC), from which his son Inaros sailed with his army, aided by some Athenian troops (Thucydides I.104), against the Persians in 465 BC, in an unsuccessful attempt to liberate Egypt.

The Byzantine Harbours (Fig. 4, Maps 1 & 2) The focal point of the town to the east of the Pharaonic/ Ptolemaic remains is the Byzantine harbour area which extends 2 km along the lake shore with four well preserved jetties projecting into the lake at an average height of nearly 2 m above the lakebed. The area comprises three harbours marked by the jetties and a natural promontory namely the west port, the middle and the east port. Along the shore there is a stone built seawall occasionally interrupted by either steps or stone ramps leading directly down to the water. The wall extends some 500 m with an average width of 1 m. At the outer extremity of the western quay of the middle port, there are a number of circular foundations about 4 m in diameter. These features are assumed to have been the base of a light beacon or lighthouse that would have aided navigation at night. Some mooring features can be detected on the upper course of this jetty in the shape of cylindrical bollard holes (Petruso & Gabel 1983: 63). The East Port (Fig. 5, Map 2) Although the east port was completely buried under the sands of the lake which by the time of excavation had dried out, digging revealed that it differs from the other ports of Marea in both shape and function. It is located between the promontory and a north-easterly island which has its own landing place formed by a short artificial dyke to the north-east and a natural promontory to the northwest. This northern island is connected to another smaller one to its south-west. A third quay connects this island to the promontory forming a wide entrance to the eastern port, at least in its first stage of construction. In the second stage, the shape, size and function of the port seem to have been modified. Almost 100 m of the quay that extended from the eastern small island to the promontory were removed and an extension to this quay constructed to the south. It was located parallel and adjacent to the eastern shore of the promontory and was about 200 m in length. Thirty metres before reaching a third island to the south, the quay ends forming an opening between its southern extremity and this small island. This opening provides an exit or entrance to the east port by way of the southern body of the lake, which extended considerably inland at this point. By this modification, the very long quays of the east port became totally isolated from both the mainland and the promontory. Accordingly, the link between the anchored ships with their cargoes and the town was severed. This fact is very important as far as the function of the port is concerned.

The Ptolemaic Q uay (Map 1) To the east of the island, a stone built jetty about 104 m long extends into the lake. Near its southern end, the jetty inclines to the east at an angle of about 175 degrees to extend a further 20 m in length. It is constructed of local regularly sized rectangular blocks, arranged in alternate courses of headers and stretchers (Haggag 1984: 280-283). The fine workmanship, medium size nature of the blocks as well as the ashlar masonry construction style with its vertical joints, are indications of a Ptolemaic date for the quay. The C emetery (Map 1) To the southwest of the island on the limestone ridge, the team came upon a group of burials that have nothing in common with Byzantine tomb types. The cemetery consists of some anthropoid pit tombs. A square shaft tomb is carved into the limestone ridge to a depth of 6 m. Loculi with gabled ceiling are cut either side of the shaft. A nearby chamber tomb is composed of a sloping passage hewn into the rock that ends with six steps leading to an open court. An opening cut in one side of the court leads

2. For detailed discussion of this pr-shaped ports of Pharaonic date as well as examples, see Haggag 1984: 263 ff. El- akharani 1991: 25, fig. 1,2.

49

LAKE MAREOTIS: RECONSTRUCTING THE PAST Map 2: Marea’s public buildings on the lake shore ( V. Atef) .

Fig.4: Byzantine harbour of Marea ( photo M.H aggag) .

depth of 0.25 m below the pavement, a square drainage canal, coated with a thick layer of plaster and covered by a row of huge rectangular blocks, extends across the decumanus. This drains into a similar canal that extends under an unpaved side street and from south to north crossing the decumanus at a right angle. Thus, the drainage waters pour directly into the lake (El-Fakharani 1983: 178-179).

El-Fakharani’s interpretation of the function of the east port 1983: 181 is that it worked as a one way traffic harbour for transit navigation: so that seafaring ships carrying goods from different parts of the Mediterranean could enter the east port via its northern opening, their cargoes were to be unloaded on the long quays in order to be reloaded on board river boats to transport goods to the Nile and vice versa.

The Shops (Map 2) The arcade of the decumanus lines a group of 12 constructions (Fig. 7), with different ground plans. The buildings are located side by side with doors that opened out onto the sheltered arcade. They seem to have been built to function as shops with residential quarters for their owners. Remains of white plaster coating are still visible on some of the walls. An upper storey for each of the shops is confirmed by the existence of stairs at the rear of some of the shops and also in some of the back rooms (El-Fakharani 1983: 178-179; Soleiman 2004: 145-164).

The Public Buildings (Map 2) Many, if not most, of the public buildings of Byzantine Marea were mainly erected along the shoreline. Parallel to the lake shore in front of the harbour installations runs the main arcaded street of the city. This decumanus (Fig. 6) extends eastward until it reaches the natural promontory which forms the eastern part of the middle port. It is 10 m wide marked on its southern side by an arcade of which some of the bases of columns can still be seen in situ. The street is paved by big rectangular limestone blocks. At a 50

M. HAGGAG: EXCAVATIONS AT MAREA/PHILOXENITÉ The Basilica (Fig. 8, Map 2) To the west of the decumanus, at the intersection of the western side street, a huge complex of structures came to light during excavations (Sadek 1992: 549-554). The structures consist of two conjoined apsidal basilicas furnished with what seems to be public bathing facilities for sanitary purposes, similar to those discovered and interpreted by Grossman at St Menas (Grossmann 1986: 12-13; Haggag 1984: 284-289). The Bakery At the promontory which forms the eastern extremity of the middle port, a huge building with an area of about 700 m 2 was uncovered. A large granite mill inside one of the building’s multiple rooms indicates the relation of this building with grain grinding processes. Inside one of the rooms, a marble tile with a carved Greek cross was uncovered. In a compartment annexed to one of the rooms in this building, a clay jar containing a hoard of coins was also found. Courses of Byzantine burnt bricks are inserted in the stone masonry of the walls. The building is surrounded by an external wall supported with buttresses typical of Byzantine methods of construction (Moussa 2002: 479480). Fig.5: ( above) Plan of the eastern harbour of Marea, after El-Fakharani 197 ( thanks to El-Fakharani) .

Fig.6: Decumanus between the eastern and middle ports of Marea ( photo M.H aggag) .

Fig.7: Plan of shops off the main decumanus ( thanks to El-Fakharani) . 51

LAKE MAREOTIS: RECONSTRUCTING THE PAST Fig.8: Basilica,M area ( photo M.H aggag) .

Fig.9: Enigma Buliding, Marea ( photo M.H aggag) .

The Enigma Building (Fig. 9, Map 2) Next to the western edge of the promontory at its meeting point with the lakeshore along the middle port, an enigmatic building has been uncovered. The building stands completely in the water. It is composed of podia and walls leaving canals 0.75 m broad between them. Some of the canals run in parallel lines with the coast, while others run at right angles with it and the other canals. The entire building is sliding towards the water leaving the outer parts of the podia under water. The canals were studded with small artefacts. The finds, which were mostly intact, included nearly one thousand pieces of small bronze coins, about 200 jugs of red or yellowish clay, St Menas ampulae, other ampulae with crosses, small statuettes of horsemen, fishing hooks and broken glass bottles. This unusual plan made scholars differ in their interpretations of the building’s function. Some suggestions included its possible function as a dry-dock owing to its distinct slope into the lake water Sadek 1978: 70 , but the huge amount of finds inside the canals stands against this hypothesis. Another explanation is that it served as a fishery ikkos Lianos pers. comm.). While the discoverer tended to interpret it as a building for votive offerings in honour of both St Menas

and St Marcus, due to the large quantity of coins, ampulae and other artefacts that were found inside (El-Fakharani 1977: 16; Soleiman 2004: 121-138). Boston Field School In 1979, the Boston University’s Summer Term and Study Abroad Program supported an archaeological field school at the site of Marea. Their work lasted for three consecutive seasons, one month each year.3 In addition to the fruitful student training, the mission made new and important discoveries, the most important of which was the Diolkos of Marea. Along the shore between the middle and western jetties, the team of Boston University came upon remnants of a dockyard (Petruso & Gabel 1982: 12). It consists of two sloping walls, each is over 7 m wide and 20 m long,

3. The team was directed jointly by Profs. El-Fakharani & Gable in collaboration with Petruso and the architect, Boyd. The author of this article had the honour of participating in this field school first as an undergraduate trainée and then as an assistant archaeologist.

52

M. HAGGAG: EXCAVATIONS AT MAREA/PHILOXENITÉ Fig.10: Winery excavated by El-Fakharani 1983( photo M. Haggag) .

pletion of the excavation of the building, it proved to be a huge double peristyle that occupied an area of more than 1,500 m2. Both El-Fakharani (1983: 184-186) and Rodziewicz (1988a: 175-178, 1988b: 267-276) asserted that this discovery represents the first and largest peristyle built for residential purposes to be discovered in Egypt hitherto.

emerging from massive stone foundations going down to the bedrock, to a depth of about 2 m. The runners slip into a central aisle that lies between the jetties. This aisle has some sort of drainage facilities in the form of holes cut into the floor in order to drain water and keep the floor dry. The entire building slides down into the lake basin at a gradient of approximately 1:16 which allows for the use of manpower and log rollers to hall ships out of the water (Petruso & Gabel 1983: 76).

In the middle of the town, some Byzantine dwellings were uncovered. Each is composed of two rooms. Byzantine local pottery ware were scattered inside. Muslim burials were dug inside the room. Green glazed pottery sherds of the Islamic period were found as well as one piece of a Fatimid coin.

The C ity’ s Residential Area Far from the harbour area to the south of the limestone ridge and south of the modern highway leading to Borg ElArab, in the area now called “Hawariya”. During a second season conducted by the Alexandria University expedition team in 1977, various buildings were uncovered in this region which seem to relate to residential and daily life activities. The Peristyle C omplex (Map 1) Adjacent to the modern highway (Map 1), some 2.5 km to the south of the harbour, a peristyle court which leads to a variety of rooms, was uncovered. One of the rooms has a staircase leading to an upper floor of the building. It seems that the building went through two stages in its construction the first stage is indicated by a layer of white plaster coating on the façade. The second stage is marked by the addition of a massive thick walled façade with two windows which have tunnels underneath leading to watercourses that are connected to cisterns. A modern ditch for irrigation water that was dug through the rooms halted a complete investigation of the remains. In 1980, Rodziewicz4 replaced that ditch with a pipe in order to continue uncovering the rest of the building. On com-

Wineries (Map 1) About 200 m to the southwest of the peristyle building, in the middle of the town, a big elaborate winery (Fig. 10) was uncovered (El-Fakharani 1983: 182-184). Another identical in layout but smaller than the first, was also discovered 2 km to the southwest (Arafa 1985: 78-80). Both are of the type known in other places in the Mareotic region.5 They represent the archaeological testimony for the excellence of the Mareotic wine praised by the classical authors (see Dzierzbicka this volume). Each of the two buildings comprises two rooms separated by a low screen wall for squeezing grapes. ne is smaller than the other. Both are entirely coated by four layers of red plaster to prevent any possible seepage of the juice. In the center of the smaller room is a raised round base covered with plaster possibly to support a movable squeezer or a hand press. By means of a lion headed spout, the juice from both rooms pours into a large square basin which is dug in the ground, with another smaller and deeper basin dug in its floor for collect-

4. Rodziewicz was then a consultant for the Egyptian Antiquity Service (EAO) and was in charge of the restoration of the area.

5. A third smaller one that lies to the south of the big winery has been reported by Rodziewicz, see Rodziewicz 1998a: 27-36.

53

LAKE MAREOTIS: RECONSTRUCTING THE PAST ing residues. The inner surface of the big basin’s walls is scratched to allow the plaster coating to hold firm. The inner sides of the basin are surrounded by a shelf approached from all sides by a flight of steps. ther steps lead to the basin s floor. It is thought that this shelf is designed to support wooden beams that held a piece of cloth for filtering the juice poured from the lion shaped spout. In the middle of the north side above the spout, there is a small podium with two funnels pierced each by a hole opening at the basin. El-Fakharani’s hypothesis is that these funnels acted as measures for adding certain amounts of some aromatic flavours that produced the famous taste of Marea s wine6 (1983: 182-184).

town Ball 1942: 114, 117, fig. 17, pl. II, III . More than two and a half centuries after Ptolemy, we hear nothing about Marea. From the 5th century onwards, Byzantine Marea appears. During this time, the city has to play another vital role in Egypt’s history. The town in which the Martyr Abu Mina was buried lies about 15 km to the south of Marea. The sanctuary of St Menas was famous for its miraculous healing capacities. The importance of the area increased gradually until it reached its climax at about the end of the 5th and beginning of the 6th centuries (Kaufmann 1910: 103). Flux of pilgrims and patients from all over the Roman provinces headed to the place seeking cures for their ailments Grossman 1986: 12-13 . lasks ampulae filled with the holy waters of Abu Mina were specially made for pilgrims to take with them on their way back home as a blessing.8

The Byzantine Tomb (Map 1) To the north of the railway line, and at the southern boundaries of the inhabited area of Byzantine Marea, lies a rock cut chamber tomb (El-Fakharani 1977: 19-22). The entrance to the tomb is an arched doorway with basalt jams and lintel. It opens onto a passage with a staircase leading to a square mourning chamber. The ceiling of the passage is barrel vaulted with a helicon vault covering the part which turns to the east towards the arched entrance of the mourning chamber. The latter is surrounded by a stone bench and covered with a cross vault with pendentives. Three burial chambers with cross vaulted ceilings are cut into the walls of the tomb, forming a trefoil plan for the Hypogeum.7

The most suitable and more convenient way for those who are coming from both the Mediterranean and the Nile is to reach the lake harbour of Alexandria and navigate the lake to the nearest harbour before taking the land route to St Menas. In this respect, Marea is the nearest point. At that time, Marea’s earlier harbours were isolated on an island and pilgrims had to be ferried to the southern coast of the lake in order to take the caravan route to the sanctuary. To solve this problem, the old town of Marea was shifted 2.5 km to the east of the earlier harbour area, where the line of the ridge inclines far from the lake shore leaving a larger area including a promontory, a bay and several islands in front of the shoreline, that combined offer a naturally sheltered harbour. This area seemed more suitable to quickly establish the new harbour facilities and provide accommodation for pilgrims (Rodziewicz 1998b: 101-103). Other factors may have contributed to the shifting of the site, the most important of which is the changing lake levels, a fact indicated by the different levels of rising water during the flood seasons over many centuries.

The N ame of the C ity More than two centuries before the establishment of Alexandria, Marea, according to Herodotus (II.30, IV-161; III.12, 15), Thucydides (I.104) and Diodorus (I.68) functioned as a strategic staging post on Egypt’s northern frontiers. After the establishment of Alexandria, the former military role of Marea appears to have come to an end and the city takes on a more civic and commercial role (Rodziewicz 1998b: 101-103). Excavation works yielded neither weapons nor fortifications of neither Ptolemaic or Roman date. At the same time, classical authors, such as Athenaeus (I.33), Vergil (G eorgics, II.91-92), Horace (O des, I.37) and Columella (De Rustica, III.24), give us glimpses of the reputation of Mareotic wines. Marea became an agricultural centre as well as an intermediate station for goods moving between Alexandria and the Nile Valley (Rodziewicz 1983: 199-208). After the time of Strabo (XVII.I.14) who spoke of the city’s wealth in papyrus, bean and vine plantation, the city seems to have shrunk to a small village as stated by Athenaeus (I.33) and as could be inferred from Ptolemy the Geographer who mentioned it as “Palae-Marea” by which name he indicates the existence of the earlier Pharaonic

This leads us to the question that Rodziewicz 2003: 27-39 first raised: Was this new site of Byzantine Marea the city of Philoxenité?9 Philoxenité was built by the Praetorian Praefect Philoxenus upon orders of the Emperor Anastasius (AD 491-518) in order to serve the needs of pilgrims on their way to the shrine of St Menas. Rodziewicz counted on the Coptic Encomium in praise of St Menas, first published by Drescher (1946: 35-72; 126-149), to reach the conclusion of interpreting the later site of Byzantine Marea as the city of Philoxenité. According to the text of the Encomium, Drescher placed Philoxenité on the shores of the main basin of the lake, somewhere close to Kom Truga 35 km to the east of the Holy Shrine. Rodziewicz believes

6. For more detailed interpretations, see Rodziewicz 1998a: 27-36. 7. Various rooms on top of the Hypogeum and its vicinity were uncovered by a team of the Supreme Council of Antiquities SCA later in 1987/88. They represent a complex of funerary chapels for the tomb (Soleiman 2004: 283-286).

8. Just as Muslims nowadays do with the waters of the holy well of Zamzam at Mecca. 9. Before tackling this problem, it is worth noting that Rodziewicz in addition to his responsibility for the restoration works carried on at the site of Marea, made additionally impressive discoveries at the site.

54

M. HAGGAG: EXCAVATIONS AT MAREA/PHILOXENITÉ Bibliography

that the site of the Byzantine harbours at Marea accurately fit the location of the landing place of pilgrims seeking accommodation, food and beverage, before taking the short caravan route to the shrine.10 In this respect, it is important to point out that the Encomium, which is the sole source we know of to date about the city of Philoxenité, includes the following passage:

Ancient Sources Athenaeus, Deipnosophists, C. Burton (transl.), 1927. London. Augusti Meinekii, 1849, Stephani Byantii, Ehnicorum quae supersunt, I. Berlin. Columella, O n Agriculture, H. Boyed (transl.), 1931. London. Diodorus Siculus, Library of History, C.B. Welles (transl.), 1963. London. Herodotus, Histories, H. Carry (transl.), 1912. London. Horace, The O des, C.H. Bennet (transl.), 1925. London. Ptolemy, G eographica, E.L. Stevenson (transl.), 1932. New York. Strabo, G eography, H.L. Jones (transl.), 1949. London. Thucydides, History of the Peloponnesian War, C.F. Smith (transl.), 1969. London. Virgil, G eorgics, J. Jackson (transl.), 1908. Oxford.

...When it was finished he gave it the name Philoxenité… He ordered to build water stations, where he placed water jars, every ten miles along the route between the hospices and the shrine in order to serve the needs of these multitudes” (Drescher 1949: 15-16).

Yet, the distance between the Byzantine harbours of Marea to the sanctuary do not exceed 15 km, that is less than 10 miles, a distance that can not take more than a few hours if riding on camels and half a day if walking on foot. Then, why did the Praefect order the building of water stations every ten miles along this route?11 Rodziewicz’ hypothesis is mainly based upon believing that there is nothing evident at the site to be dated earlier than the Byzantine era. This assumption has been proved to be incorrect through what was discovered on the island and the ridge by the Alexandria University expedition. More recently our Polish and French colleagues have also uncovered some pre-Byzantine remains (Szymanska & Babraj 2001: 37-42, 2002: 47-58 see Babraj & Szymańska this volume Pichot 2004). Moreover, if the newly established Byzantine centre had the name of Philoxenité, why did a Christian author, who lived and wrote after the time of Anastasius and his Prefect, Stephanus of Byzantium (Augusti Meinekii 1849: 432), retain the old name of Marea in his writings?12

Secondary Sources Arafa, D., 1985, Wine Production in Egypt from Alexander to the Arab Conquest: As Compared to the G raecoRoman World. Unpub. MA diss., Alexandria University. Ball, J., 1942, Egypt in Classical G eographers. Cairo. Botti, G., 1902, Studi sul III Nomo dell’egitto inferiore E piu specialmente sulla regione Mareotica. Bulletin de la Sociéte Archéologique d’Alexandrie 4: 41-84. Breccia, E., 1922, Alexandrea ad Aegyptum. Bergamo. De Cosson, A., 1935, Mareotis: Being a Short Account of the History and Ancient Monuments of the NorthWestern Desert of Egypt and of Lake Mareotis. London. Drescher, J., 1946, Apa Mena: A Selection of Coptic Texts Relating to Saint Menas. Cairo. Drescher, J., 1949, Topographical Notes for Alexandria and Distric. Bulletin de la Sociéte Archéologique d’Alexandrie 38: 13-20. El-Fakharani, F.A., 1977, A Preliminary Report on the First Season of Excavation at Marea. Excavation Report presented to the Antiquities Department, Egypt. Cairo. El-Fakharani, F.A., 1983, Recent Excavation at Marea in Egypt. In G. Grimm, H. Heinen & E. Winter (eds.), Aegyptiaca Trevernsia II, Das Röm isch-Byzantinische Ägypten. Akten des internationalen Symposions 26.- 30.Se ptember 1978i n Trier, 175-186. Mainz. El-Fakharani, F.A., 1991, The Kibotus of Alexandria. Studi Miscellanei 28: 21-28. El-Fakharani, F.A., 2002, The Pharaonic Port on the Mediterranean: Its Shape, Development and Importance. In Z.A. Hawass & L.P. Brock (eds.), Egyptology at the a n of the enty first Century Proceedings of the 8t h International Congress of Egyptologists, Vol. 2: 203-208. Cairo. El-Falaki, Mahmoud Bey, 1872, Memoire sur l’antique Alexandrie,s es faubourgs,s es environs. Copenhagen. Empereur, J-Y., 1998, Alexandria Rediscovered. New York.

Finally, one would say that it is not unusual for a certain town in antiquity to have more than one name over the different periods of its history. Rhacotis/Alexandria, Thonis/ Heracleion, are very distinctive examples in this respect. It seems reasonable that the old town of Marea shifted eastward to build new facilities suitable for the new role the city had to play.13 Whether the new site took a new name or not is a question that cannot be answered yet with any certainty. In this respect, further readings of the original Encomium are required as well as any other material evidence that may appear through investigations currently being undertaken by different missions at the site.

10. Haas and Empereur share Rodziewicz’s opinion, see Haas 1997: 349; Empereur 1998: 229-239. 11. Whether this is due to miscalculations of the wording of the text, as Rodziewicz mentions, it is something that cannot be judged unless further readings of the original Encomium are undertaken, see Rodziewicz 2003: 27-39. 12. It is not evident that the various citations of Marea in the ancient literature refer to exactly the same spot as in the case of Procopius: Iacobus Haury 1913: vol. III 2, 171. We find some confusion in the use of “Marea” and “Mareotis” used to describe both the town and the lake or sometimes the whole lake district (Petruso & Gabel 1980: 1-27). 13. In 1902 Botti stated that there were two towns named Marea, the ancient and the new (Botti 1902: 73-75).

55

LAKE MAREOTIS: RECONSTRUCTING THE PAST Fraser, P.M., 1972, Ptolemaic Alexandria, Vol. I. Oxford. Gauthier, H., 1925-1931, Dictionnaire des nomes géographiques contenus dans les textes Hiéroglyphiques, Vol. III. Le Caire. Goddio, F., Bernard, A., Bernard, E., Darwish, I, Kiss, Z., & Yoyotte, J., 1998, Alexandria: The Submerged Royal Q uarters. Oxford. Grossmann, P., 1986, Abu Mina: A G uide to the Ancient Pilgrimage Center. Cairo. Haas, C., 1997, Alexandria in Late Antiquity: Topography and ocial Con ict. Baltimore. Haggag, M., 1984, Ports in Ancient Egypt till the Arab Conquest. Unpub. MA diss., Alexandria University. Haggag, M., 2002, Two Religious Buildings at Byzantine Marea. In Z.A. Hawass & L.P. Brock (eds.), Egyptology at the a n of the enty first Century Pro ceedings of the 8t h International Congress of Egyptologists, Vol. 2: 284-289. Cairo. Jondet, M.G., 1916, Les Ports Submergés de l’ancienne Ile de Pharos. Le Caire. Kaufmann, C.M., 1910, Die Menasstadt und der national heiligtum der altchristlichen Ägypter, I. Leipzig. Moussa, .S., 2002, uelques aspects de la vie quotidiè nne représentés á Marea Byzantine. In Z.A. Hawass & L.P. Brock (eds.), Egyptology at the Dawn of the enty first Century Proceedings of the 8t h International Congress of Egyptologists, Vol. 2: 478-486. Cairo. Petruso, K.M., & Gabel, C., 1980, Marea: An Environmental and Cultural Study at Lake Maryut, Lower Egypt: A Research Prospectus. Boston U niversity African Studies Center Working Papers 25: 1-27. Petruso, K.M., & Gabel, C., 1982, Marea: A Byzantine Port in Northern Egypt. Boston U niversity African Studies Centre,W orking Papers 62: 1-23. Petruso, K.M., & Gabel, C., 1983, Marea: A Byzantine Port on Egypt’s Northern Frontiers. Archaeology (Sept./Oct.): 62-77. Pichot, V., 2004, La fouille de l’Ile de Marea: Le site, Pros pections et campagnes de fouille: http://www.cealex. org/sitecealex/activities/terrestre/marea/genef.htm, CEAlex (accessed March 2008). Rodziewicz, M.D., 1983, Alexandria and District of Mareotis. G raeco-Arabica 2: 199-208.

Rodziewicz, M.D., 1988a, Remarks to the Peristyle House in Alexandria and Mareotis. Praktika, International Congress of Classical Archaeology in Athens, September 4t h -10t h,1983 : 175-178. Athens. Rodziewicz, M.D., 1988b, Remarks on the Domestic and Monastic Architecture in Alexandria and Surroundings. In E.C.M. vanden Brink (ed.), The Archaeology of the Nile Delta: 267-276. Amsterdam. Rodziewicz, M.D., 1998a, Classification of Wineries from Mareotis. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athè nes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole franç aise d’Athè nes,1 1- 12 décembre 198: 27-36. Athens, Paris. Rodziewicz, M.D., 1998b, From Alexandria to the West by Land and by Waterways. In J.-Y. Empereur (ed.), Commérce et artisant dans l’Alexandrie héllénistique et romaine. Actes du colloque d’Athè nes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole franç aise d’Athè nes,1 1- 12 décembre 198 : 95-102. Athens, Paris. Rodziewicz, M.D., 2003, Philoxenité - Pilgrimage Harbour of Abu Mina. Bulletin de la Sociéte Archéologique d’Alexandrie 47: 27-47 Rowe, A., 1954, A Contribution to the Archaeology of the Western Desert: II. Bulletin of the J ohn Rylands Library 36: 128-145. Sadek, M., 1978, The Ancient Port of Marea. Cahiers des É tudes Anciennes VIII: 67-80. Sadek, M., 1992, The Baths at the Ancient Harbour of Marea. The Sesto Congresso Internazionale di Egittologia, Vol. I: 549-554. Turin. Shahin, B., 1983, Local Pottery in Byzantine Egypt: A tudy of the finds at the City of Marea. Unpub. MA diss., Alexandria University. Soleiman, N.M.S., 2004, Marea: An Archaeological Study and the Manner of its Tourist Investment. Unpub. PhD diss., Alexandria University. Szymanska, H., & Babraj, K., 2001, Marea: First Interim Report, 2000. Polish Archaeology in the Mediterranean XII: 37-42. Szymanska, H., & Babraj, K., 2002, Marea: Second Interim Report, 2001. Polish Archaeology in the Mediterranean XIII: 47-58.

56

V. PICHOT: MAREA PENINSULA

Marea Peninsula: ccupation and Workshop Activities on the Shores of Lake Mariout in the Work of the Center d tudes Alexandrines CEAlex, C RS SR 3134 Val rie Pichot General Introduction to the Site of Marea The site, identified as Marea on the plate of Alexandria in the Atlas géographique d’Egypte1 (Description de l’Egypt 1809: Pl. 10 and then by Mahmoud Bey El- alaki in 1866 El- alaki 1872: 96 , is one of the rare examples of a harbour town on the shores of Lake Mariout that has essentially remained untouched since antiquity and accessible for archaeological studies. Situated some 40 km to the south-west of Alexandria, it stretches more than 25 ha along the southern shore of Lake Mariout at a point where

the width of the lake from north to south is less than one kilometre ig. 1 . Since 1977 the site of Marea has been the subject of archaeological excavations and those discoveries made prior to 2003 all suggested a somewhat late occupation of the site. The excavated remains, dating from the 5th to 7th centuries AD, indicate a harbour town of considerable capacity with large storage facilities as well as public buildings of high quality ig. 2 El- akharani 1983

Fig.1: Map of Egypt: location of Lake Mariout and the site of Marea ( CEAlex Archives) .

1. Hydrographic map of Lower Egypt drawn in 1801-1802, Description de l’Egypte,E tat moderne, vol. I, Paris, 1809, plate 10.

57

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig. General plan of the site identified as Marea ( CEAlex Archives) .

Petruso & Gabel 1983 Sadek 1992 De Cosson 1935: 131 Rodziewicz 1983, 1998 Szymanska & Babraj 2008 .

will be more precise once we have the results of the ceramological and numismatic studies presently under way.

Since 2003, the Centre d Etudes Alexandrines CEAlex has been working on the peninsula situated some 100 m to the northeast of the ancient town of Marea: topographical surveys, prospecting on foot, geophysical examinations and archaeological excavations have all begun to reveal the general organisation of the peninsula and the existence of a large workshop quarter that extended over much of its surface.

An imposing building constructed of large blocks and hydraulic mortar stood on this causeway about 100 m before its junction with the peninsula. It is usually interpreted as a lighthouse or landmark. A jetty stretching 26 m into the lake is situated to the south-east of this igs. 3 & 7 . The ancient causeway is far from straight. riented south/north for less than 300 m, it takes a turn to the east to follow, firstly, a south-west/north-east direction then west/east to the point of the lighthouse or landmark. It then returns to the south-west/north-east and reaches a workshop quarter. The rather particular shape of the causeway and the jetty to the south-east of the landmark create a mooring basin that is relatively well protected from the prevailing winds.

General Organisation of the Peninsula ig. 3 riginally an island, access to the peninsula was possible either from the lake onto a pier to the north that allowed for boats to moor, or by land across a causeway that was some 5 m wide that connected with the mainland igs. 4a & 4b .

n the north of the peninsula there was a building over 100 m long also built of large blocks and hydraulic mortar ig. 3 . Its northern limit ends in a system of terraces and stairways cut into the rock and leading to a pier. Its general layout around a large rectangular courtyard might suggest a public building for commercial purposes concerned with trade and storage, though it might also be a rich villa. The presence of cistern s under this building is attested by a descending passage situated in the centre of the west wing. Structures in brick and hydraulic mortar visible on the surface of the east wing could be connected with the cisterns, or could be evidence of bath facilities. Surface surveys and drawings have not yet made it possible to understand the exact role and organisation of this building or the function of its associated hydraulic system.

In 2003 the first two excavation sectors were established upon this causeway. Sector 1 igs. 3 & 5 was opened on a very badly preserved part of the track. This excavation allowed us to reveal the presence of foundations of a structure probably associated with controlling movement on and off the causeway. Sector 2 igs. 3 & 6 was established at the point where the track disappears toward the southern part of the concession. The causeway effectively stops at this point and ends in a sort of platform made of numerous amphora fragments and mortar. The archaeological material found here was relatively homogenous and dates the construction of the causeway to between the 1st and 3rd centuries AD. The chronology of its construction 58

V. PICHOT: MAREA PENINSULA

Fig.3: The peninsula of Marea ( CEAlex Archives) . 59

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig.4: A) ( above) pier to the north of Marea peninsula,v iew from the north-west.B ) ( right) ancient causeway connecting the peninsula to the mainland,v iew from the west.P hoto V.P ichot ( CEAlex Archives) . Fig.5: ( left) Ancient causeway under excavation of Sector 1,v iew from the east.P hoto V.P ichot ( CEAlex Archives) .

Fig.6: Ancient causeway under excavation of Sector 2,v iew from the north.P hoto V.P ichot ( CEAlex Archives) .

Fig.7: Landmark and j etty to the south of the peninsula,v iew from the south-west.P hoto V. Merle ( CEAlex Archives) .

60

V. PICHOT: MAREA PENINSULA Occupation and Workshop Activi ties CEAlex is currently focusing upon the central part of the peninsula where a quarter developed and evolved during the Hellenistic and Roman periods. This was principally a workshop quarter the presence of which is attested from the 3rd century BC.

a later period, occupation re-adopts the north-north-west/ south-south-east orientation present during the Hellenistic period on the elevated part of the peninsula. Survey of the peninsula led to the discovery of numerous archaeological vestiges connected to light industry. Many zones with concentrations of slag and furnace walls, point out the presence of what were workshops, the greater part of which appear to have been connected to metallurgic activities. The numerous plots showing clear anomalies detected by the magnetic survey certainly correspond to areas of intense metallurgic activity. ther anomalies may possibly represent the presence of potteries and more certainly chalk furnaces from the later era 5th-6th centuries AD , as has been shown by the excavations.

Geophysical magnetic surveys undertaken on the central part of the peninsula have revealed the existence of three major orientations, subsequently confirmed by excavation, that correspond more or less to the layout of the area at different periods ig. 8 . In the Hellenistic period 3rd-1st centuries BC there are two orientations: one south-west/ north-east for the low-lying part of the peninsula and the other north-north-west/south-south-east for the elevated part. At the end of the Hellenistic period and the beginning of the Roman era, the major orientation used within the low-lying areas follows a north-north-east/south-southwest axis, while the north-north-west/south-south-east orientation appears to continue in the elevated areas. At

Metallurgic activities The excavations undertaken in Sector 4 see ig. 3 have revealed extensive furnace activity: extensive not so much in terms of production, which is still difficult to estimate,

Fig.8: Results of the geophysical survey undertaken on the Marea peninsula. ( T.H erbich & CEAlex Archives) . 61

LAKE MAREOTIS: RECONSTRUCTING THE PAST but rather in the number of hearths discovered. In fact, in this sector where explorations both geophysical and visual suggested an empty terrain, except for a ditch situated to the western edge, some 40 hearths have been unearthed most of which are lying a mere 0.15 m beneath the surface and all are concentrated in the west and north-western part of the sector. f different sizes the size depending to a large part on the extent of destruction they are either circular in plan or occasionally rectangular. While certain examples still hold fragments of objects and/or slag, others are practically destroyed. They are not all contemporary, but they may have functioned in groups of two, three or four, several groups being in operation at the same time. The best conserved hearths were cleared when excavating rubbish ditch S4024 situated on the western edge of the sector. Its fill composed fragments of amphora, ceramics, bones, some fragments of faience objects, charcoal, some fragments of iron and seashell. The majority of the material unearthed was characteristic of the 1st centuries BC and AD, however, certain elements can be dated to an earlier Hellenistic period, e.g. a worked lamp of the 3rd century BC and amphora and ceramic fragments of the 3rd and 2nd centuries BC.

Fig.9: Sector 4 - Hearths in FS4024,unde Photo V.P ichot ( CEAlex Archives) .

r excavation.

tic period occupation itself potentially related to workshop activities, can be broadly dated to the first centuries of the Roman era. Sector 3, presently under excavation see ig. 3 , was opened in 2003 on the western part of the island and was extended in 2007 into a zone of noticeable magnetic anomalies towards the east. Work on the western part was temporarily suspended to allow, firstly, for the completion of the study of Sector 4 and then the excavation of the eastern part of Sector 3. In 2003 two ensembles of buildings separated by an alley were unearthed in the western part of the sector ig. 10 . The ensemble oriented north-north-east/ south-south-west was disturbed in its north-eastern part by large constructions orientated on a north-north-west/ south-south-east axis. Each building is composed of small workshop rooms opening onto the alley, while at the back there was a large room used for storage or habitation. ne of the activities identified in this workshop quarter is polymetallurgical Cu and e . This is attested by the remains of hearths, flooring and occupation layers characteristic to this type of activity, but also by the discovery of refuse material connected to metallurgical activity slags, drips, hammerscales, unfinished objects, etc. .

This ditch S4024 was cut by several groups of hearths ig. 9 . Among these structures, two features are quite well preserved R4034 and R4035 . Although made up of hearths with different plans one is composed of rectangular hearths, the other of circular they are constructed in a similar fashion: a hearth of 0.30 m diameter to which is attached a much smaller but deeper hearth, the exact function of which remains to be determined. These features were constructed of clay in which fragments of ceramics and amphorae were placed to act as support. The structures are well constructed and the interior walls are perfectly smooth. Their fill is of charcoal, baked clay, small fragments of iron and small furnace slag. The debris discovered suggest that metal working activity was taking place, associated with the fabrication and/or repairing of small iron objects with the use, in certain cases, of copper alloys and lead. Associated with the hearths, traces of post-holes reveal the presence of light structures that could have been more or less permanent. Stake holes near the hearths, however, indicate the probable use of fireguards, windbreaks and/or walls marking out certain specific zones, e.g. for storage.

our workspaces are presently being excavated. The outline of walls marking out storage and circulation zones, have been recognised in three of these rooms. In Space 10 a series of re-laid beaten earth floors have been revealed beneath the last occupation level. According to the initial studies of the archaeological material, it seems that levels of workshop activity can be dated to the 1st century BC to the 1st century AD. A limited trench has been opened in the low-lying zone to the south of Sector 3. The water table was quickly reached at a little more than 1 m below the surface. Three successive floors and associated metallurgic activity layers lots of charcoal, numerous small metal drips dating to the 2nd to 1st centuries BC were revealed, as well as a wall whose orientation revealed the existence of an ensemble laid out on a north-west/south-west axis.

The remains of certain furnaces consisted of no more than the very bottom traces of stakes and posts, a few strips of beaten earth floors and a part of the first foundation course of a wall MR4040 oriented west/east set upon the bedrock. The vestiges of this sector are very worn down, hence the difficulty in envisaging the spatio-temporal organisation of these workshops. Without going into considerations of contemporanity of function and the length of occupation, and having yet to complete the study of the archaeological material, the important metal working activity in this zone, implanted upon the remains of Hellenis62

V. PICHOT: MAREA PENINSULA

Fig.10:

O verall plan of the western zone of Sector 3( CEAlex Archives) . 63

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig.1 1: O verall plan of the eastern zone of Sector 3( CEAlex Archives) . Area Devoted to Cult Activity This occupation is contemporary with that of a construction ST300 partially unearthed in 2007 in the eastern part of Sector 3 igs. 11 & 12 . A paved way some 1.15 m wide situated to the north, leads to the entry of the building. The building2 is rectangular in shape approximately 10.35 m by 11.75 m, oriented north-north-west/southsouth-east along its longitudinal axis. The peripheral walls of an average thickness of 1.4 m and with concave external surfaces, enclose a space divided into several cells separated by walls that are 0.8 and 0.9 m thick.

The building consists of five different spaces. rom the outside one enters into a large rectangular room of 3.5 m by 7.55 m, divided into two by a pavement of slabs that cuts through the middle. This pavement is approximately 0.85 m wide, thus narrower than the opening of the doorway 1 m approx. , and is paved with irregular limestone slabs. It connects the entrance with the only door within the building that is pierced through the interior transversal cross-wall. Its slight irregularity and the fact that is deliberately off the axis of the room, leads one to consider the nature of the flooring that once lay on either side. There are few clues left to satisfy this query. Three fragments of slabs attached to the pavement at three different points along its line might indicate that the missing flooring was made up of the same materiel as the pavement itself, at least on the edges. The walls of this room were covered with a plaster painted to represent alabaster. The remains of another painted coating made to represent red, white and black marble facing might have belonged to a later stage or more probably to the decoration of a room on the upper floor. In front of the interior door, a rectangular cavity 0.33 m by 0.18 m had been dug into one of the slabs of the pavement. It is centred on the median axis of the doorway and in the bottom one can see in the middle, a small depression, oval in plan and conical in section, which is filled with a plug of unbaked clay. This rectangular cavity

In the north-eastern part of the building, the double-faced masonry is composed of medium sized cut blocks of limestone with an inner filling. This construction method differs clearly from the opus incertum of the other walls of the construction. This could represent an act of restoration during a possible second phase of the building s use. The upper part of the walls were built of mud-brick that were found fallen and disintegrated in the interior of the building.

2. The preliminary study of this structure was undertaken by I. Hairy, architect-archaeologist, C RS, SR3134, Centre d Etudes Alexandrines.

64

V. PICHOT: MAREA PENINSULA Fig.12: ( left) Eastern zone of Sector 3unde r excavation.P hoto V.P ichot ( CEAlex Archives) . Fig.13: ( below) Sector 3- O ne of the two limestone sphinxes discovered in the tower building ST30.P hoto V.P ichot ( CEAlex Archives) . Fig.14: ( bottom) Sector 3Bronze candelabra after restoration in the CEAlex laboratory.P hoto A.H ussein ( CEAlex Archives) . probably held a support, a base bearing either an element of decoration, e.g. statuary, or a functional object. The interior doorway was framed by two pilasters, of which only the bases remain, that bore two little, limestone sphinxes ig. 13 that were found dismantled in the nextdoor room. This interior doorway opens onto a room some 2.40/2.35 m by 4.45 m. At the end of this space, to the south, two small assemblies of blocks, the sides of which have been covered by cleanly cut stone facing, stood against the walls in the corners. These two assemblies form the base of two staircases that led to the upper floors. The two lateral spaces situated to each side of this distribution space are completely closed and inaccessible at this level. The eastern space was filled with clay and other rubble, while within the filling of the other space there is a block of masonry in opus incertum that most probably supported the weight of a now-disappeared superstructure. When considering the different aspects of this construction, its interior layout, as well as the archaeological material discovered here, a hypothesis may be that the function of this building could relate to the representations of the tower-house on ilotic mosaics, or that it was religious. This building was part of a much bigger ensemble, the extension of which to the north-west was partially uncovered by excavation in 2008 and was still relatively intact beneath the construction levels of building ST301. Two spaces clearly similar in size and limited to the east by wall MR3114, were discovered. Their western limit has not yet been determined. The most southerly space is difficult to interpret. It was practically destroyed in its entirety by the implantation of wall MR3082 that lies in part upon wall MR3121 that separates the two spaces. The excavation of the northern space, delimited to the north by MR3116, has revealed the remains of flooring of compact clay in which were set several ceramic storage jars. Although we do not yet know for sure the duration of its use, the evidence of the archaeological material, including a Hellenistic lamp with seven wick-holes, a faience pendant and a bronze candelabra ig. 14 , would suggest that this ensemble was 65

LAKE MAREOTIS: RECONSTRUCTING THE PAST settlement and organisation of a sizeable town on the lakeshore. p until the present, our knowledge of the occupation of the site of Marea was limited to the Byzantine period 5th to 7th centuries AD . The work of the CEAlex on the peninsula has proven a settlement existed here well before the 5th century AD. The excavations undertaken in Sectors 3 and 4 have revealed a large occupation dating to the end of the Hellenistic and the beginning of the Roman era, as well as numerous signs of a Hellenistic occupation prior to the 2nd century BC. The opportunity at Marea to excavate and study a workshop quarter of this size, in an environment so well defined by the fact that it stands on a peninsula, is really quite exceptional. In addition to gathering information on the production itself and on the production lines, it is also possible to study the spatio-temporal organisation of the site in the very heart of the peninsula and to resituate it within its immediate environment habitation, religious area, etc. . Bibliography De Cosson, A., 1935, Mareotis: Being a Short Account of the History and Ancient Monuments of the North-western Desert of Egypt and of Lake Mareotis. London. Description de l’Egypte, Etat moderne, Vol. 1, 1809. Paris. El- akharani, .A., 1983, Recent Excavation at Marea in Egypt. In G. Grimm, H. Heinen & E. Winter eds. , Aegyptiaca Trevernsia II, Das Röm isch-Byzantinische Ägypten. Akten des internationalen Symposions 26.- 30.Se ptember 1978i n Trier, 175-186. Mainz. El- alaki, Mahmoud Bey, 1872, Memoire sur l’antique Alexandrie,s es faubourgs,s es environs. Copenhagen. Petruso, K., & Gabel, C., 1983, Marea: A Byzantine Port on Egypt s orthwestern rontier. Archaeology 36.5: 62-63, 76-77. Rodziewicz, M.D., 1983, Alexandria and district of Mareotis. G raeco-Arabica 2: 199-216. Rodziewicz, M.D., 1998, rom Alexandria to the West by Land and by Waterways. In .- . Empereur ed. , Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athè nes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole franç aise d’Athè nes, 11-12 décembre 198: 93-103. Athens, Paris. Sadek, M., 1992, The baths at the ancient harbour of Marea. Sesto Congresso internazionale di Egittologia, Vol. I: 549-554. Strabo, G eography, A. Meineke transl. , 1877. Lipsiae. Szymańska, H., & Babraj, K. eds. , 2008, Byzantine Marea. Excavations in 20203 and 206 , Marea, Vol. 1. Krak w.

Fig.15: Sector 3- Hoe chalk burner.P hoto V.P ichot ( CEAlex Archives) . in use from the end of the 3rd century BC. It was partially destroyed by the installation of one or several occupations to the north during the Roman period that were in part connected to workshop activity,3 and thereafter by chalk burners that were active in a later period ig. 15 , most probably in the 5th and 6th centuries AD, and that largely destroyed all construction in this zone. C onclusion The Mareotis area, today semi-desert, was once a fertile region with widespread agricultural activity. In the Graeco-Roman era, Lake Mariout was a veritable inland sea with intense traffic. As a zone of movement and exchange, it was connected to the ile by canals and with the sea at several points through the course of its history. A passage also allowed traffic access to the canal of Alexandria. The Ptolemaic capital had a lakeside port with quays and warehouses that Strabo 17.1.7 considered richer than the maritime harbours of the city. Marea is one of the best examples of Mariotic lakeside towns whose development was tied to exchange between Alexandria and its hinterland: exports of agricultural production and local industry glass, metal, etc. , imports of raw materials and pilgrimages during late antiquity. Its state of preservation allows us to study the problems of

3. The mission of 2008 has brought to light numerous structures connected to workshop activities of the Roman era to the west and north of the tower building. They are presently under study and will be the object of a supplementary mission in 2009.

66

M. RODZIEWICZ: MAREA AND PHILOXENITE’S ARCHAEOLOGICAL EVIDENCE

On Interpretations of Archaeological Evidence Concerning Marea and Philoxenite Mieczyslaw D. Rodziewicz

The largest and best preserved ancient lake harbour southwest of Alexandria was located on the map of the Mariout region by Mahmoud El-Falaki in 1872 with a description and interpretation of the site as the centre of the ancient Mareotic district being identified as the site of Marea ElFalaki 1872: 96). For a long time this interpretation was considered as correct and unshakable. It was expected that with continuous advances and development of field archaeology in the region, this interpretation would be substantiated by more archaeological evidence. However, in the first half of the past century the site was not investigated by the most active archaeologists in Alexandria such as Breccia and Adriani. Instead, two other sites in the Mareotic region captured the attention of these great archaeologists of the last century. The first such place was Abu Mina discovered in the desert south of Alexandria and excavated from 1905 by German archaeologist C.M. Kaufmann 1908 . The second important site was the large city of Taposiris Magna, west of Alexandria, that spread around the Ptolemaic temple of Osiris, with the neighbouring site of Plinthine, where Breccia 1922: 353 and Adriani 1940 concentrated their field research in the first three decades of the last century. There was also a short German campaign near Amreyia, situated in the desert south of Alexandria, where very informative architectural remains of early Christian date have been uncovered Eilmann, et al. 1930 .

ten by Breccia 1922: 335 and orster 1922 , the port of Marea is hardly mentioned compared with such important places west and south of Alexandria as Taposiris Magna and Abu Mina. The best equipped excavations in the Southern Desert were made by expeditions organized by the German Archaeological Institute in Cairo in the 1930s. They were carried out near the modern market place and settlement of Amreyia Eilmann, et al. 1930 . Specific rural types of early Christian structures characteristic of the region, although rather modest and of an introductory nature, were recorded and published very quickly. It is a great loss that the excavators did not come back to the region after the second world war to continue the research. In the second half of the past century, excavations at Abu Mina, undertaken by the German Archaeological Institute at Cairo, were resumed under the direction of architects, Müller-Wiener and later Grossmann. They extended the area of research around Abu Mina, the largest pilgrimage centre of early Christian Egypt, and surveyed the neighbouring territory of the Lake Mariout harbour. M ller-Wiener 1967 and R der 1967 , devoted their time to investigating not only the ancient settlements in the region, but also the sources of local building material, including ancient quarries located along the southern shore of the lake, particularly one located near the modern village of Bahig M ller-Wiener 1967: 104-117 R der 1967: 118-131 . Their interests also extended to an analysis of the building materials visible on the surface of the land along the southern shores of Lake Mariout and the surrounding area, with the clear intention of identifying a link between the chain of settlements in the desert and the pilgrimage centre at Abu Mina. The article of M ller-Wiener 1967 , entitled Siedlungsformen in der Mareotis”, was based on rational observation and the description of visible remnants of ancient structures on the surface igs. 1 & 2 . In the harbour of so-called Marea, prior to any excavations, he identified on the surface, a double bath, structured frames of the quays, with accompanying buildings, and the most important construction, the monumental three apsidal church M ller-Wiener 1967: 106).1 All his identifications were correct, and they have not altered as a result of later excavations. More im-

The most complete description of the whole area was offered by Anthony De Cosson, the former director of the railway west of Alexandria, in the 1930s. His book on Mareotis was based on then available literature on the subject and his extensive knowledge of the land from Alexandria westwards far beyond Marsa Matruh ancient Parethonium De Cosson 1935: 131 . In his description of Marea he follows the description and name of the extensive lake port remains situated south-west of Alexandria given by Mahmoud El-Falaki. In both, El-Falaki’s and De Cosson´ s descriptions, the site was presented as 1.5 km long, but not very broad, and limited to the chain of visible ruins of fallen walls along the lake shore. Nevertheless, in their opinions the place was the most important centre of the historical Mareotic region, as mentioned in ancient literary sources. Other modern publications concerning ancient Mareotis have dealt very little with the largest lake harbour in the region. et, in the first half of the last century the site was recorded on several maps of the Survey of Egypt series, but was rarely presented in published photographs. The best known are pictures showing the western jetty of the harbour surrounded not by water, but instead by marshes. In the most popular books of that time writ-

1. M ller-Wiener 1967: 106, n. 16: In Marea sind obwohl das aufgehende Maurerwerk weitgehend fortgeschleppt zu sein scheint noch zahlreiche Bauten erkrnnbar grosses Doppelbad, Hafenbefestigungen mit Kaianlagen und drei langen Molen, Kirche mit drei-apsidialem Grundriss usw.usw. eine n here ntersuchung des rtes w re zumal angesichts seiner Bedeutung in der r hzeit usserst w nschenswert.

67

LAKE MAREOTIS: REC

STR CTI G THE PAST

Fig.1: Central part of the plan of Mareotis drawn by Mül lerWiener in 196,w ith the pilgrimage road from the lake harbour to Abu Mina and settlements,am ong them settlement “ M” in Huwaryia village ( after Mül lerWiener 1967: Fig.1) .

Fig.2: Abu Mina with surrounding settlements/ karms,and pilgrimage road towards the lake harbour ( after Mül ler-Wiener 1967: Fig.4) .

68

M. RODZIEWICZ: MAREA AND PHILOXENITE’S ARCHAEOLOGICAL EVIDENCE portantly, he identified a separate settlement M on the plan reproduced on his ig. 1 see ig. 1 , at a distance of 1.5 km south from the ruins of the lake harbour known as Marea M ller-Wiener 1967: 110 . He presented its dimensions 600 x 800 m and described briefly the remains of a brick cistern located at the edge of the modern road, the same that later was identified as a water reservoir serving the double peristyle building, published by its first excavator El- akharani, as the Byzantine House El- akharani 1983: 175 . Thus, M ller-Wiener clearly separated all these inland situated structures from the harbour to the north, and very clearly stated that settlement M was not a part of a larger town plan comparable with ancient Alexandria, as was to be suggested a decade later, and supported by his successor heading the excavations in Abu Mina Grossmann 2003: 13 .

saved most of the monuments excavated by the Alexandria niversity Expedition, headed by Prof. El- akharani. The winery in Huwaryia village, excavated in 1977, received limited reconstruction mainly for protection but also didactic reasons), and was covered with a roof. The excavation of the so called Byzantine House, which most probably served as a hospice for pilgrims heading to Abu Mina, was completed. It was partly reconstructed to prevent further damage and also for didactic requirements. The remains of buildings at the lake harbour were treated similarly Rodziewicz 2002, 2003 . During these works many previously completely unrecognised structures were identified, thus extending the possibilities of defining more features of the local topography, its culture, economy and technology. If they would have been known to Prof. ElFakharani at the time of his activities in the area, he would certainly have changed some of his first hypothetical identifications, and his general opinion of the site.

In 1978 the first archaeological excavations of the socalled site of Marea were financed by Alexandria niversity under the direction of Professor of Archaeology, Dr Fawzi El-Fakharani, a very energetic organizer and gifted speaker. In a very short time his discoveries at the site became very popular both nationally and internationally. Prof. El- akharani published in 1983 an extensive report on the survey and excavations of a vast area of Mareotis, about 6 km long and 4 km broad. ther publications of the site of Marea by Sadek 1978: 67 and Petruso and Gabel 1983: 62 followed, although analysis of available material and structures excavated there were not completed. At that time, there arose the notion of a very large and prosperous city of Marea around the existing remains of the lake harbour, which some supposed to be equal in size to ancient Alexandria. The Mareotic settlements 1.5 km south of the port, and the Ptolemaic rock-hewn tombs situated about 5 km west at the site called uassimyia , previously investigated by Mül ler-Wiener, were included by El- akharani in the city of Marea El- akharani 1983: 176, 186). The lack of any continuation and evolution of chronologically parallel architectural structures between these distant places was not taken into consideration by the enthusiasts of this notion of the great city of Marea .

The most required change in the general knowledge of the site concerns the extension of the urban area of Marea to the west, i.e. to the tombs of the Ptolemaic period at modern assimiyia site, ca 5 km from the city), which is still considered by some archaeologists as a part of the western necropolis of the capital of the whole Mareotis region El- akharani 1983: 176 Bagnall & Rathbone 2004: 75).2 Of three hypogea, still well preserved in the 1970s, only traces of the largest, located on the southern slope of the rocky ridge, are still recognisable. The identification by El- akharani of assimiyia site as the western necropolis of Marea is not archaeologically substantiated, because similar hypogea existed on the rocky ridge near Huwaryia village 1.5 km from Marea , and all along the same ridge were located other types of tombs such as chambers, shafts or pit tombs, which were dispersed over the area, also considered by some archaeologists as belonging to the city of Marea . urthermore, the extensive ancient ruins with a subterranean tomb, now located near the modern Huwaryia railway station, which supposedly delimited the southern urban area of ancient Marea , are according to the results of the survey of 1990s EAO Delta West Inspectorate unpublished , the remnants of an isolated structure with a church inside, which may indicate one of the numerous Christian monasteries that existed here in the early Christian period.3 There were no detectable traces of urban features such as street connections between this structure and the group of low quality houses with an associated wine press, that were excavated and published by El- akharani 1983: 183-184 . This site was earlier identified by M ller-Wiener as a separate settlement M ller-

In 1972 Fraser, the author of a major work on Ptolemaic Alexandria, wrote, Although Marea had been an important town in Pharaonic times, these remains are evidently late structures, they are not Ptolemaic and may be Byzantine or early Arab raser 1972: 146 . Some years later an essential question arose about the location of the pilgrimage lake harbour, known from the Coptic Encomium on St. Menas as Philoxenite, a place that is also mentioned in other Christian sources Drescher 1946: 147-148, 1949: 15-16 Rodziewicz 1983, 2003 . The next phase of research connected with the lake harbour, went together with the protective work undertaken by the Egyptian Antiquities Organization in 1982, which required a factual analysis of all preserved archaeological material essential for planning and intensifying the protection of the cultural heritage in this area. I was involved in this process, particularly in the protective works, which

2. El- akaharani 1983: 176: The discovery of burials in the west and south marked the limits of the town westwards and southwards since ancient cemeteries were customarily located outside of inhabited area Bagnall & Rathbone 2004: 75: In the cemetery to the west of the town …” . 3. Compare these structures to those excavated by the Swiss at Kellia, Kasser 1983, 1986 with further references in Actes du Colloque de G eneve,13au 15aout 1984 , Geneva 1986.

69

LAKE MAREOTIS: REC

STR CTI G THE PAST

ing a now horizontal stratigraphy, unlike most of the Egyptian ancient settlements which grew upwards and form characteristic hills called koms . Contrary to the first and still valid identification of this archaeological site by M ller-Wiener in the 1960s in particular the water cistern belonging to the big Byzantine House that he published), as a separate construction in a rural environment, the site is still directly or indirectly considered by some archaeologists as a centre of the urban area of an extended town called Marea Bagnall & Rathbone 2004: 74-76 .5 Thus, despite the growing archaeological evidence that strongly opposes the identification of the lake port as Pharaonic Marea, we still read about the remains of an ancient town Stadtanlage located as far south as the modern village of Huwaryia Grossmann 2003: 15 . As a consequence of such interpretations of the ancient village, general conclusions were formulated concerning the location of wineries in urban environments Bagnall & Rathbone 2004: 74-76 Grossmann 2003: 15, n. 27 . urthermore, another conclusion drawn, was that the so-called by Prof. El- akharani Byzantine House at Huwaryia is the most convincing archaeological document attesting the existence of the urban centre of ancient Marea, where in reality at this location there existed a large building structured in a rural manner, surrounded by cultivated land. Judging from its huge dimensions and the church built inside, it may have served as a very comfortable and safe pilgrimage hospice from the early 6th to late 7th century AD, i.e. exactly at the time of prosperous pilgrimage activities to Abu Mina Rodziewicz 2003: 27-47 . o older or younger constructions in the area of the building have been registered. The proper understanding of the function of this large building is crucial for the interpretation of the character of the area, which was described by Mül lerWiener 1967 who undertook a detailed survey of the area in the 1960s), as a rural settlement, not the centre of any larger town. He mentioned visible surface remains of a brick built water cistern that was located right on the edge of the modern road, that are still visible today see ig. 3 . In late antiquity, this cistern was situated in a garden or in an open space surrounding the large double-peristyle building the Byzantine House of El- akharani . M llerWiener thought that the cistern may eventually be connected with the bath M ller-Wiener 1967 . However, the extensive neighbouring building the Byzantine House , whose excavation was completed in the EAO in the early 1980s, appeared to be the largest residential edifice known hitherto in the whole Mareotis region. It was equipped with a church in the central, eastern wing, between the two extensive colonnaded peristyles. The church had multi-

Fig.3: Byzantine House/ Hospitium at the pilgrimage road from the lake harbour to Abu Mina ( drawn by M. Rodziewicz) . Wiener 1967: 110 see ig. 1 . In the southern part of this settlement an underground water cistern, a group of low quality rooms and, further to the west, a medium-sized winery, were partly uncovered by EA in 1992/3. They were associated with pottery of the late 1st century BC to the middle of the 2nd century AD author s identification . This winery appeared to be much older than the winery discovered by Prof. El- akharani 1983: Taf. 36 . ther architectural remains from this settlement are younger. At the most northern part of the ruins typical Muslim burials were located, which, according to Prof. El-Fakharani, were associated with glazed pottery sherds and a probable atimid coin El- akharani 1983: 176-177 .4 urther to the north, an extensive building identified by El-Fakharani as the Byzantine House was separately built outside the settlement El- akharani 1983: 184-186 . It occupied a surface of over 1,500 m2 and was surrounded on all sides by cultivated land Rodziewicz 1983, 1988a: 267-277, 1990, 2002: 1-22 ig. 3 . The house was not covered by structures belonging to other buildings, which means that it was situated in a typically rural environment, at the far edge of the older, but still inhabited village, hav-

5. Bagnall & Rathbone 2004: 74-76 p. 76: Archaeological evidence for wine production has also come to light. South of the limestone ridge are the remains of two wine-producing establishments. The larger and more interesting is aligned with the middle of the town”. However, such installations as wine factories and pottery kilns are not known in the very centres of ancient towns, where life was regulated by city law. uite the contrary, these are very typical rural establishments.

4. El- akharani 1983: 176-177: ew pieces of Islamic pottery sherds with green glaze and some white lines, and a piece of Fatimid coin were discovered just above the dead”.

70

M. RODZIEWICZ: MAREA AND PHILOXENITE’S ARCHAEOLOGICAL EVIDENCE

Fig.4: Byzantine pilgrimage harbour/ Philoxenite ( Nos. 5- 17) and western/ coastal rural remains ( Nos.1- 4) .O n the eastern side,s tructured causeway ( No.18) ( drawn by M. Rodziewicz) . K EY 1– J etty of pre-Byzantine date 2– Rural estate with large water wheel 2a – Small kiln 3– nidentified structure of rural character 4– Structures with mooring place

tructures identified in the pilgrimage harbour: 5– Pier 6– Water wheel ( saqia) 7– Public double bath 8– Insula with shops 9– Public building 01– Public lavatory 11– K ibotos?

coloured, geometrical opus sectile mosaics on the floor and a richly ornamented northern baptistery Rodziewicz 1983, 1988a, 1988b . The baptistery floor was made of mortar and painted red in a geometrical pattern composed of triangles. After extensive cleaning of the area, it appeared that beside the large church in the centre and numerous rooms around an extensive courtyard and probably on the upper floors , a collective nine seat lavatory existed there with two smaller ones aside. They were all built in the southern part of the edifice, by the staircase leading to the upper floors, and close to the bathing equipment located in the south-western corner of the house. The bathing facilities were not sufficient to serve the needs of pilgrims eventually stationed there, but a full programme of cold and hot baths was available just about 1 km to the west, in the large public bath that was surveyed, excavated and restored by EAO Delta West Inspectorate in the early 1990s.6 The bath, datable to the Byzantine period, was not attached to

21– 13– 41– 51– 16– 17– 18– 19–

Pier Transept Basilica ( after G rossmann 20 2: Fig.9) Eastern harbour Water wheel ( saqia) Public bath Rock-hewn tomb/ hypogeum Causeway leading to the island Pilgrimage track to Abu Mina

any architectural remains. It was built in an open area and surrounded by cultivated land. The area of the settlement, which extended from the uplands to the lake, bears traces of various human activities, such as limestone quarrying, various burials on the rocky ridge, clay extraction, pottery firing in the numerous kilns registered there, and cultivation down to the shores of the lake R der 1967: 118-131 . The individual buildings, whose traces have been excavated and surveyed there, were built in a rural manner. Strictly urban construction methods of building are identifiable only around the lake harbour, erroneously identified as Pharaonic Marea El- alaki 1872: 96 . Most of the buildings at the port were constructed in one phase, with similar construction principles ig. 4 . The construction methods are comparable to buildings at Taposiris Magna, Alexandria and also the pilgrimage centre of Abu Mina Grossmann, et al. 1982 Grossmann, et al. 1984: 123-151 . They are not similar to the structures of neighbouring rural settlements and the karms/ karum” located further south. West of the Byzantine passenger harbour Philoxenite Marea , on a rough natural rocky coastline, are the pre-

6. The results of which were presented by Ahmed Abd El-Fatah at the Baln orient conference, held in Alexandria, December 2006 in press .

71

LAKE MAREOTIS: REC

STR CTI G THE PAST

served traces of a small size mooring facility and a quite well preserved short but wide and relatively high, jetty that differs remarkably from the two long Byzantine quays of the pilgrimage port ig. 4.1 . The jetty is constructed from very large sandstone blocks without hydraulic mortar. In front of it a cluster of ruins ig. 4. 2 is still visible, which surround a large courtyard, water wheel saqia), cistern and channels. They probably mark structures of a coastal rural estate similar to the one at Borg el-Arab. Early Roman building material has also been recorded. In addition, there are traces of another water wheel, a small kiln ig. 4.2a , and widely dispersed traces of furnace activities. These remains depict the kind of rustic-coastal human activities with small size mooring places ig. 4.4 , created prior to the construction of a large passenger harbour at the promontory. It is worth noting that during the very last era of the passenger harbour of Philoxenite, with large buildings already abandoned and decaying, old local activities returned, motivated by the rural economy, with an oil press and a pottery kiln installed on top of the ruined constructions.

tures at the lake port Grossmann 2003: 13-20 .7 Arguments against this proposition vary from self-contradictory to those based on out of date, unproven archaeologically theories, particularly that of the existence in that place of a large town, Marea , the Pharaonic capital of the district, with a generalised opinion that the old town experienced an unprecedented economic boom in a period of well-documented and steady economic decline of Egyptian cites.8 With a new round of excavations at the lake port starting in 2000, we expected a flow of new information and objective interpretations of freshly unearthed structures and artefacts.9 However, this was not to be and the lack of a rational argument in the presentation of newly excavated material from the harbour during the conference on Medieval Alexandria in 2002, motivated me to publish a paper entitled Philoxenite-Pilgrimage Harbour of Abu Mina Rodziewicz 2003 . Three years later, I found a statement in the pamphlet published by the Polish Mission 2006 saying that there are several reasons to assume that Philoxenite was built within the city of Marea, thus supporting the interpretation of the site as that of the location of Philoxenite, which I had first suggested in 1983 Rodziewicz 1983: 202, 2002 . However, the belief that Marea lies under the Byzantine buildings of Philoxenite, despite the lack of any older urban structures there that should date back at least to the time of Psametic I, is still maintained. In the

It is the author’s belief that the lake port in Mareotis should be interpreted as a disembarkation point for Christian pilgrims travelling to Abu Mina, and not as a town with a long tradition matching the history of the ancient capital of a Pharaonic nome Rodziewicz 1983, 1988a, 1988b, 2002). This opinion was formed on the back of extensive personal study of the whole lake harbour area from the late 1970s to the mid-1990s. The text of the Coptic Encomium and other Christian sources serve to support this view Drescher 1946: 147 . Since, in the early 1980s, the water of Lake Mariout was at a very low level, all jetties and other structures at the harbour were dry, and research, as well as all necessary examination of the construction phases, was much easier than nowadays. or identification of the chronology of the site I used my knowledge of the pottery acquired on other excavations in Egypt and elsewhere, but primarily Alexandria Rodziewicz 1976 . The most informative and numerous sherds were those spotted in the mortar of the lowest parts of the jetties. They belonged to Late Roman Amphora 1, which clearly fixes the chronological horizon of the big harbour in the Byzantine period and parallels the chronology of Abu Mina. The types of the structures, the extension of the urbanised surface, three large water wells ig. 4.6, 4.15 and a very small local cemetery hewn in a rocky hill ig. 4.17 , and a very large church ig. 4.13 and extensive harbour, lead me to conclude that the site did not develop gradually, but that is was constructed in a short time and according to a very specific programme. urthermore, there were no older or younger pottery sherds collected at the site.

7. Grossmann 2003: 13-20, considers my opinion unacceptable, but his arguments are not backed by any factual documentation concerning the chronology of the artefacts in the lake harbour and pilgrimage centre in Abu Mina which are mostly not yet published . Thus, his opinion on differences in chronology of the bath in the lake port and of the pilgrimage centre at Abu Mina can be taken only as his personal supposition, not supported by the necessary documentation and study of the pottery collected there. 8. According to Bagnall 1993: 108 , despite continuous economic decay in late antiquity, not all cities in Egypt suffered regress, and Marea in the immediate hinterland of Alexandria enjoyed a period of prosperity. He says that the flourishing development at Marea, on the lake which served Alexandria´ s back door by water, points to vigorous commerce by this route in the late period”. So, the old Pharaonic city of Marea, of which we still do not have any archaeological evidence and whose location is still unknown, in Bagnall s publication flourished in the 3rd century AD. 9. The Polish Mission, active there since 2000, is headed by Dr. Hanna Szymanska from the Archaeological Museum in Krakow. With great sadness, we observe subjective methods of interpretation of the archaeological evidence obtained there. At the conference devoted to Medieval Alexandria organized by CEAlex in 2002, Mrs. E. Wipszycka from the Polish Mission spoke about the identification of Philoxenite in the light of new discoveries. Her weak documentation mobilised me to publish the 2003 the article Philoxenite-Pilgrimage Harbor of Abu Mina . In 2006, a special exhibition was prepared by the Polish team about the excavations in Marea. It accompanied the conference Baln orient at which Mrs. Szymanska presented a paper on a new bath from the harbour. This exhibition entitled Marea - Polish Excavations in Egypt 2000-2004 conducted by archaeologists from the Archaeological Museum in Kracow under the auspices of the Polish Centre of Archaeology of Warsaw niversity , was supported by a pamphlet with a short text signed K.B. in which we can find the passage: Marea or Pholoxenite Identifying the accurate name of the town is a challenge that the Polish Archaeological Mission has undertaken. We have several reasons to assume that it was within Marea that the city of Philoxenite was built.”

Thus, in 1983 at a conference in Athens Graeco-Arabica I interpreted the harbour as a specialised port for pilgrims travelling to Abu Mina, as described in the Coptic Encomium, which also mentions its name as Philoxenite Drescher 1946: 15 . The description of the site in the Coptic Encomium matches quite well the preserved Byzantine struc72

M. RODZIEWICZ: MAREA AND PHILOXENITE’S ARCHAEOLOGICAL EVIDENCE Bibliography Abd el-Aziz, M., 1998, Recent Excavations around Abou Mina. In .- . Empereur ed. , Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athè nes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole franç aise d’Athè nes,1 1- 12 décembre 198: 65-73. Athens, Paris. Abd el-Fatah, A., 1998, Recent Discoveries in Alexandria and the Chora. In .- . Empereur ed. , Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athè nes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole franç aise d’Athè nes,1 1- 12 décembre 198: 37-53. Athens, Paris. Adriani, A., 1940, Annuaire du Musée gréco-romain 1935193. Alexandria. Bagnall, R.S., 1993, Egypt in Late Antiquity. Princeton. Bagnall, R.S., & Rathbone, W., 2004, Egypt from Alexander to the Copts. London. Breccia, E., 1922, Alexandrea ad Aegyptum: G uide de la Ville Ancienne et Moderne et du Musée G réco-Romain. Alexandria, Bergamo. Décobert, C., 2002, Maréotide médievale. Des Bédouins et des chr tiens. In C. D cobert ed. , Alexandrie Mediévale 2, É tudAlex 2, 127-167. Le Caire. De Cosson, A., 1935, Mareotis: Being a Short Account of the History and Ancient Monuments of the NorthWestern Desert of Egypt and of Lake Mareotis. London. Drescher, ., 1946, Apa Menas. A Selection of Coptic Texts Relating to St Menas. Cairo. Drescher, ., 1949, Topographical otes for Alexandria and District. Bulletin de la Société Archéologique d’Alexandrie 38: 13-20. Eilmann, R., Langsdorff, A., & Stier, H.E, 1930, Bericht Ü ber die Voruntersuchungen auf den Kurûm El-Tiuwal bei Amrijr. Mitteilungen des Deutschen Archaologi schen Instituts Abteilung K airo 1: 106-129. El-Ashmawi, F., 1998, Pottery Kiln and Wine-Factory at Burg el-Arab. In .- . Empereur ed. , Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athè nes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole franç aise d’Athè nes,1 1- 12 décembre 198: 55-64. Athens, Paris. El- akharani, .A., 1983, Recent Excavation at Marea in Egypt. In G. Grimm, H. Heinen & E. Winter eds. , Aegyptiaca Trevernsia II, Das Röm isch-Byzantinische Ägypten. Akten des internationalen Symposions 26.- 30.Se ptember 1978i n Trier, 175-186. Mainz. El-Falaki, Mahmoud Bey, 1872, Memoire sur l’antique Alexandrie,s es faubourgs,s es environs. Copenhagen. Empereur, .- ., & Picon, M., 1998, Les ateliers d amphores du Lac Mariout. In .- . Empereur ed. , Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athè nes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole franç aise d’Athè nes, 11- 12 décembre 198: 75-91. Athens, Paris.

same pamphlet, the author presents important archaeological evidence, previously published by Grossmann 1993 . The pamphlet informs us that: The most interesting structure in the town is a basilica situated on an eminence overlooking the lake shore near the longest of the harbour piers… The greatest surprise however was concealed under the apse. About 1.80 m below the surviving tops of the apse walls, the floor of the firing chamber of a great kiln for the production of amphorae was discovered. The part of the church where liturgical ceremonies were held had been founded on these manufacturing remains. The kiln, which is one of only a few known from the Delta so far, turned out to be 8 m in diameter and the firing chamber floor was ca. 50 cm thick. Inside it, fired amphorae of the 2nd-3rd century were discovered next to clay stacking rings that had separated the vessels during firing.

Because no further evidence was presented, I feel obliged to conclude that, firstly, this kiln alone proves that the hill is to a great extent artificial, because it was formed by industrial refuse produced by the ruins of a vast pottery workshop but that secondly, these ruins should be understood as part of an extensive rural establishment of Early Roman type, comparable to several groups of similar remains hitherto uncovered around the lake Rodziewicz 2002: 12, ig. 3 Abd-el atah 1998 El-Ashmawi 1998 Abd El-Aziz egem 1998 . ne site discovered in the 1980s by the Marsa Matruh Inspectorate of the Egyptian Antiquities Organization, was located on the southern shore of this same Lake Mariout, near modern Borg el-Arab village El-Ashmawi 1998 .10 Thus, the kiln discovered under the church in Marea ig. 4.13 belongs to a chain of Early Roman rural enterprises located all along the southern shore of the Mareotis Lake and has nothing to do with the great city of Marea. Therefore, we should retain the view that the Byzantine port of Philoxenite, created for pilgrims travelling to Abu Mina, which was built not upon the urban ruins of Marea ancient capital of Mareotis as described by Herodotus II.149 but on the ruins of a nameless, extensive Early Roman rural estate, similar to many others in the region, a number of which have been located along the shores of Lake Mariout.11

10. El-Ashmawi 1998. I had the privilege to work there in 1987/88 as an adviser in the protective works of this large Early Roman rural estate, situated on Lake Mariout. 11. Preliminary reports on excavations in Marea are scarce see Polish Archaeology in the Mediterranean XII-XVII, 2001-2007, Warsaw niversity Press . The exhibition illustrating results of the Polish excavations in Marea presented in the Bibliotheca Alexandrina, December 2006, was supplied with texts describing photos. A pamphlet was also presented. Both ignored all previous extensive research and restoration contributions to the site done by individuals and institutions. Reconstruction done at the site in the 1980s shown there on big panels that was carried out by the Egyptian Antiquities Organization was not properly described and the suggestion was that the works had been done recently by the Polish Mission. See Szymanska & Babraj 2007, with references to previous reports.

73

LAKE MAREOTIS: REC

STR CTI G THE PAST

Forster, E.M., 1922, Alexandria: A History and a G uide 5th edn.). New York. Fraser, P. M., 1972, Ptolemaic Alexandria, Vol. I. Oxford. Grossmann, P., Kosciuk, ., Severyn, G., & Severyn, H.G., 1984, Vorl. Bericht. Mitteilungen des Deutschen Archaologischen Instituts Abteilung K airo 40: 123-151. Grossmann, P., aritz, H., & Romer, C., 1982, Vorl. Bericht. Mitteilungen des Deutschen Archaologischen Instituts Abteilung K airo 38: 131-154 Grossmann, P., 1986, Abu Mina. A G uide to the Ancient Pilgrimage Center. Cairo. Grossmann, P., 1993, Die uerschiffbasilika von Hauwariya und die Bauten dieses Typus in Ä gypten als Repr sentanten der verlorenen r hchristlichen Architektur Alexandreias. Bulletin de la Société Archéologique d’Alexandrie 45: 107-121. Grossmann, P., Arnold, ., & Kosciuk, ., 1997, Report on the Excavations at Abu Mina in Spring 1996. Bulletin a la Societe d´ Archeologie Copte 36: 83-98. Grossmann, P., 2002, Christliche Architektur in Ägypten. Leiden. Grossmann, P., 2003, ochmals zu Marea und Philoxenite. Bulletin a la Societe d´ Archeologie Copte 42: 13-20. Herodotus, The Histories. A. de S lincourt transl. , 1971. Middlesex. Kasser, R., 1983, Survey Archeologique des K ellia ( BasseEgypte) . Rapport de la Campagne 198 1, Vol. I-II. Louvain. Kasser, R., 1986, Le site monastique de K ellia. Sources historiques et explorations archeoloqigues. Geneva. Kaufmann, C.M., 1906, Die Ausgrabung der MenasHeiligtüm er in der Mareotiswüs te. Cairo. Kaufmann, C.M., 1908, La Découverte des Sanctuares de Ménas dans le Désert de Maréotis. Cairo. Kaufmann, C.M., 1924, Die Heilige Stadt der Wüs te. U nsere Entdeckungen, G rabungen und Funde in der altchristrlichen Menasstadt. Kempten. M ller-Wiener, W., 1967, Siedlungsformen in der Mareotis. Archäol ogischer Anzeiger 82.2: 103-117. Petruso, K., & Gabel, C., 1983, Marea. A Byzantine Port on Egypt´ s Northwestern Frontier. Archaeology, Sept/ ct: 62-63, 76-77. Röde r, J., 1967, Antiken Steinbrüc he der Mareotis. Archäol ogischer Anzeiger 82.2: 118-131.

Rodziewicz, M.D., 1976, La céramique romaine tardive d´ Alexandria, Alexandrie I. Varsovie. Rodziewicz, M.D., 1983, Alexandria and District of Mareotis. G reco-Arabica 2: 199-216. Rodziewicz, M.D., 1988a, Remarks on the Domestic and Monastic Architecture in Alexandria and Surroundings. In E.C.M. van den Brink ed. , The Archaeology of the Nile Delta, Proceedings of the Seminar held in Cairo, 19- 2 O ctober 1986 : 267-276. Amsterdam. Rodziewicz, M.D., 1988b, Remarks on Peristyle House in Alexandria and Mareotis. Praktika 1983 : 175-178. Rodziewicz, M.D., 1990, Taenia and Mareotis, Archaeological Research West of Alexandria. Acta of the First International Colloquium of the Egyptian Society of G reek and Roman Studies I: 62-81. Cairo. Rodziewicz, M.D.,1995, Eco-Archaeology of Ancient Alexandria and Mareotis. In A.A. Hussein, M. Miele & S. Riad eds. , Proceedings of the Seminar on G eosciences and Archaeology in Mediterranean Countries, Cairo,N ovember 28- 30/ 193 : 127-139. Cairo. Rodziewicz, M.D., 1998a, Classification of Wineries from Mareotis. In .- . Empereur ed. , Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athè nes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole franç aise d’Athè nes,1 1- 12 décembre 198: 27-36. Athens, Paris. Rodziewicz, M.D., 1998b, From Alexandria to the West by Land and by Waterways. In .- . Empereur ed. , Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athè nes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole franç aise d’Athè nes, 11- 12 décembre 198: 93-103. Athens, Paris. Rodziewicz, M.D., 2002, Mareotic Harbours. In C. Décobert ed. , Alexandrie Médiévale 2, É tudAlex 8: 1-22. Rodziewicz, M.D., 2003, Philoxenit Pilgrimage Harbour of Abu Mina. Bulletin de la Société Archéologique d’Alexandrie 47: 27-47. Sadek, M., 1978, The ancient port of Marea. Cahiers des É tudes Anciennes VIII: 67. Szymańska, H., & Babraj, K., 2007, Marea. Sixth Season of Excavations. Polish Archaeology in the Mediterranean XVII. Reports 2005: 55-66.

74

K. BABRAJ & H. SZYMANSKA: MAREA OR PHILOXENITE?

Marea or Philoxenite? Polish Excavations in the Mareotic Region 2000-2007 Krzysztof Babraj & Hanna Szymańska

Harbour installations of all kinds can be identified all along the south-western shores of Lake Mareotis. The most imposing ruins of Byzantine date located just 45 km southwest of Alexandria, extend along the coast for 1.5 km and are currently under exploration by a Polish mission ig. 1.

gion remained the most important agricultural producer in northwestern Egypt Haas 2001: 47 . The lake was an important communication route in Ptolemaic and Roman times. Goods were transported down the channels from inland, reloaded in Marea and sent to Alexandria, from where they were shipped to other parts of the Mediterranean. The information contained in the Periplus of Pseudo-Scylax Desanges 1978: 404-405 , dated presumably to the early 4th century BC, indicates that Lake Mareotis may have been navigable in Pharaonic times and that the water was potable. Strabo VII, 1, 7 reports that the lake port in Marea handled more goods than ever came to Alexandria by the sea route oyotte & Charvet 1997: 83 . South of Alexandria, there was a harbour, no longer visible, called Portus Mareoticus it handled the goods that were shipped via the lake. rom there they were transported down a canal named Kibotos through Alexandria to the Mediterranean. ne of the canals linking Mareotis with the so-called Canopic Branch of the ile Rodziewicz 1983, 1998: 101 . The fertile Mareotis region was celebrated in antiquity for its vineyards, olive groves, fruit orchards and papyrus plantations. It produced food for the capital city and rich landowners had their estates here. In the mid-7th century it was still a place for Egyptians to come to appease hunger History of the Patriarchs 1. 14, 501 . Ancient writers spoke warmly of the quality of the local wine exported to Rome Virgil, G eorg. II, 91 Strabo, VII, 1, 14 Horace, O des I, 37, 14 Athenaeus, Deipn. I, 33 . Even today one of the most renowned Egyptian grapevine growers and wine producers is based on the southern shores of the lake. Almost 30 furnace sites for firing amphorae from the early Hellenistic to late antiquity have been identified on the southern lake shore Empereur & Picon 1998: 75-91 El-Ashmawi 1990: 55-64 . In 2003, the Polish expedition unearthed beneath the basilica at Marea, a pottery kiln for firing amphorae of 2nd-3rd centuries date see below . This evidence, when considered together with the numerous remains of glass workshops in the region, gives an idea of the crafts that were an important part of the economic life of the region Kucharczyk 2008: 129-143 .

In 2000, excavations at the site were undertaken by a Polish team working under the auspices of the Polish Centre of Mediterranean Archaeology of Warsaw niversity and the Archaeological Museum in Krak w. The license granted by the Supreme Council of Antiquities SCA of Egypt, covers an area of c. 19 ha. A hypsometric grid was put in place and a surface survey of architectural remains carried out 16 objects were found Szymańska & Babraj 2008: ig. 1 . In the course of the following eight field campaigns, the team has fully cleared a Byzantine bath of the 5th to 7th century connected with a well that supplied it with water, and a funerary chapel. The present focus is on the exploration of a huge Christian basilica standing on the promontory above the port. History of the Site The site was identified as ancient Marea Μαρεία by Mahmud Bey El- alaki, court astronomer to Egypt s Viceroy Ismail Pasha El- alaki 1872: 96 Kees 1930: 16761678 Am lineau 1973: 241-243 Calderini 1980: 233 Goma 1980: 647 , but today this identification is being reevaluated as indicated in the discussion below. The extensive town ruins situated on Lake Maryut ancient Mareotis , 45 km from Alexandria, near the village of Hauwariya, have fascinated researchers for a long time. De Cosson prepared one of the first maps of the coast presenting the location of mainly architectural remains De Cosson 1935: 131 . The history of this imposing harbour town with four grand jetties for ships to dock and port basins to protect against wind and waves, can be traced in the ancient sources. Herodotus II, 30, 2 reported that in the times of Psammetich I of the 26th Dynasty, Marea was a strategic position garrisoned by troops defending the border with Libya. General Amasis defeated the armies of Pharaoh Apries at Marea and took the throne in 570 BC Diodorus Siculus I, 68 . nder Persian rule, the town was allegedly the capital of an independent Libyan-Egyptian kingdom extending from the Canopic Branch of the ile to Cyrenaica. Its ruler by the name of Inaros, challenged the Persians and lost his life after the fall of Memphis in 454 BC Thucydides I, 104 Winnicki 2006: 135-142 .

The channels that fed the lake silted up in the 8th and 9th centuries due to lack of maintenance, and the fact that fresh water from the ile, mainly the Canopic Branch, was cut off. The lake began to dry up Rodziewicz 2002: 9 leading to the decline of towns particularly along its southern shores. In 1801, the lake bed was filled with sea water after the English opened the sluices in order to cut off the apoleonic army from fresh water. The present surface of the lake is c. 90 km2, and its depth averages at 1.50 m Blue & Ramses 2005: 7 .

Marea never lost its importance as a commercial harbour despite the founding of Alexandria, and the Mareotis re75

LAKE MAREOTIS: REC

Fig.1.R K EY 1. 1a. 2. 3. 4. 5.

STR CTI G THE PAST

emains of structures along the southern shore of Lake Mareotis.

Tholos baths Stone pier 41m long Stone pier 111m long Stone pier 125m long Harbour facilities nidentified structure presumably dry dock uncovered by Petruso

6. .7 8. 9. 10.

ra n by M. iepok lc ycki and . as c yk .

teps belonged to the unidentified structure y antine baths Polish e ca ations Well operated with saqiyah ( Polish e ca ations Double baths uncovered by F.E l-Fakharani Decumanus along the shops cleared by F.E l-Fakharani

11. Remains of saqiya 12. G ranary uncovered by F.E l-Fakharani 13. asilica Polish e ca ations 14. Funerary chapel ( Polish e ca ations 15. Tombs carved in the rock 16. Remains of the ancient road leading to the island

time of Byzantine domination the region had lost its leading role. The town fell due to lack of water as discussed above, but the decline was also due to political and social factors. Mareotis was overrun by semi-nomadic Bedouin tribes who plundered whatever stood in their way D cobert 2002 . At the beginning of the 8th century, in the rule of the Abbasids which coincides with the period when all life seems to have disappeared from the site, a new wave of Arab nomads appeared in the region. These people were politically active and they increasingly took control of local church institutions D cobert 2002: 137 . bviously, the atmosphere of the place was no longer favourable to

Mareotis was Christianised eventually but the relations with the Patriarchate in Alexandria passed through different phases Timm 1988: 1593-1603 . The first evidence of Christianity is linked to waves of repression, the effect of which resulted in adherents of the new religion being banished to the region. In 538, by the power of ustinian s edict, the entire Mareotis region, previously part of the Roman province of Aegyptus Prima, was incorporated into Libya Edict 13.1, 9, 17-22 . In the period immediately prior to the conquest of Egypt by the Arabs in 641, Mareotis was a Byzantine province. The infrequent reference to Mareotis in Coptic writings is proof that by the 76

K. BABRAJ & H. SZYMANSKA: MAREA OR PHILOXENITE? Christians, particularly as a stopover for pilgrims. The wine from Mareotis, however, still retained its reputation through to the 7th century at least this is what we are led to believe from a story about the abstinence of monks cited by ohn Moschus 1946 .

difficult in view of continued industrialization and numerous tourist villa complexes being built along the shore. Remains of harbour facilities can be seen along both the southern shore and the northern shores of the lake. The biggest of these is Marea. The quay at the waterfront in the eastern end of the site preserves stone blocks that were used in its construction. Beside it there are four almost wholly preserved jetties where the ships docked. These jetties formed the eastern, central and western harbours. They were built of several courses of stone blocks of different size three or four courses have still been preserved and can be studied thanks to the overall drop of the level of water in the lake. The dimensions of the jetties are counting from the west to east : I L. 41 m, W. 6.50 m, blocks averaging: 0.50 m x 1.35 m II L. 111 m, W. 5 m III The longest jetty is located by the promontory with the ruins of the basilica, L. 125 m, W. 7 m, blocks averaging 0.65 m x 1.70 m IV L. 0.35 m presumably only the part on land has survived , W. 4 m found next to the island that was connected in antiquity with the mainland by an artificial causeway.

However, identification of early harbour remains at Marea was questioned mainly on the strength of field survey results, which revealed nothing earlier than 6th to early 8th century pottery. The architectural remains also proved to be essentially Byzantine in nature raser 1972: 146 . n these grounds Rodziewicz 1983: 202-204 proposed to identify the ruins recorded on the surface with a town established as a transfer point on the southern shores of the lake for pilgrims on their way to the sanctuary of Abu Minas some 20 km away. The Coptic Encomium of St. Menas written by the Patriarch Ioannes IV 775-789 , mentions numerous conveniences like hospices by the lake and rest-houses for those wishing to rest Drescher 1946: 147148 , a market place, porticoes and even a facility that we would refer today as a left-luggage office. This impressive rest stop was said to have been founded by Philoxenos, Prefect of the Emperor Anastasius 491-518 , hence the town s name of Philoxenite. Wipszycka 2002 has identified this official as the consul Philoxenos Soterichos, who held office in 525 and who is known also from consular diptychs Martindale 1980: 879-880 .

Marea or Philoxe nite? The size of the harbour installations, which are currently being researched by a Greek-Egyptian team of underwater archaeologists headed by ikos Lianos from the Hellenic Society for the Study and Preservation of Marine Cultural Heritage and the Department of nderwater Archaeology in Egypt when considered in connection with the considerable size of the basilica on the lakefront and the innumerous but conclusive evidence of Early Roman potsherds, indicates that a large urban agglomeration flourished on the spot already in pre-Byzantine times. either can the extensive sewage network be considered as belonging to a town of little consequence. Philoxenos may have founded his city adjacent to existing harbour installations that had formerly been part of Marea, before it was abandoned as an urban centre prior to the silting up of the fresh-water channels that supplied the lake, but there is little doubt that settlement existed here prior to its foundation. Already in the 2nd century AD Athenaeus 33d wrote of Marea as the source of a wine of fine quality he called Marea a village κώμη . Athanasios 85 col. 400b-c had the same impression of the town 200 years later. Grossmann 2003: 16 thinks that the Hellenistic architecture of Marea was of mud-brick which is easily disintegrated. It should be noted that at the margins of the settlement mud-brick is a building material used in the lower layers of the basilica, chronologically contemporaneous with the pottery kiln of the 2nd-3rd centuries AD.

Location of the Site The geographical coordinates of the site are: 30o 59 32.85 and 29o 38 58.34 E to 30o59 48.84 and o 29 39 28.32 E. The excavation area measures 19.7 ha. To the west it is limited by a double-lane road linking the waterfront with Shakush settlement the lake is a natural boundary on the north and east, and to the south there is the desert which extends as far as Shakush settlement. The overall height difference does not exceed 8 m anywhere on the site. The site grid consists of 42 geodetic points, mostly dictated by the natural topography. The coordinates were assumed locally, while the heights are listed by the Egyptian state geodetic service. The topographical plan of the site so far only covers the area excavated in 2000-2003 see ig. 1 . rom a geological point of view, Marea lies in a broad stretch of land between Mallahet Mary t, the long westward arm of Lake Mareotis, and a parallel ridge of oolithic limestone. The soils on either side of the ridge are mostly calcarenite, very favorable for cultivation with proper watering Warne & Stanley 1993 El- akharani 1983: 175 Rodziewicz 1995 Mycielska-Dowgia o & Woronko 2008: 17-18 . A similar ridge lies on the opposite side of the lake, separating Lake Mareotis from the sea. This is the Taenia ταινία Ridge which extends along the coast from Alexandria all the way to Abu Mina and Libya. Today these ridges are mostly destroyed as a result of heavy exploitation of limestone quarries. In the 4th and 5th century, Christian monasteries appeared in the Taenia they were called after the mile-stones that served to measure the distance from Alexandria. Their localisation today is very

There is however, opposition to the idea that this was the site of Philoxenite. The distance from Philoxenite to Abu Mina is one area to question. According to Drescher s 1946 translation of the Encomium, the distance amounted to about 30-35 km. The text also mentions rest stops 77

LAKE MAREOTIS: REC

STR CTI G THE PAST

where water was available, which the Prefect had arranged every 10 miles or so 14.80 km between the two towns. However, the distance in reality is some 20 km, not enough for even two water stops. et the key argument against the identification with Philoxenite is Grossmann s research on the frequency of pilgrimages to Abu Mina. Pilgrimages to the sanctuary did not peak until the second half of the 5th century, and afterwards surges in the number of pilgrimages accompanied the feast days of St Menas Grossmann 2003: 18 . The ruins currently being investigated on the site are considered earlier the double baths explored by Elakharani 1983 see both Haggag s papers in this volume originate from the end of the 4th century and the pottery kiln under the apse of the basilica operated in the 2nd-3rd centuries AD. The size of this latter facility is unmistakable proof of a flourishing centre existing on the spot possibly Marea prior to the Byzantine foundations, perhaps producing the wine that Mareotis was famous for and expediting it to other towns of the Greco-Roman world in amphorae produced at the site. One of the sherds from the fill of the pottery kiln bore the stamp of a potential owner of the vineyard ιονυσίου Szymańska & Babraj 2005: 130, n. 41 . Many artifacts discovered in the neighborhood of Marea confirm the presence of pottery kilns for firing amphorae in Roman times Empereur & Picon 1998 .

block-joining technique that went out of use in Italy in the 1st century AD, evidently point to an earlier origin Adam 2008: 57 . Moreover, the apsed funerary chapel, discovered by the Polish mission, with its three chambers intended for an affluent family, could not have been erected by and for pilgrims transiting through the town. n the other hand, there can be no doubt that pilgrimages to the temple of the holy martyr must have passed through this area, if anything because of the close proximity to the lake. More evidence is provided by a Late Roman villa rustica, uncovered in Hauwariya village, transformed into a dormitory in the middle of the 6th century and furnished with a small church inside it Rodziewicz 1988: 271-273, ig. 2 . A large urban centre with developed harbour facilities must have been perfectly suited as a stopover for wayworn and surely often ailing pilgrims. D cobert 2002: 129 identifies ancient Marea with the locality of kinj Mary t to the east of the site, where excavations are currently being conducted. et the only arguments in favour of this theory is the strategic position of this locality, affording control of the road from Cyrenaica and Egypt to Alexandria and being accessible also from the opposite side of the lake thanks to a considerable narrowing of its width at this point.

The great jetties are of key importance for the dating of the port. The structural variety of jetty construction different size blocks suggests several building stages, but the characteristic imprints left by swallow-tail metal clamps, a

Fig. . Marea. y antine baths.

78

ra n by . arara .

K. BABRAJ & H. SZYMANSKA: MAREA OR PHILOXENITE? Baths Szymańska & Babraj 2008a: 27-83 The freestanding building follows an orthogonal plan, encompassing the baths itself ig. 2 and the well, which is furnished with a s iyah installation drawing water for the baths. The actual building of the baths, covered an area of 642 m2, including two courtyards and was encircled by a wall of stone blocks. The baths itself was built entirely of red brick on a brick foundation ig. 3 . Shops and a latrine lined the north wall on the outside. Colonnaded courtyards were located to the east and west.

The underground part of the building included the service area cellars, two heating furnaces and one for heating water in a boiler, finally, there were four hypocaust cellars: two in the men s part and two in the women s part. The interior decoration of the building consisted of marble pavements, columns with Corinthian capitals, and a few coatings of painted wall plaster featuring a colorful vegetal frieze. The floors and selected walls of pools were faced with marble slabs. This was evidently a luxurious complex and definitely not the only one in town to judge by the solid water channel system to which it was connected.

The building was a double bath, consisting of two uneven parts separated by a brick wall: the southern one, for men, bigger than the northern one for women. Each of these parts was heated by a separate furnace supplying warm air to four hypocaust systems. The entrance for men was in the western facade it led from the courtyard, following widely accepted circulation principles, that is, directly from the apodyterium into the tepidarium, and then into the two caldaria. The women s part was not as extensive, accessible from a rather spacious courtyard to the east side. It consisted of three rooms: apodyterium and two caldaria. ourteen small pools were used for bathing, eight of which were located inside the two parts of the baths, by the outer wall of the caldaria. Two of them were accessible from the courtyard, leading to the men s part. our pools were located outside the baths, symmetrically in the two courtyards, abutting the walls. The pools were differentiated in shape, from semicircular through rectangular to round, depending on the rebuilding phase. Calculations made on the basis of six fragments of brick arches found in the debris permit a reconstruction of the width of baths halls, which were presumably covered with barrel vaults. The hypothetical height of the building, derived from the proportions of the surviving walls, was 3.50 m.

owhere in the building was a clear stratigraphy of the construction and occupation of the baths visible and the fact that the building had collapsed vide brick arches from the vaults lay immediately on the floors precludes precise dating. An analysis of the pottery finds has provided a timeframe for the operation of the complex between the first half of the 6th century and the first 30 years of the 8th century, that is, before the appearance of Islamic Early Lead Glazed wares, of which not unsurprisingly, not one sherd has been found at the site Majcherek 2008 . Arab coins already issued after the monetary reform, that is, after 686, and remaining in circulation for 750 years, confirm the continued operation of the baths after the damages incurred during the Persian occupation and the Arab invasion Malarczyk 2008 . The most intensive use of the baths occurred in the first half of the 7th century, a fact confirmed by both the pottery finds and the prevalence of coins issued by Chosroes II Lichocka 2008 . Sāqiyah Szymańska & Babraj 2008: 85-99 A well operated by a s iyah stood about 5 m north of the baths. The well itself was 5 m deep, rectangular in plan and built of stone blocks igs. 3 & 4 . The water in the

Fig.3.M area.B aths and western courtyard.V iew from the south-west ( photo P. us ek .

79

LAKE MAREOTIS: REC

STR CTI G THE PAST

Fig.4.M area.W ell operated with sāqiyah. View from the south photo by . alarus .

ity for collecting water in case of drought. n top of this installation was a treading circle for the animals turning the s iyah wheel. A very simple hydraulic system based on connected vessels was employed. Water was collected in the reservoir of the s iyah and carried to the baths via a system of terracotta pipes

well came from a still functioning underground spring in its northwestern corner. The mineral content of this water is much more abundant compared to the lake water. South of the well and level with its rim, there was the floor of a reservoir made of fired bricks in waterproof mortar. The south side of this basin, which was filled with water from the well, was reinforced with a solid platform built of stone blocks. A capacious cistern closed off the complex of the s iyah to the north. It was presumably intended as a facil-

s

nly one conduit reaching Basin in the baths south of the iyah see ig. 3 has been preserved by the west wall Fig.5.M area. Funerary chapel. View from the west photo P. us ek .

80

K. BABRAJ & H. SZYMANSKA: MAREA OR PHILOXENITE? Fig.6.M area.P lan of the basilica ( drawn by . arara .

of the baths see plan in Szymańska & Babraj 2008: ig. 1 . The pipe, a fragment of which is visible under the last block of the s iyah reservoir with two blocks of the limestone bedding, ran to the north wall of the baths. The pipe turned into Room B1, which it crossed until it reached the south wall, where it was raised c. 1.50 m in order to empty into Basin . There is every reason to believe that there were more s iyahs operating in the town complex of Marea. A ground survey of the site has identified the remains of at least one other facility of the kind in the area to the west of the baths Szymańska & Babraj 2008: 88 .

Basilica Szymańska & Babraj 2006: 107-117 The most interesting building at the site, however, is a basilica situated on the hill near the longest harbour jetty. It was discovered by M ller-Wiener 1967: 106, n. 16 , former director of excavations at the nearby sanctuary of Abu Mina. But it was Grossmann who first determined the plan and dimensions of the building in 1986 Grossmann 1993: 107-121 . Surprisingly, none of the ancient written sources mention this church, which was one of largest buildings of its type yet known in Egypt see Grossmann 2002, passim .

Funerary chapel Szymańska & Babraj 2008: 177-185 . One hundred meters to the south there were the ruins of a building which were identified as a funerary chapel. It was furnished with an east-oriented apse and three masonry grave chambers containing 23 burials associated with very poor grave goods ig. 5 . They contained both genders of different age, all clearly of family character. It functioned for a little over a hundred years and consequently must have belonged to the town s inhabitants. A Gaza amphora under the floor of the apse verified a 6th century date for the construction of this complex.

It was a squat-shaped building, divided by columns into three aisles, furnished with a wide transept terminating in rounded arms and a relatively tiny apse Szymańska & Babraj 2004: 53-56, 61-64, 2005, 63-67, 2005a, 43-54 . The liturgical rooms were not on the outside, as was the rule at other basilicas, but incorporated into the body of the building on either side of the apse ig. 6 see Grossmann 2002, passim . The baptistery with round baptismal font apparently remained from an earlier building, possibly a small chapel, which preceded the great basilica at the site. 81

LAKE MAREOTIS: REC

STR CTI G THE PAST

Fig.7.M area. Apse of the basilica.V iew from the west. ‘ A’ and ‘ b’ locate the two burials photo by . alarus .

A kind of money exchange building intended presumably for pilgrims visiting the basilica, was discovered beyond the southwestern corner of the building. Rare examples of bronze weights used to verify coin weight were found inside this room Szymańska & Babraj 2005a: 54 .

rated columns of various sizes, all of them in Proconesian marble and probably imported via Alexandria. The interior decoration of the basilica also included pavement mosaics of which small marble cubes found in the building are the only surviving evidence.

Two burial chambers with multiple burials were discovered under the floor of the apse ig. 7, a & b . Anthropological examination identified over 100 individuals: men, women, children and even unborn children. They appear to have been buried there during the invasion of Chosroes II in AD 619 when Persian troops torched Alexandria and ravaged the region.

Amphora Kiln Szymańska & Babraj 2004a: 53-56, 6164, 2004b: 26-28, 2005: 63-67 . A complete surprise awaited archaeologists digging under the church apse. At a depth of 1.80 m below the preserved tops of the walls, a grate belonging to a large amphora kiln was subsequently discovered. The part of the church intended for liturgical practices used this earlier kiln as a foundation ig. 8 . More of the grate appeared once the floor of the burial chambers under the apse had been cleaned.

The rich interior decoration of the basilica included several fragmentary column shafts and Corinthian capitals deco-

Fig.8.M area. Basilica. Amphora kiln.V iew from the west ( photo by J. ucy .

82

K. BABRAJ & H. SZYMANSKA: MAREA OR PHILOXENITE? Bibliography

The kiln had a diameter of 8 m, and the thickness of the grate was 0.50 m. ne of only a few discovered in the ile Delta, this kiln still contained upon discovery the last batch of amphorae. These vessels were dated to the 2nd and 3rd centuries AD. The wall of the kiln still stood 0.93 m high and was built of mud-bricks. The structure of the kiln is not unlike others from the period with pillars under the grate pierced with holes to support easy circulation of hot air El-Ashmawi 1998: 55-64 .

Ancient Sources Athenaeus, The Deipnosophists, Ch.B. Gulick transl. , 1927. Cambridge, MA. Athanasios, Apologia contra Arianos, H.G. pitz transl. , 1934. Berlin, Leipzig. Diodorus Siculus I, Library of History, C.H. ldfather transl. , 1989. Cambridge, MA. erodotus Book II, A.B. Lloyd transl. , 1976. Leiden. History of the Patriarchs of the Coptic Church of Alexandria Patrologia rientalis 1.14, B.T.A. Evetts ed. , 1904. Paris. Horace, Odes and Epodes . Rudd transl. , 2004. Cambridge, MA. John Moschus, Pratum Spirituale = .H. Baynes, 1947, The Pratum Spirituale. O rientalia Cristiana Periodoca 13: 404-414. Justinian, Edict Corpus Iuris Ci ilis. Code Justinianus Vol. II, P. Kr ger ed. , 1877. Berlin. Strabo, G eography. In Strabon. Le voyage en Egypte. U n regard Romaine . oyotte & P. Charvet transl. , 1997. Paris. Thucydides, History of the Peloponnesian War, Ch. orster Smith transl. , 1928. Cambridge, MA. Virgil, Eclogues. G eorgics, H. Rushton airclough transl. , 1999. Cambridge, MA.

C onclusions The site of Marea with its Late Antique agglomeration remains a puzzle, generating considerable controversy as to its chronology and even its very name. The identification of the town ruins lying on the southern shore of Lake Mareotis, just 45 km south-west of Alexandria, is amongst the priorities of a team from the Polish Centre of Mediterranean Archaeology of Warsaw niversity and the Archaeological Museum in Krak w, which have conducted excavations here since 2000, based on a concession granted by the Supreme Council of Antiquities of Egypt that covers an area of 19.7 ha. The Polish expedition has concentrated on the extant ruins, investigating the role and importance of this Byzantine town as a religious center built around a huge basilica. During nine field seasons 2000-2008 , three independent architectural complexes have been investigated: the baths with s iyah-well, a funerary chapel, and the Christian basilica. All the structures were dated to the 6th end of 5th through early 8th centuries. A site survey included an inventory of all the ruins discernible on ground level within the limits of the SCA concession see ig. 1 .

Secondary Sources Adam, J.P., 2008, La construction romaine. Paris. Am lineau, E., 1893, reprint 1973 , G éographie de l’É gypte à l ’époque copte. snabr ck. El-Ashmawi, ., 1998, Pottery Kiln and Wine- actory at Burg el Arab. In .- . Empereur ed. , Commerce et artisanat dans l’Alexandrie hellénistique et romaine. ctes du collo ue d th nes organis par le C R le Laboratoire de céramologie de Lyon et l’Ecole franaise d th nes d cembre 55-64. Athens, Paris. Babraj, K., & Szymańska, H., 2004, Die Vierte Grabungssaison in Marea, Aegypten. Grabkapelle und Basilike. K emet 13.3: 61-64. Babraj, K., & Szymańska, H., 2005, nfte Grabungssaison in Marea, gypten: Basilika. K emet 14.3: 63-67. Babraj, K., & Szymańska, H., 2008, iyah. In H. Szymańska & K. Babraj eds. , Byzantine Marea. Excavations in 20203 and 206 , Marea Vol. 1: 85-99. Krak w. Babraj, K., & Szymańska, H., 2008, Funerary chapel. In H. Szymańska & K. Babraj eds. , Byzantine Marea. Excavations in 20 -203 and 206 , Marea Vol. 1: 177-185. Krak w. Blue, L., & Ramses, S., 2005, Lake Mareotis Research Project Preliminary Report from Pilot Survey August 2004. Newsletter of the Archaeological Society of Alexandria February : 5-16. Calderini, A., 1980, i ionario dei nomi geografici e topografici dell Egitto Greco Romano Vol. III, fasc. 3. Milano.

This large port operated in Roman and later Byzantine times and perhaps also in the Ptolemaic era. However, the question of the actual identity of the site remains still uncertain. The current excavated ruins of Byzantine date extend along the coast for 1.5 km. The great jetties are of key importance for the dating of the port. The size of the harbour installations, the small but important quantity of Early Roman potsherds, as well as huge sewage network of substantial build, indicates that a large urban agglomeration flourished on the spot from pre-Byzantine times. Philoxenos may have founded his city prior to its abandonment as an urban center due to the silting up of the sweet-water ile channels supplying the lake. However, evidence suggests that a site existed prior to the establishment of Philoxenite, being located next to existing harbour installations that may have been part of Marea. The recent discovery of the pottery kiln under the apse of the basilica that operated in the 2nd-3rd centuries AD, is unmistakable proof of a flourishing center existing on the spot possibly Marea prior to the Byzantine foundations. Perhaps this site was were the wine that Mareotis was famous for was being produced, and from here the product was being exported to other towns of the Greco-Roman world in amphorae produced at the site. 83

LAKE MAREOTIS: REC

STR CTI G THE PAST

De Cosson, A., 1935, Mareotis eing an ccount of the History and Ancient Monuments of the North-West Desert of Egypt and Lake Mareotis. London D cobert, Ch., 2002, Mar otide m di vale. Des B douins et des Chr tiens. In Ch. D cobert ed. , Alexandrie Médiévale 2, É tudes Alexandrines 8: 127-167. Le Caire. Desanges, ., 1978, P riple du Pseudo-Scylax 90 107 M P riples libiques, 7. In . Desanges ed. , Recherches sur l acti it des M diterran ens aus confins de l’Afrique, Collection de l’É cole Franç aise de Rome 38: 404-405. Rome. Drescher, . ed. , 1946, Apa Mena. A Selection of Coptic Texts relating to St.M enas. Cairo. Empereur, .- ., & Picon, M., 1998, Les ateliers d amphores du lac Mariout. In .- . Empereur ed. , Commerce et artisanat dans l’Alexandrie hellénisti ue et romaine. ctes du collo ue d th nes organis par le C R le Laboratoire de c ram ologie de Lyon et l Ecole fran aise d th nes d cembre 75-91. Athens, Paris. El- alaki, Mahmoud Bey, 1872, Mémoire sur l’antique le andrie ses faubourgs et et en irons d cou erts par les fouilles sondages ni ellement et autres recher ches. Copenhagen. El- akharani, .A., 1983, Recent Excavation at Marea in Egypt. In G. Grimm, H. Heinen & E. Winter eds. , egyptiaca re ernsia II as R misch y antin ische Ägypten. Akten des internationalen Symposions 26.- 30.Se ptember 1978i n Trier, 175-186. Mainz. raser, P.M., 1972, Ptolemaic le andria I. Oxford. Goma , ., 1980, s.v. Kom el-Idris Marea , Lexicon der Äegyptologie III: 647. Grossmann, P., 1993, Die uerschiffbasilika von Hauwar ya-Marea und die brigen Bauten dieses Typus in gypten als Repr sentanten der verlorenen fr hchristlichen Architektur Alexandreias. In N. Swelim ed. , In Memoriam Daoud Abdu Daoud ( Bulletin de la Societé Archéologique d’Alexandrie 45 : 107-121. Grossmann, P., 2002, Christliche Architektur in Aegypten. Leiden. Grossmann, P., 2003, ochmals zu Marea und Philoxenite. Bulletin de la Societé d’Archéologie Copte 42: 13-29. Haas, Ch., 2001, Alexandria and the Mareotis Region. In S. Burns & .W. Eadie eds. , U rban Centers and Rural Context in Late Antiquity: 47-62. Ann Arbor, MI. Kees, H., 1930, s.v. Marea, Mareotis. Real Encyclopedie IV.2: 1676-1678. Kucharczyk, R., 2008, Glass finds from the baths and s iya. In H. Szymańska & K. Babraj eds. , Byzantine Marea. Excavations in 20203 and 206 , Marea ol. : 129-143. Krak w. Lichocka, B., 2008, The early Byzantine coins. In H. Szymańska & K. Babraj eds. , Byzantine Marea. Excavations in 20203 and 206 , Marea ol. : 145-152. Krak w.

Majcherek, G., 2008, The pottery assemblage from the baths and s iyah. In H. Szymańska & K. Babraj eds. , Byzantine Marea. Excavations in 20203 and 206 , Marea ol. : 105-127. Krak w. Malarczyk, D., 2008, mmayad coins. In H. Szymańska & K. Babraj eds. , Byzantine Marea. Excavations in 20203 and 206 , Marea ol. : 153-154. Krak w. Martindale, .R. ed. , 1980, The Prosopography of the Later Roman Empire, II. Cambridge. M ller-Wiener, W., 1967, Siedlungsformen in der Mareotis. Archäol ogischer Anzeiger 82.2: 103-117. Mycielska-Dowgia o, E., & Woronko, B., 2008, Evolution of the natural environment in the region of Marea. In H. Szymańska & K. Babraj eds. , Byzantine Marea.E xcavations in 20203 and 206 , Marea Vol.1 : 17-26. Krak w. Rodziewicz, M.D., 1983, Alexandria and District of Mareotis. G raeco-Arabica II: 199-216. Rodziewicz, M.D., 1988, Remarks on the Domestic and Monastic Architecture in Alexandria and Surroundings. In E.C.M. van den Brink ed. , The Archaeology of the ile elta Proceedings of the eminar held in Cairo 19- 2O ctober 1986 : 267-276. Amsterdam. Rodziewicz, M.D., 1995, Eco-Archaeology of Ancient Alexandria and Mareotis. In A.A. Hussein, M. Miele & S. Riad eds. , Proceedings of the Seminar on G eosciences and Archaeology in the Mediterranean Countries, Cairo o ember : 127-139. Cairo. Rodziewicz, M.D., 1998, rom Alexandria to the West. In .- . Empereur ed. , Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du collo ue d th nes organis par le C R le Lab oratoire de céramologie de Lyon et l’Ecole franaise d th nes d cembre 93-103. Athens, Paris. Rodziewicz, M.D., 2002, Mareotis Harbours. In Ch. D cobert ed. , Alexandrie Médievale 2, É tudes Alexandrines 8: 1-22. Le Caire. Rodziewicz, M.D., 2003, Philoxenit Pilgrimage Harbor of Abu Mina. Bulletin de la Societé Archéologique d’Alexandrie 47: 27-47. Szymańska, H., & Babraj, K., 2004a, Marea. ourth season of excavations, 2003. Polish Archaeology in Mediterranean V: 53-63. Szymańska, H., & Babraj, K., 2004b, The ancient port of Marea, Egypt. our seasons of excavations by the Polish Archaeological Mission. Minerva April 2004 : 26-28. Szymańska, H., & Babraj, K., 2005, ouilles arch ologiques de Marea en gypte. Saisons 2002-2003. Archeologia LV: 119-130. Szymańska, H., & Babraj, K., 2005a, Marea. ifth Season of Excavations, 2004. Polish Archaeology in Mediterranean VI: 43-54.

84

K. BABRAJ & H. SZYMANSKA: MAREA OR PHILOXENITE? Warne, A.G., & Stanley, D. ., 1993, Late uaternary Evolution of the orthwest ile Delta and Adjacent Coast in the Alexandria Region, Egypt. J ournal of Coastal Research 9.1: 26-64. Winnicki, .K., 2006, Der libysche Stamm der Bakaler in pharaonischen, persischen und ptolem ischen gypten. Ancient Society 36: 135-142. Wipszycka, E., 2008, Remar ues sur l identification de Piloxenité à la lumiè re de fouilles récentes. Paper delivered at the Third Conference on Medieval Alexandria at the Centre culturel fran ais in Alexandria, 8-10 ovember, 2002 unpub. .

Szymańska, H., & Babraj, K., 2006, Polish Excavations in the Basilica at Marea Egypt . Bulletin de la Societé Archéologique d’Alexandrie 45: 107-117. Szymańska, H., & Babraj, K., 2008, Baths. In H. Szymańska & K. Babraj eds. , Byzantine Marea. Excavations in 202003 and 206 , Marea ol. : 27-83. Krak w. Szymańska, H., & Babraj, K. eds. , 2008, Byzantine Marea. Excavations in 20203& 206 , Marea I. Krak w. Timm, S., 1984, Das christlich-koptische Ägypten in arabischer Z eit. Wiesbaden. Timm, S., 1988, Teil 4 M - P , Reihe B Geisteswissenschaften o. 41/4, Wiesbaden.

85

LAKE MAREOTIS: REC

STR CTI G THE PAST

86

M.-F. BOUSSAC & M. EL AMOURI: LAKE STRUCTURES AT TAPOSIRIS

The Lake Structures at Taposiris Marie-Françoise Boussac & Mourad El Amouri

Introduction Among the harbours of Mareotis, which in recent years have been the subject of renewed attention, Taposiris on the north shore is of particular interest. Amongst many features of interest, the city offers the only example, besides that of Alexandria, of a closed basin which allows controlled maritime traffic management ig. 1 . These structures were briefly mentioned at the beginning of the 19th century by Pacho, who in 1824 devoted a few lines to the dam running from east to west ... built in the south of the city , a structure intended, he said, to prevent floods Pacho 1827 and sketched by Coste1 in 1820 ig. 2 . Thereafter, Breccia 1914 , De Cosson 1935: 110-111 ig. 3 , chsenschlager 1979, 1999 and Rodziewicz 2002 , amongst others, have pointed out the peculiarities of the closed basin, speculating on its chronology and its connection with the wall of the Barbarians which closes, to the west of the city, the very narrow spit of land that separates the Mediterranean from Lake Mareotis at this point, and its relation to the causeway which blocked the lake to the south. At some point in their use, both systems acted as locks and seem related, but we do not know if they are part of the same feature, nor do we know during which period they were used.

When the French Archaeological Mission of Taposiris3 was launched in 1998, it was determined that an understanding of the harbour system would be a research priority. What was the chronology of the currently visible structures? To which phases of the site did they correspond? Because of the dimensions of the harbour basin the eastwest artificial levee or causeway is about 1,700 m long , areas that were deemed likely to provide some answers with limited resources were selected for excavation, supplemented with an environmental study and geophysical survey. Since most of the data acquired has already been the subject of articles Boussac 2007, 2009 , this paper will briefly present the harbour structures and the conclusions reached concerning the southern causeway. The east side of the system will be the main focus as this gives the latest possible date for all the studied lake constructions. Ove rvi ew of the Remains The ancient city of Taposiris, located on the south side of the taenia ridge, is organised into three sectors linked by a network of north-south routes: the upper town with the Brescia terrace and the temple, the middle town, and the lower town on the shores of Lake Mareotis. To the west, the Wall of the Barbarians forms the western limit of the city see ig. 1 .

To answer these questions, the American mission from Brooklyn College carried out several soundings in 1975, during a one month campaign. Anxious to locate the harbour area,2 chsenschlager worked on two sectors, north and north-west of the dug-out channel and the causeway. either gave the anticipated results: in the north one A , an elevated area, interpreted as a lake front, proved to be an accumulation of rubble waste from amphorae workshops, as shown thereafter by Empereur & Picon 1998 . In the north-west one C , American archaeologists concentrated their efforts on what initially seemed to be a warehouse along the shore. They described it as a platform which had been redesigned several times the so-called platform building , they did not understand its function, but they dated it to the 3rd century AD. They also uncovered a sophisticated system of water tanks carrying water towards the harbour from a terrace further north. No research was published to specify the chronology, and no interpretation was suggested unpublished .

The topographical survey of the lower town area carried out in 1999 and 2000 ig. 4 , and supplemented in 2009, extends from the Wall of the Barbarians in the west to Plinthine in the east, and highlights the irregular outline of the lake shore. A partially anthropogenic spit of land promontory divides the lake shore into two marshy plains, one of which, to the east, forms a bay which is interpreted as a harbour basin.4 This basin covers an area of approximately 8 ha and is closed off to the south by a causeway approximately 1,700 m long , which extends east-west in the shape of a ridge, but it is not straight as the sketches of Coste 1820 see ig. 2 or De Cosson 1935: 110 see ig. 3 imply. It is also interrupted by at least six openings that may be deliberate or may be related to the disappearance of the retaining walls which in places supported the north slope of the causeway.

3. Since 1998 La Mission ran aise in Taposiris Magna has been supported by the Commission des fouilles of the French Ministry of Foreign and European Affairs. Authorisation has been kindly given by SCA and wee appreciate their assistance. 4. Analysis by C. laux, PhD niv. Aix-Marseille, in 2008, in a PhD thesis on the geomorphology of Lake Mareotis.

1. Thanks to the Biblioth que Municipale Vocation R gionale in Marseille, especially o lle Colombi , who allowed us to reproduce two sketches by Pascal Coste from their archives. 2. Thanks to the American mission, especially M. Venit, who shared their unpublished documentation esp. Preliminary Report of the 1975s eason .

87

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig.1.G eneral map of the site and location of areas under study at Taposiris Magna and Plinthine.© Archaeological Mission in Taposiris.

88

MAFT,F rench

M.-F. BOUSSAC & M. EL AMOURI: LAKE STRUCTURES AT TAPOSIRIS

Fig.2.D rawing by Pascal Coste,m ade in 1820.a2: G eneral plan of Alexandria to Abousir.2b: Detail of plan showing the Temple of Taposiris ( 1) ,t he Tower of Arabs ( )2 ,br idge ( 3) ,e astern j etty ( 4) and causeway across the lake from the levee to the south shore ( 5) .2c : Plan and isometric view ( south) of the bridge of Taposiris.© BMVR Fig.3.M ap of the remains of the ancient city of Taposiris by A. De Cosson,1935 .

89

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig.4.L ocation of the different sectors of the site of Taposiris that were studied along the lake shore.© MAFT

The Elements of the Lake System The Causeway and the Channel Several operations 1998, 2000-2005 were conducted at the west end of the causeway, on both sides of the channel and near the bridge, where the basin forms a bend: it

expands to the nortwest to form a loop against which is aligned by a series of buildings, whose different orientations and especially superposition indicate different phases. This visible state has significantly altered the previous interpretation of the phases: the channel was dug in the first half of the 2nd century AD Boussac 2009 7 in a location where a densely occupied area was previously developed whose architectural design suggests prosperity. This area was used for storage and trade, judging from the quantity of imported amphorae material found there Boussac 2009 .8 n the north shore Sector 1 there is a row of shops, abandoned at the end of the Hellenistic period un-recovered items left in place perhaps as a result of flooding,9 before the digging of the channel. On the other side of the channel Sector 3 , earth that had been dredged from the lake covered buildings last occupied at the end of the 1st century BC and which had been in use since at least the 2nd century BC. The excavated building has the same

5. Breccia 1914. He only mentions the 1 km-long dyke, running parallel to the hills, and a fine bridge whose construction certainly dates to Roman times. It seems clear that the lake extended to Taposiris, and that the dyke blocked the water in a kind of harbour . 6. Prospecting by C. Benech, C RS. See the map in Boussac 2007: 452.

7. See the analysis by S. Marqui in Boussac 2009. 8. Boussac 2009 see the appendix by S. Marqui and the study by K. Senol . 9. Study suggests during the 1st century BC: see the analysis by S. Marqui in Boussac 2009: 137.

Topographic mapping and geophysical survey revealed the density of structures in the lower town, briefly mentioned as walls and ruins by Coste 1820 and ignored by Breccia 1914 .5 Although the causeway that blocked the lake has virtually disappeared only a stretch of about 10 m could be uncovered , its outlet was located near a warehouse built on the causeway. Most importantly, the geophysical survey revealed in the West Bay and south of the bridge many ruins which are currently silted over.6 These results underscore the extent of changes in the landscape, the mobility of the shoreline and the extent of the harbour area.

90

M.-F. BOUSSAC & M. EL AMOURI: LAKE STRUCTURES AT TAPOSIRIS architectural features as the north shops foundations and architectonic elements in stone, mud-brick walls covered with a coating . It was supplied with drinking water by rainwater collected and discharged into a cruciform tank located to the north. The angle of the buildings on both sides of an east-west axis indicates the prior circulation of water in this area, which was later replaced by the channel. The level of the tank indicates that the site chosen to dig the channel was originally an area not liable to flooding.

harbour traffic to pass, while minimizing the size of the humpback bridge. This would partly explain the asymmetry of the two arches, with a humpback passage which best fits the top of the vaults, which are all of different widths and heights approximately 4.5 m and 4 m . Furthermore, the construction of the bridge can only be explained by the existence of the causeway and channel system. A limited investigation 6.2 m by 3.2 m conducted in the foundation pebble bed radier of the north pier of the bridge, to the flood level of the lake, confirms that the construction of the bridge is contemporary with the digging of the channel, however it does not allow precise dating. Indeed, the stratigraphic study of this area shows that the bridge was built over a level that dates back to the Imperial Roman period it confirms that the channel, the bridge and the closed harbour system can be dated to the Imperial Roman period.

The whole area was disrupted by major works after a period of stagnation and neglect. The objective was to create a navigation canal by digging a channel, and ensure access by strengthening and shoring the banks formed. The channel was dug and the waste was deposited to the north and south creating two sedimentary causeways, of unequal length but of a similar configuration the causeway to the north extends for about 150 m, to the south for 1,700 m.

The East Part of the Lake System The lake basin is closed to the east by a built jetty, which was the subject of two campaigns in 2005 and 2006, following a preliminary phase of topographical mapping in 1999 and 2000. The cancellation of the following two campaigns 2008 and 2009 , due to an exceptional rise in the water level, meant that certain points remain to be determined, in particular the chronology: while the date of abandonment is well attested by the coins and ceramics first half of 7th century AD , uncertainties remain concerning the previous phases.

An anthropic embankment was formed by successive heaps of backfill piled up, stabilised by a gypsum screed. The most extensive works were carried out to the west, where the edge of the promontory needed to be cut back. This explains why there are only two artificial embankment slopes in this sector, to the north and the south, and why the causeway ends near the bridge at a height of some 5-6 m and then drops in irregular steps towards the east. The two sides are asymmetrical: the north slope, which follows the dug-out channel is steeper and supported by a series of retaining walls to prevent the collapse of the earth and therefore seal the passage, the south side has a shallower slope.

The jetty is located approximately 1,700 m to the east of the bridge west entrance of the harbour basin below the Temple of Taposiris. Between the end of the causeway and the tip of the jetty, there is an opening which is currently about 100 m wide. This probably allowed boats to enter the harbour basin from the east. We have not yet observed a structure that could act as a checkpoint like the bridge that closes access to the west or like the causeway running north-south which blocked the lake between the harbour complex and the southernmost bank Boussac 2009 .

The Bridge A dressed stone10 bridge11 guards the west entrance of the channel. Its present appearance has hardly changed since Coste s sketch in 1820 see ig. 2-c : 47 m long including the ramps, and 9 m wide. Its maximum height is 2.55 m. It consists of two massive abutments of equal length 6.02 m for the south abutment and 6.55 m for the north abutment . The space of approximately 8.35 m between the two abutments is divided by a narrower intermediate pier 1.20 m . This pier is off-centre compared to the axis of the abutments, thus forming two passages of different widths 4.10 m to the south and 3.05 m to the north which only allowed access to the channel to small boats Bougia 1996 . Study of the structure leads to several conclusions. First of all, the bridge was probably supported by masonry vaults igs. 5-7 .12 The flatness of the natural terrain required the creation of a passage sufficiently high for the

Excavations have identified several structures: the jetty, which is interrupted by three openings two flushes and a water supply channel which cross its entire width two basins, one square, the other in the shape of a hoop, located inside the dock and attached to the jetty, between the north flush and the water supply channel. The East Je tty The east jetty extends north-south and has a visible length of 230 m and a width varying of 5-6 m.13 It ends in the lake and it is possible to follow its extent for a further

10. Rodziewicz 1998: 102 only mentions a structure . In fact, the bridge was necessary to cross the channel for those coming from the south of the lake using the causeway and following the road north. 11. As is the case for most stone buildings in the region, it is calcarenite, white dune limestone, extracted from the quarries on the taenia. 12. This architectural analysis, and the model ig. 7 are the work of students at M2Pro niv. Lyon II, especially R. Brunier and . imenez, under the supervision of T. ournet C RS, IRAA/HIS MA .

13. It is longer than the jetties built on the south shore at Marea 60 and 40 m long at the site referred to as Kibotos Blue & Ramses 2005: 10 and is similar to structures noticed during prospecting on an island of Mariout, at a site which played a major role in the management of lake traffic ibid., Sites 23, 12 .

91

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig.5.V iew of the Taposiris bridge from north.©

MAFT

Fig.6.( above) Plan and elevation of the north pile of the bridge,N ovember 204.L ay out,dr awing and DAO by M. El Amouri.© MAFT

Fig.7.( left) Model and reconstruction of the bridge in Taposiris Magna ( students of the Master Pro-Lyon II) . © MAFT 92

M.-F. BOUSSAC & M. EL AMOURI: LAKE STRUCTURES AT TAPOSIRIS

Fig.8.O verall plan and detailed outline of the eastern j etty.L ay out,dr awing and DAO by M.E l Amouri. © MAFT

20 m or so under water ig. 8 . To the north it gradually disappears first towards the west and then the east. This asymmetry could indicate that the point of attachment of the causeway on dry land is different than at the southern lake end from that at the lake end. urther north, the silting of the area and recent constructions prevent the correct reading of its route. Situated on the axis, about 50 m to the north, is a masonry structure ST900114 which may or may not be related to the jetty.

Fig. 9. The southern end of the eastern j etty. View from north.© MAFT

The jetty ig. 9 is built in blocks of regular size 0.50 x 0.20 x 0.25 m on average arranged in headers, and the design is comparable to that of the harbour structures present at the neighbouring sites at Gamal and useir igs. 10 & 11 . Excavated sections of the eastern jetty revealed four well preserved courses, and two additional courses which have now disappeared. Over its entire length, both sides of the jetty are punctuated with buttresses ig. 12 . These 27 buttresses 16 on the east side and 11 on the west side are maintained on several courses three or four . Every second course has two headers while the alternate blocks are stretchers on their edge. They are on average 0.50-0.60 m wide and protrude by about 0.30 m. They stabilised the entire jetty, while encouraging micro areas of

14. An investigation was carried out in 2006 on this structure. The report on surface ruins shows an angle similar to that of the jetty. If it belongs to the jetty this would alter our interpretation of the extent of the closed harbour basin.

93

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig.10.J

etty at G amal.V iew from north-west.©

MAFT

Fig.1 1.J etty at Q useir.V iew from south-east.©

Fig. . Plan and ele ation of the inner south face of the north ush and buttress. Lay out dra ing and Amouri.© MAFT

94

MAFT

O by M. El

M.-F. BOUSSAC & M. EL AMOURI: LAKE STRUCTURES AT TAPOSIRIS Fig.13.V iew of the north ush during excavation,f rom east. © MAFT

In addition, the east part of the flush is designed in order to circulate as much surface fresh water as possible McCann 2003: 32 16 towards the interior of the harbour basin: a base of grey mortar17 at the base of the projection formed by the paving of the western part, has a slope to the east which hinders the entry of sediment into the harbour basin, while allowing the passage of a stream of fresh water. The western part of the bottom of the flush, paved in worked stone, is thus located more than 0.40 m above the mortar.

sediment accumulation, thus reducing the risk of sapping erosion due to currents and saline seepage.15 It is telling that particular attention was given to the construction of the eastern face of the jetty where the buttresses are more numerous. Indeed one notices a difference in treatment between the east side facing the open lake the outside of the basin that is subject to currents, and the west wall facing the interior of the basin. Investigation of the north flush revealed four courses of blocks to the east whereas there are only two to the west. A row of worked stones and a stone blockage line the east facing, but do not exist on the west side. In addition, the level of the upper beds of the blocks on the east side is slightly higher than that of the west side. These differences in the construction show that the jetty is a rampart against the open waters of the lake located to the east. To the west, the water in the basin is calmer and the construction of the jetty does not require these precautions.

Carved into the faces of the flush conduit on the paved western half, there are several vertical notches with corresponding horizontal grooves in the paving, which allow the recreation of the system of closing and filtering of the flush. It would appear that three of these locking systems could operate simultaneously. They probably held filter sluice gates or simple marteliè res type gates, which allowed the control of the flow of water according to currents and seasons while limiting the entry of sediments to the basin.18 Two twin notches have no corresponding grooves which implies the redesign of the paving and locking systems. A lightweight and retractable system to cross the flush probably existed for passage across the jetty, but no archaeological evidence shows its design.

The North Flush There are two flushes, designed to bring as much running water as possible into the basin and thus prevent silting one near the south tip of the jetty, the other to the north, more than 70 m from the present shore line, only the second flush has been the subject of investigation igs. 12 & 13 . Its east opening is shaped like a funnel with two facing walls built in a quadrant. The west opening of this passage is straight and 1.2 m wide. The axis of the flush is not perpendicular to the jetty but shifted 15 degrees to the north, which probably gives an indication of the direction of the wind and/or dominant currents in the region during antiquity.

16. Mc Cann 2003: 32 the author states that the constant flow of water helps control the temperature, oxygenation and salinity of the water, adding that the proximity of a source of fresh water is an advantage. 17. Grey mortar seems to have been chosen for the bed of the east entrances of the north flush and the water supply channel. However, pink hydraulic mortar is used around the upper areas of the north flush and water supply channel. 18. Does this system indicate seasons with high levels of sedimentation Are these seasons connected to ile floods Indeed, the ile is joined to Lake Mareotis via canals but underground water must also have raised the level of the lake during floods, bringing a great deal of sediment.

15. There is also a risk of erosion of the sides of the causeway facing the channel this was prevented by the stone constructions, the ramp and the retaining wall.

95

LAKE MAREOTIS: RECONSTRUCTING THE PAST Fish Farming Systems ne of the most interesting findings is the discovery of an original system for breeding/keeping fish. At the current stage of excavations, this consists of a water supply channel which crosses the jetty, a hoop-shaped basin and a fish-tank attached to the jetty inside the harbour basin see ig. 8 . This complex, which has possibly not been entirely uncovered, and which shows signs of repairs, indicates intensive fish farming Lafon 2001: 161, 165 , i.e. breeding in artificial pools. rom a typological point of view, these installations built on the lake are one of the types of fish farming installations analyzed in various recent studies, mostly devoted to the Roman world Kron 2008 : basins built by digging in the sand and masonry work Type III see below Lafon 2001: 171 . Plato tells us that fish farming was practiced on a large scale on the banks of the Nile and papyri from the Hellenistic and Roman eras reveal much information about ichthyotropheia, piscinai and other apodocheia Chouliara-Ra os 2003 Brewer & riedman 1989 of which our system is an example. The region of Canope is home to several piscinae dugout of the rocks, some of which have long been known Breccia 1926 Abd El-Maguid forthcoming but the installation uncovered in Taposiris is unique to this day around the shores of Lake Mareotis Lafon 1998: 573 .19 However, the association of harbour installations and fisheries is not surprising and the facilities found at Taposiris are similar for example to those found at Kenchreai dating to around AD 80 Scranton, et al. 1978 :20 basins are created near the jetty Rothaus, et al. 2008 Lafon 2001: 162 21 and connected to one another and with the sea through channels.

tion basin. As with the flush, the grooves holding the filter sluice gates or locking marteliè res were modified during the rehabilitation of the water supply channel. The walls of the water supply channel are built in worked stones bound with pink mortar. The blocks which form the entrance to the channel are chamfered to facilitate the entry of water. It is a simple funnel system. At the entrance, the bottom of the channel, like the bottom of the flush, is made of grey mortar sloping towards the east. The water supply channel, like the flush, might have been covered by a crossing system, but we did not find any trace of it, although we noticed the last course on its north wall was slightly battered. The western part of the south interior face of the channel ig. 14 appears to have undergone repair with small unworked stones which contrast with the usual blocks of worked stone of the walls of the channel. ther repairs are visible in the south wall of the hoop-shaped basin. The meticulous cleaning of the water supply channel led to the discovery of seven coins wedged between the paving stones of the channel, and thus protected from the cleaning work ield nit 9240 . ive of them belong to the Later Roman Empire AE4 and two are Byzantine dodekanoummia of the 6th-7th century AD, including one dated to Heraclius AD 610-641 .23 Thus, the water supply channel must have functioned at least from the 4th century to the 7th century AD.

The Water Supply Channel The water supply channel crosses the jetty perpendicularly over its entire width igs. 14 & 15 . This channel is 0.70 m wide, 6 m long and 0.85 m deep at the point where the third course is still preserved. It consists of a small decantation basin, a ledge and three closing systems with notches carved opposite one another, 0.50-1 m apart: they were used to create a kind of lock, like in Kenchreai Scranton, et al. 1978 .22

Its period of abandonment is suggested by traces of blocking up. Chamfered blocks the same size as those still in place, and a block with a groove identical to that found on the blocks still in place, were discovered during excavations. These blocks thus come from the courses of the water supply channel. The chamfered blocks were placed at the entrance of the water supply channel to block it: they fit perfectly in the entrance of the water supply channel while keeping one layer horizontal, indicating that they were probably deliberately placed there. The backfilling of the channel thus seems a deliberate action indicating the stopping of the water supply or even of fishing, but ensuring the continued functioning of the jetty.

One of these systems of notches still has its corresponding grooves cut in the bottom of the paving blocks ig. 16 . This is the first paving block to the east and acts as a dam to stop the sediment accumulated in the small decanta-

The Hoop-shaped Basin The west end or exit of the water supply channel flows into a hoop-shaped structure about 6 x 3 m the base of which is paved igs. 8, 17 & 18 . It is built in blocks of worked stone, arranged in a single row and preserved to two or three courses. The blocks are arranged in stretchers and their bases are covered with a hydraulic coating that forms a slight slope on the inside edge of the basin see ig. 18 . This poorly preserved coating is only applied at the junction between the base pavement and the elevation.

19. Lafon 1998: 573, n. 3, notes that coastal speculation has caused the disappearance of many installations in the western Mediterranean over the past fifty years. Lake Mareotis is also threatened by intensive property speculation: during our first works on the east jetty area, it was divided up into plots of land, shown by boundary makers. 20. Scranton, et al. 1978. Their construction dates back to around AD 80. 21. Also see Lafon 2001: 162, n. 89, for Cosa: the author notes that in the port of Cosa, “canals joining the lagoon which has various dams, flow directly into the port, and form part of the same project . 22. Scranton, et al. 1978. ote the presence of several sluices per channel, necessary to form intermediary basins and fish traps: the dimensions of the basins 1 x 0.7 m are similar to those found at Taposiris.

23. Study by T. aucher, Appendix 2, infra.

96

M.-F. BOUSSAC & M. EL AMOURI: LAKE STRUCTURES AT TAPOSIRIS

Fig.14.L ay out,e levations and sections of north and south walls of the water supply channel through the eastern j etty. Lay out,dr awing and DAO by M.E l Amouri.© MAFT signed so as to create a slope down towards the fish-tank basin.

The hoop-shaped basin, fed with water through the supply channel, is closed by a double sluice gate system, to the north-east and south: to the north-east, vertical and horizontal grooves were cut into the blocks forming the junction between the channel and the basin. To the south, the funnel-shaped end is fitted with the same system, indicating the method of control of the flow of water into the fish-tank, a large square structure enclosed by walls with double facing ig. 19 . The junction was also de-

The north wall of the basin is an extension of the north wall of the water supply channel and seems to be a renovation or addition. Indeed, this wall is not linked to the channel and is built with smaller, roughly hewn stones, some of which are reused. Perhaps the link with the channel did not initially exist, at least in this form. 97

LAKE MAREOTIS: RECONSTRUCTING THE PAST The Fish-tank The fish-tank is a square basin whose interior dimensions are about 8 m x 8 m. It features walls angled north-south and east-west, all of different form. The west face of the east jetty serves as its east wall. The northeast corner of the basin corresponds to the end of the west face of the jetty. Investigations carried out on the extension of this face showed that no repairs had been made at this location and that the jetty had its original form. It appears that the south and west walls of the basin were entirely built under water. The south wall of an average width of 1.5 m, consists of two parallel faces built in stone and filled with green clay sediment 9221 to the east and 9225 to the west . This wall has an opening, built on at least two courses and equipped with a sluice gate system to control the flow of water. The proximity of the flush installed in the jetty provides a fresh water supply to this southern sluice gate Leatham & Hood 1958-59: 265 .24 The position of this pool is therefore also chosen because of the flow of fresh water into he harbour basin. Sluice G ates and Water Flow A system of fresh water flow was set up between the various structures composing this fishery. The flow or the blocking of water, or at least of fish in the case of filter sluice gates, is ensured by a system of sluice gates or marteliè res type gates. These are all designed in the same way: the vertical grooves cut in the facing blocks allow their positioning across passages, while horizontal grooves cut in the paving slabs ensure a good grip. Some of these gates must have let water filter through to ensure a good regeneration of water in the fish tank. However, no archaeological evidence indicates their construction material.25 The size between 0.10-0.20 m wide for the different grooves in the facing or paving blocks suggests that at least the external framework of the sluice gates was made of wood. They may have been solid and entirely made of wood, or with openings, made with a wooden frame and metal grid.

Fig.15.W ater supply channel through the eastern j etty opening into the hoop-shaped basin.V iew from east. © MAFT Fig.16.G rooves and notches in the water supply chanel. View from north.© MAFT

The water from outside enters from two openings located to the south and north of the whole system: In the south, fresh water arrives through the flush located in the jetty. It enters the harbour basin and indirectly feeds the fish tank by the passage located in its south wall. It is closed by a sluice gate installed upstream of the paving inclined towards the fish-tank. In the north, the water supply channel feeds fresh water into the hoop-shaped basin. The arrival of water is regulated by four successive sluice gates, which leave water to flow towards the hoop-shaped basin while preventing fish from escaping. The frequency of these sluice gates also facilitates maintenance work cleaning, repairs, etc. . A rearrangement may explain the position of the sluice gate located to the west end of the water supply channel, which opens directly onto the northeast corner of the hoop-shaped basin. The space partitioned off by this sluice gate and the one before it, is the largest in the channel. The last sluice

24. Leatham & Hood 1958-59: 265: n. 8, concerning the need for a flow of fresh water into the fish tanks the authors cite . Hortensius criticising M. Luculus whose fish tank does not have enough. 25. Sciallano 1997: 17 notes bronze sluices with holes, like those recommended by Columella viii.17,6 because they are fine nets which allow water to pass through while trapping the fish Leatham & Hood 1958-59: 265, n. 9 . But other systems exist in wood or composite wooden frame and metal grid for example , or in stone like at Mochlos Leatham & Hood 1958-59: 275 and plate 63 a .

98

M.-F. BOUSSAC & M. EL AMOURI: LAKE STRUCTURES AT TAPOSIRIS Fig.17.V iew of the hoopshaped basin,t he water supply channel and the fish tank. ie from north. © MAFT

Fig.18.T he pavement in the hoop-shaped basin.V iew from north.© MAFT

Fig.19.T he opening between the hoop-shaped basin and the fish tank. ie from south west.© MAFT

99

LAKE MAREOTIS: RECONSTRUCTING THE PAST gate is fitted across and at an angle to the axis of the channel and is therefore directed towards the centre of the loop, which may also be due to a repair. This all suggests that this sluice gate dates from a later date.

9216/9220 filling the hoop-shaped basin, principally uncovered storage containers Amphorae LRA 1, 4 LRA and LRA 5/6 , indicate a final phase of use of the fish tank during the first half of the 7th century. This hypothesis is supported by the absence of Egyptian amphorae Egloff 167, well documented on Alexandrine and Mariout sites from the mid-7th century. Similarly, among the coins wedged between the paving stones on the channel bed, there are two Byzantine dodekanoummia from the 6th-7th century AD, including one dated to Heraclius AD 610-641 see Appendix 2 .

The passage between the hoop-shaped basin and the fishtank, which is also blocked by a sluice gate, is equipped with the same system. Thus, all these adjustments help to prevent the basins from silting up while encouraging the constant flow of water, which is a key element for the survival of fish in captivity Sciallano 1997: 17 . urthermore, the shallow and muddy bed is the perfect environment for fish farming Lafon 2001: 159 .26

This last phase of use of the system also relates to the last phase of the warehouse excavated in 2000 on the causeway, strategically located near the causeway which crosses the lake Boussac 2009: 129 . The finds included three Heraclius dodekanoummia and late amphorae LRA 1, 4, 5/6 and 7 .

The Fish The fish species identified in the ield nit related to the use of the fish tank 9215 and 9220 in particular at this stage of the study are mainly marine fish or fish which live in brackish water Lafon 1998: 575 .27 A preliminary diagnosis carried out using photographs to visually identify the species28 in advance of a comprehensive study, indicates the presence of sturgeons and garfish Belone belone . At the present stage of work, all the excavated structures seem too small to have housed real breeding farms Guest-Papamanoli 1986: 301 . We suppose it was for the temporary storage of live fish, like at Kenchreai.

The intensity of lake activities during late antiquity corresponds to what the texts and archaeology tell us of the role of Taposiris during this period: the identification in 2009 of late antique thermal baths in the middle town, comparable to those at Marea, further strengthens these findings. However, nothing is yet known about the Hellenistic installations, although we are collecting more data on the city from the 2nd - 1st century BC, and even from the end of the 3rd century: the temple, according to Dr. . Hawass, was founded by Ptolemy IV, suggesting that the city s development is linked to or accelerated by this royal intervention Hawass 2008: 29 . However, it implies nothing about lake constructions. evertheless, in the west, the area covered by the backfill from the digging of the channel in the first half of the 2nd century AD Boussac 2009: 137-141 , revealed many imported amphorae Rhodians and Cnidians which show the intensity of trade from the outside via Alexandria, at least between the mid-2nd century BC and the early 1st century BC. Although their fragmentation prevents us from being precise, the site where these containers were found, close to the shore, suggests that we are not dealing with land transport, but transport on the lake Bernard 2009 . This storage area suggests specific installations of which nothing remains.

C hronology In the east sector of the harbour system, the various phases of installation of the fish farming facilities and the chronological link between the visible system and the construction of the jetty are currently difficult to specify. Whereas the west harbour system bridge and channel can be dated to the 2nd century, the state of visible installations in the east remains unclear: the traces of repairs in the system between the hoop-shaped basin and the channel do not make it possible to date its construction, nor that of the visible jetty. All that can be said about the jetty is that it cannot be later than the 4th century. The presence in the filling of the south wall of the fish tank, of amphorae remains LRA 1 and fragments of containers LRA 4, some of which are dated between the last third of the 5th century and the middle of the 6th century,29 provides an approximate dating for its construction terminus post quem .

othing is known about the structures from the beginning of the Imperial era: in describing Taposiris as the gate eisbolè ) of Egypt in his Letter to the Alexandrians in AD 41, Claudius refers to the fiscal and military role of the city, which was probably associated with specific facilities Boussac 2001 . However, the oldest phase of the harbour that we were able to identify is that which includes the digging of the channel, which is not earlier than the first half of the 2nd century AD Boussac 2009 .30 If one considers that the warehouses and the fish farming facilities operated until the 7th century, then one understands to what extent the intermediate stages of evolution still escape us.

However, ceramic and monetary data provide consistent dates for the last phase of use of the system: ceramic material discovered in the last archaeological layer

26. Lafon 2001: 159: breeding requires very sheltered water with a shallow sandy or silty bed: lagoon shores provide the most favourable environment”. 27. Lafon 1998: 575. The author insists that brackish water is favourable for breeding as it helps the development of young fish and attracts fish. 28. S. Cravhino, pers. comm. PhD niv. Aix-Marseille I . 29. Study by D. Dixneuf, I A . See Appendix 1, infra.

30. See S. Marqui s study in Boussac 2009.

100

M.-F. BOUSSAC & M. EL AMOURI: LAKE STRUCTURES AT TAPOSIRIS

Appendix 1 The C eramics in the Harbour Area: the Eastern Je tty ( Sector 9)

Fig. 20. Main types of ceramics found in Taposiris Magna ( Sector 9) during the Late Roman era.

Archaeological excavations carried out in the area of the harbour basin, and more precisely on the eastern jetty Sector 9 uncovered a few ceramics often very fragmentary and eroded. They amount, except for fragments of bodies, to more than 243 sherds, or 119 individuals.31 However, most of them, i.e. more than 66% of the ceramic assemblage MI 79 , consist of storage and transport containers, both locally produced and imported, characteristic of the Late Roman period, dating which is confirmed by the study of coins by Thomas aucher see Appendix 2 this paper .

pereur & Picon 1989: 236-243 ,33 and were used for the storage and transport of wine, and possibly olive oil. Fragments found in Area 9 ig. 20, a are related to the B1 type according to the classification of Late Eastern Amphorae established by Dominique Pi ri 2005 . This form, generally attributed to the 6th and 7th centuries, is characterized by thick rolled rim an edge to mid-neck recalls the moulding of the previous type Pi ri 2005: 75 . The handles are now almost unribbed. The clay is rather dense and a general trend emerges: marl clay, fine texture and a red to light brown section. Inclusions are of large size, and consists mainly of grains of quartz, some white and sometimes red.

Delphine Dixneuf, I A , Cairo

1. The Imported Amphorae and C eramics he LR mphorae32 Among the imported containers at the site, LRA1 amphorae are the second most common type MI 10 . This evidence finds parallel in all Late Roman Egyptian sites, from the second half of the 4th century until the mid-7th century, or slightly beyond. These jars were produced on the southern coast of Turkey, Cyprus and Rhodes Em-

he LR mphorae LRA 4 productions from southern Palestine represent 32.7 of MI, i.e. 39 individuals. The general form of these amphorae ig. 20, b is morphologically close to Pieri’s type B and dated between the middle of the 6th and the 7th century Pi ri 2005: 106-107 . Many of these amphorae have a sandy and gritty marl clay, with a medium dense fabric and contain several grains of quartz, some grey and white particles of various sizes. The section is usually homogeneous, from buff to orange. Shaping and surface treatment are quite coarse, and clay accretions are frequently observed on the neck. The macroscopic charac-

31. All the sherds are analysed using the criteria of the clay and shape to determine the M I Minimum umber of Individuals by context first, then for the whole sector. Rims, handles, bottoms and bodies fragments were counted. NMI or the global minimum, estimated for the sector, is the highest number among the different morphological features, except the bodies. Regarding the amphorae, the number of handles is divided by two. 32. The term Late Roman Amphorae LRA 1 to 7 refers to the typology of Late Eastern amphorae established by Riley 1981: 85-122 for the material found in Carthage.

33. However, only two workshops have been excavated in Cyprus: Paphos and iggy. See Demesticha & Michaelides 2001: 289-296 Demesticha 2003: 469-476.

101

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig. . Ceramics from the filling of the hoop-shaped basin ( FU 9216/ 920) .

teristics help us to locate the region of production in southern Palestine, especially in the area of Gaza, Ashkelon and Ashdod.34 However, it is interesting to note the similarity between the clay of Palestine amphorae and ceramics manufactured from Mareotic clays.

Sector 9 are morphologically close to Pi ri’s “bag-shaped amphora, type 4 Pi ri 2005: 117, fig. 76 . It is a globular amphora of small size the neck is short and cylindrical or convex, ending with a small round and/or flared rim, with a sinuous profile ig. 20, c . n the Kellia site, rançoise Bonnet-Borel sets the appearance of these containers during the second quarter of the 7th century Bonnet 1983: 442 , however it is not impossible that they started as early as the late 6th century. They are still present during the Fatimid era, with no major change in the shape Vogt 1997: 256, 258 . As regards the identification of site productions, only one group of clay could so far be identified: marl clay, medium density fabric, sandy the section is usually brown to buff. The inclusions consist of several grains of quartz, some grey and white particles of various sizes. The macroscopic examination of the amphorae, which can logically be assumed to be local or at least regional productions, questions the distinction between Palestinian and Mareotic productions. Thus, two hypotheses can be put forward: they are either Egyptian or Palestinian containers exported with the LRA 4 Amphorae. nly macroscopic and petrographic analysis would answer this question.

In addition to transport containers of types LRA 1 and LRA 4, a few fragments of amphorae and ceramics could not be identified with the exception of a fragment of the body of an African sigillata and two elements of Cypriot sigillata. It is a ring-base low bottom and a rim which seems to correspond to Hayes 9B shape and date from late 6th to late 7th centuries Hayes 1972: 378-382 . 2. Egyptian Production Egyptian Production in Marl Clay With the exception of 12 residual fragments MI 6 of AE 3 amphorae, marl clay productions are mainly illustrated by globular or ovoid containers, more commonly referred to as “bag-shaped”, which were produced in many eastern Mediterranean sites, more specifically in the Levant, Palestine and Egypt MI 15 . Containers found in

Quite logically, common ware is illustrated primarily by marl clay products, some with alluvial content, manufactured without any doubt in Mareotic workshops, on the site

34. or a state of question on LRA 4 amphorae and bibliography, see Dixneuf 2005: 54-62.

102

M.-F. BOUSSAC & M. EL AMOURI: LAKE STRUCTURES AT TAPOSIRIS

Appendix2 The C oins from Sector 9

or at least nearby Majcherek 2001: 60-64 . The shape repertory includes dishes with rounded rim ig. 20, d , some of them with a high-placed carination in their external wall, cups with flattened carination ig. 20, e , fragments of storage jars and vases for liquids and some cooking pots ig. 20, f .

Thomas Faucher, Sorbonne, Paris A number of coins whose study is currently in progress were found during the excavations in Sector 9. After some preliminary restoration, it is possible to make a few remarks here.

Egyptian Production in llu ial Clay These productions from the Delta and/or the ile Valley are only a small part of the ceramics: 17 individuals, i.e. 14.2 of all ceramics identified. Apart from some fragments of three AE amphorae, a bottom and a carinated shoulder of a wine LRA 7 container ig. 20, g have been identified. Common wares are illustrated by bowls ig. 20, h ,35 dishes with rounded rims, and two cooking pots with a C shaped rim dated at the Kellia to the 7th century Egloff 1977: 103, Pl. 52 no. 4 & 6, type 138 and to the first half of the 7th century in Baouit.36

Most of the coins belong to the Late Roman and Byzantine periods. o coin dating to earlier than the 4th century AD was found, whether in or out of stratigraphical context. All the Late Roman coins 12 items are small denominations ca 10 mm commonly called AE4 AE for bronze and 4 for the size . The identification of these coins is not possible until the restoration is completed but it is unlikely that more information about the type and date of these coins will be gained, for two reasons. First, the proximity of the lake water has facilitated the corrosion of these coins which are often in poor condition. Secondly, these coins, commonly dated to the 4th-5th centuries, were often molded coins, unlike struck coins, produced more or less legally to overcome the lack of currency. This technique allowed for the production of a large number of coins in a short time, but makes coins almost illegible, even when they are just produced. Therefore, it is logical that their circulation and subsequent deposition in the ground has erased any information.

3. The Assemblages he Fishery Comple The southern wall of the fish-tank consists of two faces with a filling of green clay 9221/9225 . ew ceramics were found in the filling but they provide a chronology for the construction of the fish-tank and probably the whole complex. They consist of bodies of some LRA 1 amphorae and fragments of LRA 4 containers including a bottom and a rim close to Pi ri s B1 form this form is dated from the last third of the 5th to the middle of the 6th century Pi ri 2005: 105-106 .

In addition to these Roman coins, there are also three dodekanoummia a coin of 12 noummion easily recognisable by the letters IB on the reverse, produced using the same molding process. Among the latter, were identified one coin from the reign of ustin I 518-527 and a second from the reign of Heraclius 610-641 . It is difficult to give a precise date for molded coins since the coin might have been molded later than the date of production of the model.

The ceramic material discovered in the last archaeological layer 9216/9220 of the hoop-net shaped basin consists mainly of storage containers, i.e. LRA 1, LRA 4 and LRA 5/6 Amphorae, and fragments of common wares which might be linked to food consumption by the fishermen ig. 21 . Thus, this material attests to a final phase of use of the basin during the first half of the 7th century. This hypothesis is confirmed by the absence of Egyptian Egloff 167 amphorae, well documented on Alexandrine and Mariout sites from the mid-7th century onwards.

The Alexandrian excavations emphasised the importance of this phenomenon whose magnitude was often underestimated in the past. This manufacturing technique is attested as early as the Ptolemaic era and continues in Roman and Byzantine periods despite its apparent restricted use during the Early Roman Empire in any case it is difficult to say if it was outlawed by the state and successfully repressed.

To conclude, the study of Late Roman pottery discovered at Sector 9 offers some interesting insights into the commercial activities and trade in the harbour area, which was the western customs of Alexandria on the Mareotis Lake since Roman times if not before. In addition to local marl clay productions, the repertory includes several imports, mainly of wine containers from the eastern Mediterranean and some ceramics from the Delta and/or the ile Valley, reflecting the economic vitality of the region and fishing activities.

When fully restored, the coins from Sector 9 will hopefully provide a more precise datation and help discern more accurately the different periods of use of the eastern jetty area.

35. ne should tentatively compare this rim fragments with a series of dishes found at Kellia and dated to the first half of the 7th century however these are of alluvial clay and with a painted decoration on white slip. See Bonnet 1994: 378-379, fig. 231 o. 138 . 36. npublished material.

103

LAKE MAREOTIS: RECONSTRUCTING THE PAST Bibliography Bernard, S.G., 2009, Alexandrian Tainiai and Land Traffic Patterns: A ote on the Amnesty Decree P. Teb. I 5 in Light of the Topography. Z eitschrift für Papyrologie und Epigraphik 168: 265-270. Blue, L., & Ramses, S., 2005, Lake Mareotis Research Project Preliminary Report from Pilot Survey August 2004. Newsletter of the Archaeological Society of Alexandria 15( February) : 5-16. Bonnet, ., 1983, Poteries, verres, monnaies, d cors et inscriptions. In R. Kasser & .-M. Al s eds. , Survey archéologique des K ellia ( Basse-É gypte) . Rapport de la campagne 198: 423-480. Louvain. Bonnet, ., 1994, Le mat riel arch ologique r colt en 1977, 1982 et 1983 aux ou o r er-Roub iy t. In R. Kasser & . Bonnet eds. , Explorations aux Q ouç oût er-Roubâ’ î yât . Rapport des campagnes 1982 et 1983 : 349-406. Louvain. Bougia, P., 1996, Ancient bridges in G reece and coastal Asia Minor. PhD thesis, niversity of Pennsylvania. Boussac, M.- ., 2001, Deux villes en Mar otide: Taposiris Magna et Plinthine. Bulletin de la Société franç aise d’égyptologie 150: 42-72. Boussac, M.- ., 2007, Recherches r centes Taposiris Magna et Plinthine, gypte 1998-2006 . Comptes rendus des séances de l’Académie des inscriptions et belles-lettres ( j anvier-mars) : 445-479. Boussac, M.- ., 2009, Taposiris Magna: la cr ation du port artificiel. In . Dumasy & . ueyrel eds. , Archéologie et environnement dans la Méditerranée antique: 123-142. Paris. Breccia, E., 1914, Alexandria ad Aegyptum. Bergamo. Breccia, E., 1926, Monuments de l’É gypte gréco-romaine, I: Le rovine e i monumenti di Canopo. Bergamo. Brewer, D. ., & riedman, R. ., 1989, Fish and fishing in ancient Egypt. Warminster. Chouliara-Ra os, H., 2003, La pê che en É gypte à la lumiè re des papyrus grecs. Ioannina. De Cosson, A., 1935, Mareotis: Being a Short Account of the History and Ancient Monuments of the NorthWestern Desert of Egypt and Lake Mareotis. London. Demesticha, S., 2003, Amphora production on Cyprus during the Late Roman Period. In C. Bakirtzis ed. , De Rome à Byzance; de Fostat à Cordoue. É volution des faciè s céramiques en Méditerranée ( Ve – IX e siè cles) , actes du VIIe congrè s international sur la céramique médiévale ( Thessalonique, 11- 16 octobre 19) : 469476. Ath nes. Demesticha, S., & Michalides, D., 2001, The Excavation of a Late Roman 1 Amphora kiln in Paphos. In E. Villeneuve & P.M. Watson eds. , La céramique byzantine et proto-islamique en Syrie-J ordanie ( IVe - VIIIe centuries AD. J .- C.) . Actes du colloque d’Amman ( 3- 5 décembre 194) : 289-296. Beirut. Dixneuf, D., 2005, Production et circulation des biens Gaza durant l Antiquit tardive: le t moignage des amphores. In C. Saliou ed. , G aza dans l’Antiquité Tardive. Archéologie, rhétorique et histoire: 54-62. Salerne. 104

Egloff, M., 1977, K ellia. La poterie copte. Q uatre siè cles d’artisanat et d’échanges en Basse É gypte. Genve. Empereur, .- ., & Picon, M., 1989, Les r gions de production d amphores imp riales en M diterran e orientale. Amphores romaines et histoire économique: dix ans de recherché. Actes du colloque de Sienne ( 224m ai 1986) : 236-243. Rome. Empereur, .- ., & Picon, M., 1998, Les ateliers d amphores du Lac Mariout. In .- . Empereur ed. , Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athè nes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole franç aise d’Athè nes, 11- 12 décembre 198: 75-88. Athens, Paris. Guest-Papamanoli, A., 1986, Arch ologie, ethnographie ou ethnoarch ologie des ressources marine de sites cotiers. Le cas de la pê che aux muges dans les lagunes de la Gr ce ccidentale. In . Gerard ed. , VIè mes rencontres internationales d’archéologie et d’histoire d’Antibes. L’exploitation de la mer de l’Antiquité à nos j ours. II: La mer comme lieu d’échanges et de communication : 281-303. Valbonne. Hawass, ., 2008, In search of Cleopatra and Mark Antony. Horus August/September : 26-29. Hayes. .W., 1972. Late Roman Pottery. London. Kron, G., 2008, Reconstructing the Techniques and Potential Productivity of Roman Aquaculture in the Light of Recent Research and Practice. In E. Hermon ed. , Vers une gestion intégrée de l’eau dans l’empire romain: 175-185. Rome. Lafon, ., 1998, Piscinae et pisciculture dans le bassin occidental de la M diterran e. J ournal of Roman Archaeology 11: 573-582. Lafon, ., 2001, Villa Maritima, recherches sur les villas littorales de l’Italie romaine ( 3è me s. av. J .- C. / 3è me s. ap. J .- C.) . Rome. Leatham, ., & Hood, S., 1958-59, Submarine Exploration in Crete, 1955. The Annual of the British School at Athens 53-54: 263-280. El-Maguid, A., forthcoming, Maamourah Surveys: Preliminary report. In H. Tzalas ed. , Tropis VIII: Eighth conference on ship construction in antiquity, Hydra, G reece, August 20 . Athens. Majcherek, G., 2001, Marea 2001: ote on the Pottery. Polish Archaeology in the Mediterranean 13: 60-64. McCann, A.-M., 2003, he Roman Port and fishery of Cosa,a s hort guide. Rome. chsenschlager, E.-L., 1979, Taposiris Magna: 1975 season. First International Congress of Egyptology: 503-506. Berlin. chsenschlager, E.-L., 1999, Taposiris Magna. In K.A. Bard ed. , Encyclopedia of the Archaeology of Ancient Egypt: 759-761. xford. Pacho, .-R., 1827, Relation d’un voyage dans la Marmarique et la Cyrénaï que. Paris. Pi ri, D., 2005, Le commerce du vin oriental à l’époque byzantine ( Ve – VIIe siè cles) . Le témoignage des amphores en G aule. Beyrouth.

M.-F. BOUSSAC & M. EL AMOURI: LAKE STRUCTURES AT TAPOSIRIS Riley, .A., 1981, Pottery from the Cisterns 1977.1, 1977.2 and 1977.3. Excavations at Carthage 19 7 conducted by the U niversity of Michigan 6: 85-122. Rodziewicz, M.D., 1998, rom Alexandria to the West by land and waterways. In .- . Empereur ed. , Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athè nes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole franç aise d’Athè nes, 11- 12 décembre 198: 93-103. Athens, Paris. Rodziewicz, M.D., 2002, Mareotic harbours. In C. D cobert ed. , Alexandrie médiévale 2: 1-22. Le Caire. Rothaus, R.M., Reinhardt, E.G., & oller, .S., 2008, Earthquakes and Subsidence at Kenchreai: sing Recent Earthquakes to Reconsider the Archaeological

and Literary Evidence. In W.R. Caraher, L. ones Hall & R.S. Moore eds. , Archaeology and History in Roman, Medieval and Post-Medieval G reece: Studies on Method and Meaning in Honor of Timothy E. G regory: 53-66. Aldershot. Sciallano, M., 1997, Poissons de l’antiquité. Catalogue d exposition, 1er avril-31 ao t 1997. Mus e d Istres. Scranton, R., Shaw, .W., & Ibrahim, L., 1978, K enchreai Eastern Port of Corinth, Vol. I. Leyde. Vogt, C., 1997, Les c ramiques ommeyyades et abassides d Istabl Antar- ostat: traditions m diterran ennes et influences orientales. In G. D mians d Archimbaud ed. , La céramique médiévale en Méditerranée. Actes du VIe congrè s de l’AIECM2 ( Aix-en-Provence, 13- 18 novembre 195) : 243-260. Aix-en-Provence.

105

LAKE MAREOTIS: RECONSTRUCTING THE PAST

106

M. BERGMANN, ET AL.: SCHEDIA ON THE CANOPIC NILE

Schedia, Alexandria’s Harbour on the Canopic Nile. Interim Report on the German Mission at Kom el Giza/Beheira (2003-2008) Marianne Bergmann, Michael Heinzelmann & Archer Martin Historical Sources and Research History Shortly after the foundation of Alexandria, the new metropolis was connected to the Canopic ile by an artificial, 30 km long channel. At the point where the canal branched off the Canopic Nile a new town, Schedia, was founded (see in general: RE 1921: 401-403; RE 1899: 2030; Fig. 1).1

to display them. At the same time they are an indication of the formation of a prosperous citizenry with an understanding of itself as inhabitants of Schedia. In Roman times Christianity seems to have established itself early under Alexandrian influence. Schedia received its own episcopal seat. In spite of its importance the site as well as the Nile Branch, is still virtually unknown. The river and the canal have now vanished from the landscape. Schedia itself was still a prominent stretch of hills at the beginning of the 20th century with a length of about 2.5 km and a width of 1.5 km. Intensive sebakh extraction and the increasing encroachment on the ancient settlement site by the villages of Kom el Giza, Kom el Hamam and Kom el Nashwa have, however, made the topography so unclear that the place is hardly to be noticed in the landscape. Only in a few places can the flat remains of the koms and single structures be recognized. The greatest height in the territory of Kom el Hamam is occupied by the present cemetery.

In its important river harbour all goods coming from Upper Egypt had to be transferred to smaller vessels. It was also a customs station wherefore a pontoon bridge was installed (σχεδία), which gave its name to the place. Schedia seems to have been a flourishing Greek-dominated polis throughout Ptolemaic and Roman times and was inhabited until the early Byzantine period. Ancient sources suggest the existence of ample harbour and storage installations with the personnel and administrative infrastructure that belonged to them (Strabo, G eogr. 17,1,16; Procopius, Aed. 6,1). The inscriptions found in the area of Schedia provide further evidence of the character of the place (Bernand 1966: 45 ff., 1970a: 329 ff.; Botti 1902). Thus, cults of Zeus Soter and Athena Polias are attested, along with a ship and military station already in the late 4th century BC. About 115 BC the garrison station at Schedia dedicated a Kleopatreion. These cults, as well as early Ptolemaic grave inscriptions with Greek names, appear to indicate a high percentage of Hellenic inhabitants or at least of strongly Hellenized people in the city founded ex novo. There was, however, a Jewish colony as early as the time of Ptolemaios III Euergetes (246-221 BC) with one of the oldest known synagogues attested by an inscription – evidence of the early multicultural character of this harbour city. In the Roman Imperial period Schedia calls itself a polis in inscriptions and apparently has its own magistrate (archon). Various honorary statues and bases for private citizens and emperors found in the area of Kom el Giza suggest that there must have been appropriate public spaces in which

After the first vague attempts to localize Schedia in the 18th century, it was the Napoleonic expedition of 1798-1802 that first concerned itself with Schedia and the Alexandrian canal (Jomard & Jacotin 1818/25; Bernand 1970b: 422). In this connection the place is shown on the first accurate map as one of largest tells in the Nile Delta. A participant in the expedition, ean Michel Le P re, was the first to propose the localization of the beginning of the canal at El-Karyun and the identification of the group of mounds of ashwa and Kom el Giza as Schedia. Subsequently, Ptolemaic-Roman finds begin to be reported in the area later identified as Schedia, mostly chance finds by farmers or people engaged in quarrying sebakh. These objects, which include a series of honorary and funerary inscriptions and a Nilometer, finished in part in private collections but mostly in the museums of Cairo and Alexandria.

1. This project is being carried out by the Institute for Classical Archaeology of the University of Goettingen in cooperation with the Archaeological Prospection Service of the University of Southampton. Further support came from the German Archaeological Institute, the Department of Geodesy at the University of Stuttgart, the Geology Department of Mansoura University and the Centre d’Études Alexandrines. The project has been funded first by the ritz Thyssen oundation and since 2004 by the German Research Foundation. We thank the Supreme Council of Antiquities and its Director General, Dr. Zahi Hawass, for the permission to work at Schedia. For help and support we thank the head of the Antiquities Service for the Department of Beheira, Fawzi el Choulani, and the director of the SCA’s excavations at Schedia, Ahmed Abd el Fattah.

It was only in the 1980s, though, that the first regular archaeological work was done at the site, when the Egyptian Antiquities Service carried out some rescue excavations because of the growing threat to the ancient site through modern construction (Abd el Fattah 1988, 1998, 20002001). Several areas of varying sizes were investigated on the eastern edge of Kom el Giza and at Kom el Hamam, that is in the eastern part of the ancient settlement (see Fig. 4). Numerous building structures of the late Hellenistic, Roman and Byzantine periods were uncovered. A high number of Hellenistic objects were found, among them numerous coins. The excavated areas were restored and protected by the Egyptian Antiquities Service. 107

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig.1: Alexandria and its hinterland in 1925,

Survey of Egypt,s heet 4 ( detail) .

The N ew Evi dence Since 2003 thanks to the Supreme Council of Antiquities (SCA), the Institute of Classical Archaeology of the University of Goettingen has conducted an interdisciplinary project consisting of a topographical survey, large scale geophysical surveying, geological investigations, selective stratigraphic excavations and a comprehensive study of the finds Bergmann & Heinzelmann 2007 .

boundaries can also disclose characteristics of the ancient topography, because some of them seem to document ancient boundaries, like the former edge of a river branch. G eophysical and G eo-archaeological Investigations Three seasons of geophysical surveys have been conducted at the site since 2003 by the University of Southampton, to assess the archaeological remains of the area surrounding the Schedia excavations (Figs. 2 & 3). The surveys were carried out using magnetometry. Furthermore, in 2005 a series of deep drillings and sedimentary analyses was carried out in collaboration with the Geology Department of the University of Mansoura, which will give further evidence of the course of the Canopic Nile and the canal.

Topographical Survey In order to integrate old maps and the results of the excavations, geomagnetic and geological investigations into a unified GIS system, it was necessary to establish a raster of fixed points for the whole site investigated. After the introduction of this geodetic net, relevant points of the newly established fixed point system were measured using a real time GPS to determine the parameters of transformation for the implementation into the Egypt Blue-Belt-system. Subsequently an up-to-date cadastral map was created on the basis of recent Satellite images (Fig. 2). This mapping of the modern surroundings of ancient Schedia was indispensable for several reasons. With regard to the planning of the geomagnetic surveying and the stratigraphic sondages, detailed knowledge of the modern topography especially of the canals, streets and field borders is of great importance. Furthermore, it is possible to draw some conclusions about the ancient landscape from its current shape. This holds true for the relief of the terrain, which reveals areas of settlement as well as the course of the probable harbour basin, but to a certain degree modern field

At present these results seem to suggest that the modern Kanubiye Canal follows the south-western embankment of the ancient Canopic Nile which may have passed between Kom el Giza and Kom el Hamam. In that case, its breadth amounted to 150 m. As there is also clear evidence that there was once a lagoon or a second branch of the river to the north of Kom el Giza, it can be assumed that the settlement of Kom el Giza actually was isolated on a kind of long and narrow island. Maybe it was this topographical situation which gave the stimulus to construct a pontoon bridge here. The investigations have also revealed a possible localisation of the Alexandrian Canal with a strict east-west orientation between Kom el Nashwa and Kom el Sherif and a 108

M. BERGMANN, ET AL.: SCHEDIA ON THE CANOPIC NILE

Fig.2: Schedia.T opographical map based on quickbird satellite image ( 205) ( A) and corings ( H) .G erman Mission at K om el G iza/ Beheira.

indicating excavated areas

Fig.3: Schedia.M ap with reconstruction of the ancient topography.G erman Mission at K om el G iza/ Beheira.

109

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig.4.Sc 20306.G

hedia.P lan of the southeastern settlement ( K om el G iza,K om el Hamam) with excavation areas 19 80- 92and erman Mission at K om el G iza/ Beheira.

possible harbour basin immediately north of it. A second canal could have existed between Kom el Hamam and Kom el Sherif, branching at a right angle from the Canopic Nile and then turning in a westward direction. We thus appear to have a complex system of canals.

the architectural remains documented as a first step. Secondly, some stratigraphic trenches were excavated in order to learn more about the chronology and function of the structures. Furthermore, a series of new stratigraphic sondages have been carried out in previously untouched areas.

As for the extension of the ancient city, it is quite clear that the ancient settlement once occupied the whole area around the possible harbour basin (today’s Nashwa), the whole southern bank of the former Canopic Nile for a length of ca. 2.5 km (today’s Kom el Sherif, Kom el Hamam, el Karyun), as well as parts of the northern bank (today’s Kom el Giza). However it seems that the built up areas did not reach a significant depth but stuck rather close to the edges of the waterways, because of the necessities of a river harbour with its ongoing transhipment of goods. Thus, the picture that emerges is that of a rather amorphous and unhomogeneous settlement, consisting of a series of elongated narrow strips along the waterways, which also separated the different parts of the settlement from each other. Its nucleus has to be assumed in the vicinity of the harbour, from where it probably grew step-by-step without over-all planning along the river embankment. Unfortunately, the former kiman in the area of the harbour have suffered most from sebakh extraction, while we have to assume that the better preserved areas of the former settlement at Kom el Giza and Kom el Hamam were actually on the outskirts.

Summing up the results the following picture appears. In the western part of the older excavations (Area 3), there is a double-tholos-bath with bath-tubs to sit in (Fig. 5). It belongs to a Hellenistic type, which is well attested in the Nile Delta, the Fayum and other parts of Egypt, where it was used as a Hellenistic heritage much longer than anywhere else in the ancient world (see Boussac, et. al. forthcoming) A special feature of the Schedia bath is the visible division between the baths for men and women, otherwise attested in the contemporary papyri. Coin finds and the deep foundations suggest that the bath was built in Hellenistic times, but it is one of the rare examples whose long duration can be shown. Provided with additional rooms with bath tubs for immersion and with new high quality terrazzo floors, it was used at least until around AD 200 (Bergmann & Heinzelmann forthcoming). In its immediate vicinity, still in Area 3, the remains of a huge foundation are visible. Its dimensions seem to indicate that they belonged to a big public building, maybe a temple. Its facade was probably oriented westwards to the Canopic Nile. Unfortunately, however, this part is inaccessible because of modern constructions. Again all findings in this area, including the coins, indicate a (late?) Hellenistic date. Both buildings, the round baths and the possible temple foundation, consist of fired bricks, which is quite unusual for Ptolemaic architecture outside Alexandria.

Excavations at K om el G iza ne important aim of the field archaeological activities was to explore further the parts of the town found during the SCA’s rescue excavations in the 80s and early 90s (Fig. 4). Therefore, these areas were cleaned and all 110

M. BERGMANN, ET AL.: SCHEDIA ON THE CANOPIC NILE

Fig.5:

K om el G iza, Area 3.P lan of Hellenistic baths.G erman Mission at K om el G iza/ Beheira.

Fig.6: K om el G iza, Area 1.P lan of Roman villa,t ombs and later basins. G erman Mission at K om el G iza/ Beheira.

111

LAKE MAREOTIS: RECONSTRUCTING THE PAST Fig.7: K om el Hamam.P lan with excavations areas 203206.G erman Mission at K om el G iza/ Beheira.

Fig.8: K om el Hamam.P lan of Area 6w ith Roman storage building and later phases.G erman Mission at K om el G iza/ Beheira.

112

M. BERGMANN, ET AL.: SCHEDIA ON THE CANOPIC NILE

Fig.9: K om el Hamam.R econstruction of the Roman storage building ( J .Sc humann) . G erman Mission at K om el G iza/ Beheira.

While these two structures of Area 3 seem to be in a certain sense public buildings, Areas 1-2 farther to the east reveal a completely different picture (Fig. 6). In both areas the earliest traces of use were numerous constructions that are possibly to be interpreted as funerary monuments dated on typological grounds to the Roman period. Especially one type of monument is very characteristic: a stepped pyramidal structure of ashlar blocks, probably with a pillar on it, which has close parallels in the necropoli of Alexandria and Marina el-Alamein (Daszewski 1998).

particular interest is a large building measuring 70 x 40 m, which is visible on the surface of Kom el Hamam. Three stratigraphic sondages were carried out down to the foundations of the building (Figs. 8 & 9). They show that it was constructed around AD 100 above earlier mud-brick structures, which were demolished for the purpose. The building was erected in an interesting building technique: the supporting structure consisted of 10 x 8 rows of massive brick pillars with cores of opus caementicium, which bore massive cross vaults of caementicium, collapsed fragments of which were found. It was an enormous fired brick building, particularly by Egyptian standards, that also demonstrates an in-depth technical knowledge of Roman buildings from central Italy of the period. However, two features indicate that the responsible architect was not really familiar with this building technique: contrary to the Roman prototypes, all pillars were constructed without foundations, instead of which huge mudbrick pylons were added already in the original phase of the building on the outside in order to absorb the pressure of the vaults. Furthermore, mud-brick walls running in a north-south direction were constructed between the pillars, which means that the building was separated into a series of nine parallel naves each ca. 5 m wide, 40 m long and 5.5 m high. All these naves opened on the south side through huge doors. The floor consisted of a simple mud-brick pavement. In the middle of the naves long low benches (ca. 0.25 m high, 1.50 m wide) made of mudbricks, seem to have been installed. Unfortunately, no specific finds were found that could give an indication of the original function of the building; obviously it was completely emptied before the changes of the second phase. However, judging by the typology of the building, it most probably formed a large storage facility for some kind of precious goods.

One of the less well preserved pyramidal structures has been investigated. Its core consisted of a cube of fired bricks. Around and underneath these bricks a stratum containing ashes, bones and pottery has been found. All bones belonged to animals of different types, such as at least four jaws of sheep as well as a complete skeleton of a young goat missing the head. They do not show signs of burning. However, a certain portion of the bones concentrated in one area was burnt and was in a very fragmentary state of conservation. They have yet to be analyzed, but it seems likely that they belong to a human cremation, while the unburnt animal bones might be connected to some funeral rites. Because of a Trajanic coin found within the ash layers this structure probably belongs to the 2nd century AD. The area must be regarded as suburban cemetery. At a somewhat later time, an isolated villa was built apparently in the midst of the sparse graves of Area 1 (Fig. 6). This villa of the middle Imperial period, was organised around a central atrium-like courtyard and was furnished with mosaics and opus sectile floors. It was in use for a long time and repeatedly rebuilt and renovated. During the 4th century AD the villa was abandoned, and numerous basins were installed over its ruins and the neighbouring graves, as well as in most of the other areas excavated up to now. Undoubtedly, these were production units that were used intensively and were repaired repeatedly up to the 6th century.

Later, the building underwent several different phases of use and renovation. Only a short time after its erection, additional mud-brick walls were added during the later 2nd and 3rd centuries AD in order to create simple apartments. These habitations show two main phases of use. In one of the sectors several rooms were examined, some of them containing cooking installations. They revealed a huge amount of pottery, mainly amphorae, as well as coarse and

Excavations at K om el Hamam New archeological examinations were carried out at Kom el Hamam and other parts of the ancient city (Fig. 7). Of 113

LAKE MAREOTIS: RECONSTRUCTING THE PAST Fig.10: K om el Hamam, Area 9.R oman enclosure wall.G erman Mission at K om el G iza/ Beheira.

cooking wares, mostly of local production. In the late 4th or early 5th century AD these habitations were abandoned and levelled in order to install a granary. On a level ca. 1 m above the pavements of the prior habitations, a special brick floor was constructed supported by some dozens of small parallel vaulted chambers – a typical Roman system for keeping grain dry. This grain storage facility was in use until the late 5th or 6th centuries AD, when it was destroyed by fire. n its ruins simple dwellings and agrarian installations were installed, which were in use until the Arab conquest.

Once again in the 4th century, a radical change took place. As in the case of the villa suburbana at Kom el Giza, most of the earlier buildings at Kom el Hamam were destroyed or changed their function. Instead of the dwellings or monumental buildings, a whole series of simple working installations occur. Among these are several types of basins, which seem to have served various purposes. Some have the characteristic arrangement for wine production with a higher platform for treading and a lower catchment basin accompanied by further basins for fermentation (Fig. 11) (Rodziewicz 1988). Others are isolated or have furnaces nearby and appear to have served other productive processes that cannot be identified at present. Aside from their function, the basins testify to an intensive agricultural use of the territory in late antiquity. These agricultural installations were in use until the beginning of the 7th century, when Schedia was completely abandoned at the time of the Arab invasion.

Immediately south of this huge building, the remains of another monumental structure was uncovered (Figs. 7 & 10). At a distance of ca. 4.5 m a huge wall runs parallel to the southern front of the above-mentioned storage building. The wall has impressive dimensions: with a thickness of about 2 m its foundations consist of a massive foundation of opus caementicium and three layers of huge ashlar blocks, each with a height of 0.45 m, while the elevation consisted of two more layers of ashlar blocks and a massive mud-brick wall on top. The stratigraphic evidence shows that the wall was constructed shortly before the huge storage building, probably in the first half of the 1st century AD. Its function is not yet clear. The dimensions as well as the massive construction technique indicate that the wall must originally have reached an impressive height. Therefore, it seems probable that the wall actually formed a high enclosure belonging maybe to a temple or some other kind of public building. Further investigations showed that the wall actually belonged to a much bigger structure, consisting of at least two rectangular courts. Strangely enough the northern half of it was abandoned shortly after its construction and substituted by the huge pillar building, while the southern part was continuously in use until late antiquity.

Analysis of the Finds The detailed analysis of our excavated material, especially coins and pottery, will give further information in the future concerning Schedia’s economic development. However, because of the huge amount of finds we are just at the beginning. Coins Around 4,000 coins found at the ancient site of Schedia have been analyzed by Ch. Noeske. Among them are 1,400 stray and stratified finds from the recent excavations. n the whole they cover the period from ca. 320 BC (Ptolemy as a satrap) to the time of Arab conquest (Heraclius, AD 641). Hellenistic coins from the SCA excavations at Kom el Giza are numerous, while in the excavations on the northern side of the Kanubiye Canal (Kom el Hamam, Kom el Sherif) the Roman Imperial and late-antique coins 114

M. BERGMANN, ET AL.: SCHEDIA ON THE CANOPIC NILE prevail. The analysis of these coins has different aims. The dates they offer for the stratigraphic units are important, but they are also analyzed in their own right, as sources for minting policy, monetary circulation and its regions, and for economic change. Comparison to the few other places where coins have been noted, such as Elephantine, Abu Mina and parts of Alexandria, is informative in this respect.

phorae make up approximately 2/3 or 3/4 of the assemblages, which is a characteristic percentage for Mediterranean sites that engage actively in commercial exchange, while in the latter amphorae are attested at about 1/3. Consideration of some of the specific functional groups confirms the difference between the earlier and the later layers. The characteristics of the fine wares and amphorae vary markedly between the two periods. There are so few lamps that it is not possible to speak of trends in them. The utilitarian wares (coarse and cooking wares, which at Schedia appear in the same fabric) are essentially of local production throughout, and their typological development remains to be determined.

Pottery At Schedia the programme of pottery study aims at as complete a documentation as possible. The protocol used is predicated on the idea that all fragments, even typologically unidentifiable body sherds, have some information to give. Therefore, all the sherds per stratigraphic unit are attributed to a functional group and fabric, counted and weighed. For rim sherds the percentage of the diameter preserved is also registered, in order to calculate estimated vessel equivalents. Diagnostic fragments are, of course, classified typologically, and chronologically sensitive items are used for dating purposes. This degree of quantification permits our material to be compared with other assemblages throughout the Roman world, where various standards prevail. It also allows us to make statistical considerations at various levels (Martin 2008; Martin forthcoming).

Among the fine wares in the earlier layers it is noticeable that Schedia did not participate in the fine-ware boom of the Augustan period, when most parts of the Roman world wanted to import or produce red-gloss wares reflecting the repertoire of Italian sigillata, that is to present themselves as Roman at the table (Martin 2008: 268). There are some very occasional imports of Eastern Sigillata A and less frequently yet of Italian sigillata, but for the most part Schedia seems to continue to employ red and black-slipped wares in the Hellenistic tradition. In the late-antique layers a certain quantity of Cypriot and also of African Red-Slip Ware appears, and a larger amount of Egyptian products inspired by them is attested. It is only then that one can say that Schedia participates in the Mediterranean-wide trends in fine wares.

A first level concerns the overall composition of the assemblages by functional groups. There is some evidence that the percentage in particular of transport amphorae with respect to fine table wares, lamps, cooking and coarse wares, reflects the degree of integration of a site into trade networks. Here Schedia shows a decided difference between the late-antique layers up to the abandonment no earlier than the 7th century and the earlier ones. In the former, am-

In the earlier levels Egyptian amphorae greatly outnumber imported ones – the majority were the Amphore Egyptienne 3 type which were first developed at the beginning of the Empire. Other examples included the earlier Amphore

Fig.1 1: K om el Hamam.P lan of Area 8w ith basins for wine-production. G erman Mission at K om el G iza/ Beheira.

115

LAKE MAREOTIS: RECONSTRUCTING THE PAST Egyptienne 2 and the contemporary Amphore Egyptienne 4 types (Empereur 1998: 79). They appear most frequently in the same fabric as the utilitarian vessels, decidedly more rarely in the northwest coastal fabric associated with production around Maryut, and only occasionally in the Nile silt fabric. Imported amphorae are represented by a diffuse scatter of various Aegean and Levantine pieces. In late antiquity, on the other hand, the provenience of the amphorae is much more focused and the percentage of imported vessels much higher. Egyptian vessels appear in the Nile silt fabric associated with Egloff 172 and the Carthage Late Roman Amphora 7, while the imports come almost exclusively from Cilicia with the Carthage Late Roman Amphora 1 and Gaza with the Carthage Late Roman Amphora 4.

During the 2nd and 3rd centuries the city reaches its maximum extension. That the brick-pillar building investigated here was transformed for residential purposes contrary to its original destination, can be considered a further indication of population pressure and of an increasing process of urbanization. Parts of the city reveal a strong Greek tradition (the Hellenistic bath and the stepped bases of grave monuments and Roman influence villa suburbana). Other monuments, like the pillar building with its contemporary use of fired and mud-bricks show an interesting mixture of imported and local building traditions. At the same time, Schedia appears to have been a city with interesting cultural contrasts: while a prosperous upper class lived in houses of clearly Roman influence, a large number of inhabitants lived in simple dwellings of local Egyptian character.

In short, Schedia in the Imperial period presents, paradoxically for the major river harbour of Alexandria of which the written sources inform us, the picture of a site little integrated into the exchange networks and cultural trends of the time, while late-antique Schedia appears to participate more actively in them. How to explain this situation remains to be seen. Could it be that in the earlier period little needed to be offered in exchange for Egyptian products (e.g. grain collected in tax, stone from the Imperial quarries), while late-antique Egypt was in a similar situation to other provinces, with imports and exports. In any case, merit is due to the pottery study that serves to highlight a line of research to be pursued that might not have been highlighted in any other circumstance.

In late antiquity the explored parts of the city underwent a fundamental change. They seem to lose a large part of their urban character with the transformation of the older buildings into productive installations, of rural and other determination. At the same time the new installation of a granary in the pillar building coincides with the undiminished export of grain to Constantinople. How these changes are to be interpreted against the background of new discussions about the flourishing agriculture and renewed trade and imports of late antique Egypt (Banaji 2001; Kingsley & Decker 2001) and against the situation in Alexandria itself, remains to be seen. Although the new transformation of the pillar building into a large granary points to the still uninterrupted importance of the city as a supply base, it changes into a centre for agricultural production that possibly still has regional importance in supplying Alexandria but appears to a large extent to lose its role as an emporium, perhaps because of the overall decrease in trade. This hypothesis, which for now can take only the late Ptolemaic and Roman settlement of Schedia into account, needs to be checked by further investigation and especially to be complemented by a greater consideration of the Ptolemaic phases.

C onclusions The following preliminary hypotheses concerning the development of Schedia seem possible. After the foundation of Alexandria and the construction of the Canopic Canal the early Ptolemaic settlement of Schedia existed probably around an artificial harbour basin at the starting point of the canal. Because of the artificial bridge, the schedia’, and a very dynamic development, the settlement, extended on both sides of the Canopic Nile towards the southeast. Here, on its southern borders, monumental and public buildings (baths and temples) were erected not later than middle and late Ptolemaic times. The settlement, therefore, already had urban character, which is also documented by the inscriptions.

Bibliography Abd el Fattah, A., 2000-2001, Brief report on excavations carried on at Kom el Giza during the season 1989. Annales du Service des Antiquités de l’É gypte 76: 9-12. Abd el Fattah, A., 1998, Quelques nouveaux moules alexandrins à Kom Giza. In A. Abd el-Abd el Fattah & P. Gallo (eds.), Aegyptiaca Alexandrina. Monuments pharaoniques découverts récemment à Alexandrie 1: 65-73. Alexandrina. Abd el Fattah, A., 1988, Recent discoveries in Alexandria and the Chora. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athè nes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole franç aise d’Athè nes, 11- 12 décembre 198: 47-53. Athens, Paris.

In the course of the Late Hellenistic and Early Roman periods a substantial necropolis with monumental tombs developed along the southeastern outskirts of the city, but because of the continuous growth of Schedia, luxurious residences for a prosperous upper class soon penetrated the periphery previously used for burials. At the same time on the southern embankment a monumental building, maybe a temple complex, as well as large storage facilities like the pillar building, arise, apparently as a consequence of massive investments. On one hand, they attest to the economic attractiveness of Schedia in the Imperial period, and, on the other, their distance from the harbour suggests that the more favourable zones were already built up. 116

M. BERGMANN, ET AL.: SCHEDIA ON THE CANOPIC NILE Banaji, J., 2001, Agrarian Change in Late Antiquity. G old, Labour and Aristocratic Dominance. Oxford. Bergmann, B., & Heinzelmann, M., 2007, Schedia, Alexandrias Hafen am kanopischen Nil. Zwischenbericht zu den Arbeiten 2003-2007. Hefte des Archäol ogischen Seminars Bern 20: 65-77. Bergmann, B., & Heinzelmann, M., forthcoming, The tholos-bath at Schedia. In M.F. Boussac, T. Fournet & B. Redon (eds.), Le bain collectif en Egypte. Des balaneia antiques aux hamams contemporains: origine, évolution et actualité des pratiques. Actes du colloque Balnéorient d’Alexandrie 1- 4 Décembre 206 . Alexandria. Bernand, A., 1966, Alexandrie la G rande. Paris. Bernand, A., 1970a, Alexandrian Canal. Le delta égyptien d’aprè s les textes grecs I. Les confins liby ues: 329-380. Cairo. Bernand, A., 1970b, Schedia/Menelais. Le delta égyptien d’aprè s les textes grecs I. Les confins liby ues: 381-442. Cairo. Botti, D.G., 1902, Studio sul III° nomo dell’Egitto inferiore. Bulletin de la Société d’Archéologique d’Alexandrie 4: 41-61. Boussac, M.F., Fournet, T., & Redon, B. (eds.), forthcoming, Le bain collectif en Egypte. Des balaneia antiques aux hamams contemporains: origine, évolution et actualité des pratiques. Actes du colloque Balnéorient d’Alexandrie 1- 4 Décembre 206. Alexandria. Daszewski, A.W., 1998, La nécropole de Marina el-Alamein. In S. Marchgay, M.-Th. Le Dinahet & J.-F. Salles (eds.), Nécropoles et pouvoir: idéologies, pratiques et interprétations. Proceedings of a conference Lyon 195 : 229-241. Lyon. Empereur, J.-Y., 1998, Les ateliers d’amphores du lac Mariout. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athè nes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole

franç aise d’Athè nes, 11- 12 décembre 19 :8 75-91. Athens, Paris. Jomard, E.F., & Jacotin, P., 1818/25, Carte Topographique De L’É gypte Et De Plusieurs Parties Des Pays Limitrophes (1818/25), M 1: 100 000, Blatt Nr. 37. Kingsley, S., & Decker, M., 2001, New Rome, new theories on inter-regional exchange. In S. Kingsley & M. Decker (eds.), Economy and Exchange in the East Mediterranean during Late Antiquity. Proceedings of a conference at Somerville : 1-27. Oxford. College.O xford 29 th May 19 Martin, A., 2008, Pottery from Schedia near Alexandria (Egypt). Rei Cretariae Romanae Fautorum Acta 40: 263-269. Martin, A., forthcoming, The pottery from a late-antique settlement at Schedia (Western Delta, Egypt). In S. Menchelli, M. Pasquinucci & S. Santoro (eds.), The 3rd International Conference on Late Roman Coarse Ware, Cooking Ware and Amphorae in the Mediterranean: Archaeology and Archaeometry ( LRCW3) ( Parma – P isa,25- 30M arch 208) . Oxford. Procopius, e edificiis, Vol. VII, H.B. Dewing (transl. & ed.), 1961. Cambridge, MA. RE 1899, Wissowa, G. (ed.), Paulys Real-Encyclopädi e der classischen Altertumswissenschaften, Vol. I 1, 2030 s.v. αιρ ου (Sethe). Stuttgart. RE 1921, Wissowa, G., Kroll, W., &Witte, K. (eds.), Paulys Real-Encyclopädi e der Classischen Altertumswissen-schaft, Vol. II A 2, 401-403 s.v. Schedia (Kees). Stuttgart. Rodziewicz, M.D., 1988, Classification of wineries from Mareotis. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athè nes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole franç aise d’Athè nes, 11- 12 décembre 19 8: 27-36. Athens, Paris. Strabo, G eography, Vol. VIII, H.L. Jones (transl. & ed.), 1959. Cambridge, MA.

117

LAKE MAREOTIS: RECONSTRUCTING THE PAST

118

P. WILSON: SURVEY WORK IN MAREOTIS

Recent Survey Work in the Southern Mareotis Area Penelope Wilson

Introduction The ‘Sais and its Hinterland Project’ based at Sa el-Hagar, Sais in Gharbiyah Province, was designed partly to determine the impact of river distributary systems on settlement in the north and north-west of Egypt. In order to compare the Canopic with the Rosetta riverine systems, 70 sites were surveyed and visited in Beheira and Kafr esh-Sheikh Provinces. This paper describes and discusses sites in the southern Mareotis area, which formed an interesting group in themselves. They represent places influenced by another type of water-system, an inland lake located at sea-level, that is fed by a series of distributaries from the Canopic system. The location and extent of Lake Mareotis fluctuated over time and this may have been an influencing factor in the life-span recorded at some of the sites in the southern Mareotis area.

sides. The lower parts of its sides have been left largely untouched, but are used for animal husbandry. The village of Kom Ishu was formerly on the western, southern and eastern sides of the hill but has developed in linear streets radiating to the east. It is possible that the original heart of the village was sheltered from the north wind by the hill. The field boundaries beyond the village suggest that the kom may have covered a larger area at one time. The whole of the upper surface of the hill is now covered by a modern cemetery. Pottery gathered at the site is mostly of Late Roman date.

K om Ishu and K om el-Hagg The two sites share a similar geological nature as they both consist of rocky limestone outcrops at the edge of a limestone ridge formation and seem to have been the focus of human activity in Lake Mareotis (Embabi 2004: 260-2).

Kom el-Hagg (SCA 100155) is surrounded by a small ezbet, called Ezbet el-Bank. The mound has a sloping, rounded edge to the east and the other sides are slightly steeper at the edges. The area is used for stabling animals and as a manure store. The southern edge of the kom has been cut away, leaving a metre-high section above field level containing pottery fragments. The fields to the south of the main mound are covered in sherds and may have been part of the ancient site, perhaps representing a settlement against the rocky outcrop. It is also possible that the sherds to the south were washed or swept down off the rocky hill. It would be interesting, therefore, to define the exact nature of the site at this point because of its location and the apparent use of the local geological features as at Kom Ishu and also at Kom Lemsan, situated to the north.

Kom Ishu is approximately 12 km west-south-west of Kafr ad Dawar. The rocky outcrop has steep, almost sheer,

The pottery sample from the site gave a broadly consistent date range from the Ptolemaic through to the Early Roman

The Sites (Figs. 1-2) The sites are discussed in an anti-clockwise direction around Lake Mareotis. Although the pottery collection and survey are of a preliminary nature, some interesting observations can be made arising from a study of the results.

Fig.1: Ancient sites in the Delta,hi ghlighting the Mareotis region ( drawn by P. Wilson) .

119

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig.2: Map of the sites surveyed,w ith 0m etre sea-level contour area highlighted in grey.B y R.D ickinson,af ter 1: 50, Survey of Egypt maps ( 197) .

120

P. WILSON: SURVEY WORK IN MAREOTIS periods. The different types of vessel from the site included both amphorae and domestic vessels and so did not indicate a particular function for the site, but there certainly seems to have been some kind of settlement here. Most of the pottery was Egyptian in origin, and presumably was produced locally, with only one of the amphorae of Early Ptolemaic date perhaps from Rhodes, possibly a reused wine or water container. Both Kom el-Hagg and Kom Ishu are situated on the western edge of the main body of Lake Mareotis and could have had strategic value in guarding the lake shores or in watching over the desert areas to the west, the river to the south and marshes or fields on all sides Bernand 1970: 866). Although the two places are within signalling distance, the pottery represents two distinct phases, with that from Kom el-Hagg covering the Ptolemaic and Early Roman periods, while that at Kom Ishu is of the Later Roman period. There may be a number of reasons for this, mostly due to the nature of the collection of the pottery sample. But it could be that Kom Ishu replaced Kom el-Hagg as a watching place or simply that it continued for a longer period of time, perhaps because it was situated upon a more long-lived canal which later became the Tirat Abis. In such a case, Kom Ishu had a better strategic value in the later period because of its water channel communication.

Fig.3: The main mound at K om el-Mahar ( photograph by P.W ilson) . 2nd century. Most of the pottery sample consists of tablewares, with some imported finewares and glass, and may have come from domestic contexts. Kom el-Mahar could have been a small town or village focussed on a series of elite houses and villas on the southern shore of the lake, taking advantage of the elevated land (Botti 1902: 55).

K om el-Mahar and Sidi G hazi Kom el-Mahar (SCA 100167) lies isolated in farmland to the east of the Nubariyah Canal and south-west of Sidi Ghazi village. The main mound is covered in pottery, including finewares ig. 3 . The name of the site in Arabic means ‘shell’ and shelly deposits are apparent all over the mound and in the sections. The shell derives from the fact that the mud-brick used to build the houses of the town was made from silt mixed with crushed shell from degraded, local oolitic limestone. When the mud-brick disintegrated, a large amount of shell was left lying upon the surface.

The modern village of Sidi Ghazi (SCA 100142) lies approximately 10 km south of Kafr ad Dawar and to the north of the Masraf al-Umum Canal which runs to the west of the Baslaqun-Luqin group of tells. There is now no tell at the location of Sidi Ghazi, but modern maps indicate that the village lies upon a small area of raised ground. It is likely that the original tell has been subsumed under the village. A site here may have formed one of the islands inside Lake Mareotis. South-western Sites Kom el-Gel (SCA 100217) is the site at Ezbet Abd el-Qadir at-Tawil and lies approximately 17 km north-west of Abu Matamir. The site consists of a mound with a flat top and gently sloping northern side and a secondary mound to the east. The top of the mound is covered with a modern cemetery, while the sides are used for animal husbandry. The houses on the western side of the tell are built on flat land at the base of the hill. The fields to the south of the mound contain noticeable quantities of pottery and some limestone fragments, suggesting that this was once a more sizeable mound. Satellite imagery suggests that the area to the south may have been part of the mound because of the shape of the fields.

The lower slopes of the eastern and south-eastern side of the mound are covered by a modern cemetery. The northern and north-western side of the mound has sheer sides, where they have been dug away and it is likely that the original site extended in all directions away from the mound. It is now surrounded by deep irrigation ditches and fields, which are encroaching on the outer parts of the site. One of the northern sections contains a substantial, stratified sequence, including ashlar limestone blocks, a wall made of sandy mud-bricks and shell-filled, mudbricks. There are also fragments of red granite on the mound and glass fragments lying upon the surface of the site. The southern area of the site has been flattened and there are building outlines visible on the surface. There are some lower, smaller mounds covered in scrub to the northwest and south-east sides of the site.

The pottery sherds from the site comprised fine wares, domestic material and amphorae. The pottery dated mainly to the Late Ptolemaic period, with some Early and Late Roman sherds included in the collection.

The main pottery component from the site ranges from Late Ptolemaic to Early Roman in date (Fig. 4), suggesting that it may be a Ptolemaic foundation continuing as a town into the Roman period, before being abandoned around the

The original site at Kom el-Farag lies under a modern town, approximately 13 km north-west of Abu Matamir. It may be the location of Kom el-Hanache (Toussoun 1922: 121

LAKE MAREOTIS: RECONSTRUCTING THE PAST Fig.4: Late Ptolemaic to Early Roman pottery sample from K om el-Mahar ( drawings and analysis by D.G rigoropoulos) .

lake next to the village which may be a relic of an ancient river distributary or ox-bow lake, extending for about 500 m in a north-west to south-east orientation. The lake may link with some ancient depressions in the landscape to the west or have formed one of the distributary channels from the Canopic Branch. If there was once an ancient settlement here, it may have been located upon the river, lying along a raised levee as with other tells in this part of Beheira.

Pl.II), but as the Ezbet Farag grew to the south, the tell seems to have been subsumed underneath the houses and removed by the Delta Light Railway. The cemetery part of Kom el-Farag lies to the north-west and may be a remnant of the older mound, being a maximum of 1.5 m above ground level. The old part of the village upon the mound has a less organised street plan than the more recent areas of the town with their orthogonal layout. The interest of this location is that it lies on the Tirat Abdel al-Fatah and Tirat al-Hajar drainage channels and, perhaps, on the channel of one of the earlier Canopic distributaries on the western side of Beheira. Kom Farag is also at the centre of a cluster of small villages and ezbets with the word kom in their name, such as Kom es-Siwayd to the south, Kom Kifri, as Sa’ayidah and Jaradah to the south-west and to the north-west Ezbet Abd el-Qadir atTawil (Kom Makhboura), which has a cemetery upon a kom. They may well be part of a town cluster focussed on an irrigation basin network dating at least to the medieval period.

Abu Guduur (SCA 100212) is 2.5 km north-west of Abu Matamir and lies to the south of the main road, upon a low ridge of silt along an ancient distributary course, at the edge of the Mareotis basin. It once covered an extensive area, but now consists of some low, sandy mounds covered in scrub and a larger flat area with scrub and bushes. Most of the site is now used as a municipal rubbish dump, having once also been a gravel extraction and processing area. The highest mound is only about 2 m above the level of the surrounding land and the whole area is very sandy. There was pottery lying on the tell as well as some bronze objects, including a stirrup ring and a three-flanged arrowhead. Both could date to the Late Dynastic period. A small diagnostic set of pottery fragments were collected from the site. As far as can be judged, they mostly date to the Late Ptolemaic period, but there were some Late Roman sherds and some Late Dynastic pottery as well.

South of the Lake Kom el-Adda SCA 100230 or Kom Abu El-Eida lies 8 km north-west of Abu Matamir. Although there is a cemetery mound to the south-east of the main village, the main interest of the area is the curious crescent-shaped, enclosed 122

P. WILSON: SURVEY WORK IN MAREOTIS The place may have been a strategic Dynastic location for a small military establishment related to the other western delta fortified sites and to other mounds running from Abu Matamir to the south east. Southern Port on the Naukratis Canal Kom Trugi or Turuga (SCA 100251) lies 6 km due north of Abu Matamir, close to several other places with kom in their name and on the former edge of Lake Mareotis. Turuga is known to have been an important inland port on the lake for people going from Alexandria to the Wadi elNatrun and for goods and supplies coming out of the Western Desert and delta towards Alexandria for trade (Timm 1984-1992: 2545-6 De Cosson 1935: 79, 151 . It is possible that the site had two halves on either side of the waterway into the lake, one known as Psenemphaia, according to a stela from the site Bernand 1970: 899-913 and the other known as Theroge or Therange, which lends its name to the whole site after the Ptolemaic period. Roman period material was noted at the site in 1895 by D.G. Hogarth Spencer 1999: 305, 308-11 and Botti noted that marble column capitals could been seen there and that a headless white marble statue was found at Kom Trugi (Botti 1902: 58). Kom Trugi suffered greatly from sebakhin digging (De Cosson 1936 , with the authorities only intervening at the discovery of important objects or remains such as a Ptolemaic-Roman bath-house El-Khashab 1957: 127-139 .

Fig.5: Red brick structures at K om Trugi,e xcavated below foundation level.T he scale is 2m etres high ( photograph by P.W ilson) . Fig.6: Section through the cemetery mound at K om elQ adi ( photograph by P.W ilson) .

It is clear that this was once an extensive site and perhaps a high mound. It now consists of flat, sandy areas interspersed with several excavated red brick structures and limestone upon mounds. To the north of the area, there is a modern village with an adjoining cemetery and both may have been built upon part of the original site. Amongst the extant red brick structures there is part of a pavement left standing upon its sand-brick foundation and a pillar of earth, as well as red brick and plastered tanks, perhaps from a winery. There is also a fragment of a limestone wall consisting of five courses of ashlar blocks for a length of 4.75 m, with some plaster still adhering to one side (Fig. 5). All of this area seems to have been dug out to below the floor level of these buildings and the site seems very denuded. On one of the mounds to the south there are the foundation walls of a limestone building. The flat, sandy area between the mounds shows building plans in wet weather, according to the local people, but is now mostly used as a football pitch. There are also a number of red granite grindstones at the site, lying upon the ground. Although it is difficult to make sense of the standing remains, they do suggest that an impressive town with large red brick and limestone buildings once stood here.

the medieval period. The visible levels date perhaps to the Early and Middle Roman periods and information about the later levels may be lost. The height of the mound suggests that there is still buried material at the site, but that it is disappearing fast. North-east of the Lake Kom el-Qadi (SCA 100165), south of the main Cairo to Alexandria road, and Kafr ad Dawar may be part of a cluster of sites around Al Baslaqun, forming a township or irrigation basin group of sites. The sprawling area at Kom el-Qadi has several distinct zones and a main archaeological area to the south of Kom el-Qadi village. It consists of a high mound, around 8 m high, the southern and northern faces having been cut away to form sheer sections. The southern sections show significant stratified sequences, including sand-brick walls, pottery deposits and human burials in both pottery coffins and brick vaults. Skulls of adults and children can be seen in the sections and on top of the mound (Fig. 6). The remains may be interpreted as

The pottery sample collected from the site dates mainly from the Early Roman period, but there are also some Ptolemaic, Middle-Late Roman and Medieval Arab period sherds in the collected sample. Because of the disturbed nature of the site, however, it is difficult to associate particular pottery with specific areas of the site. The range of the pottery agrees well with the historical material, suggesting that the site was used from the Ptolemaic through to 123

LAKE MAREOTIS: RECONSTRUCTING THE PAST Summary (Table 1) Lake Mareotis created its own special topography and landscape, with the shallow flooded area bordered by reeds and rushes forcing people onto the high ground around the lake edge, islands in the lake and between the distributaries to the south and the Canopic and Alexandria Canal to the north. The strange, thin tongues of land, with their dykes and paths, must have made the desert routes seem attractive to travellers and so created opportunities on the Western Desert edges for monastic hostels and stopping places. The topography of the area shows that the presence of the lake had a profound impact upon social organisation and economic management in the North-West Delta. Pharaonic period attempts to control the Western Desert edges may always have led to limited settlement in the north-west area, as suggested by Abu Guduur and the fortress line in the Ramesside period (Habachi 1980). The foundation of Alexandria seems to have acted like a magnet, attracting service settlements and industries upon a huge scale, as well as enabling the development of agricultural lands around the city and bringing settlement in favourable locations. The preliminary dating evidence from the sites in Beheira shows that the earliest material at the surveyed sites in the majority of cases comes from the Ptolemaic period, contemporary with the development of Alexandria Wilson & Grigoropoulos 2009: 42-43 . The sites at Kom el-Mahar, Kom el-Gel and Kom Trugi may all be part of this wave of development, with the creation of farming towns, a process which continued into the Early Roman period. This seems to have been a busy time for the lake, with many settlements established either as internal trans-shipping ports for the movement of goods and people

a reuse of the abandoned town mound as a cemetery. This may be the Roman cemetery from which three inscriptions were published (Botti 1902: 57). On the eastern side of this mound a substantial mud-brick wall is visible in the section. In the village to the west, there is a Roman bath-house, which has been restored recently and includes part of the concrete floor of a tholos, some possible underfloor furnaces and rectangular tanks. More uncovered red brick chambers as well as surface building plans, can be seen on a sandy area to the east of the village and main mound. A further sandy area lying to the east of this part of the village and a patch of extensive low scrub to the north-east of the village may have been part of the original site. There were some larger fragments of limestone lying in the mounds and the village. The pottery collected was mostly from the mound. It consisted of Early and Middle Roman material, including some imported fine wares, many cooking pots which showed evidence of burning on the outside and some Egyptian amphorae. The material seems to indicate the domestic nature of the site in the Roman period, but the relationship of the cemetery remains to the settlements was not clear from a preliminary examination of the mound sections. The site seems to be a small town and its associated cemetery, perhaps dating from the height of Roman management of this area. Table 1: Summary table of sites surveyed by the ‘ Sais and its Hinterland Proj ect’ in the region of southern Lake Mareotis. Site & current dimensions

Early Roman Ptolemaic Early Late 1st c. BC rd st 3 -late 1 c. BC to 2nd c. AD

Middle Roman Late Roman 3rd-4th c. AD 5th-end 7th c. AD

Kom Ishu 150 x 200 x 10 m

Early medieval Arab 8th-16th c. AD

X

FUNCTION Watch post

Kom el-Hagg 120 x 200 x 4 m

X

X

Watch post

Kom el-Mahar 440 x 240 x 12 m

X

X

Main Ptol-Roman town and villa site

Sidi Ghazi Under town Kom el-Gel 270 x 220 x 4 m

Island village X late

X

Town/villa

Kom el-Farag Under town

Farming village

Kom el-Adda 220 x 300 x 3 m

Riverside town

Abu Guduur 300 x 250 x 2 m

X late

Kom Trugi 350 x 300 x 3 m

X

X late X

X

Kom Hassan nder fields Kom el-Qadi 370 x 250 x 8 m

Fort?, riverside town X

Inland harbour town Unclear

X

X

Main Roman town

124

P. WILSON: SURVEY WORK IN MAREOTIS Bibliography SCA numbers refer to site registration numbers of the Supreme Council of Antiquities,E gypt.

to and from Alexandria, or as points of control around the lake, such as at Kom Ishu, Kom el-Hagg and perhaps Sidi Ghazi. The apparent abandonment of places, such as Kom el-Mahar and Kom el-Gel, may be part of the ‘Antonine Plague’ effect with depopulation of some areas (Bagnall & rier 1994: 173-8 . In the later periods, the lack of Late Roman and Early Arab continuity at the sites suggests that, in common with other parts of Beheira province, once the Canopic Branch of the Nile and its distributaries had silted up, the sustainability of some of the settlements was not possible and that people moved away from earlier settled land. They may have gone to a few major centres such as Abu Matamir or Dilingat or Damanhur or Alexandria itself, rather than carrying on trying to eke out an existence on the shores of the Mareotis Lake.

Bagnall, R.S., & Frier, B.W., 1994, The Demography of Roman Egypt. Cambridge. Bernand, A., 1970, Le Delta égyptien d’aprè s les textes grecques, I-III. Cairo. Botti, G., 1902, Studio sul IIIo nomo dell’egitto inferiore: e più specialmente sulla Regione Mareotica. Bulletin de la societé royale d’archéologie - Alexandrie 4: 4184. De Cosson, A., 1935, Mareotis: Being an Account of the History and Ancient Monuments of the North-West Desert of Egypt and Lake Mareotis. London. De Cosson, A., 1936, El Barnugi, Bulletin de la societé royale d’archaeologie - Alexandrie 31: 113-116. Daressy, G., 1929, Ménélais et l’embouchure de la branche Canopique. Revue de l’É gypte ancienne 2: 20-51, Pls. I-II. Embabi, N.S., 2004, The G eomorphology of Egypt. Landforms and Evolution. Vol. I. The Nile Valley and the Western Desert. Cairo. El-Khachab, A.M., 1957, Les Hammams du Kom Trougah. Annale du Service des Antiquités de l’ Egypte 54: 117-139. Habachi, L., 1980, The Military Posts of Ramesses II on the Coastal Road and the Western Part of the Delta. Bulletin de l’Institut franç ais d’archéologie orientale 80: 13-30. Spencer, P., 1999, Hogarth’s 1895 Report to the Committee of the Egypt Exploration Fund. In W.J. Tait & A. Leahy (eds.), Studies on Ancient Egypt in Honour of H.S. Smith: 302-311. London. Survey of Egypt 1997. 1:50,000 Maps; “Abu al-Matamir”, “Kafr ad-Dawwar”, “Ikinij Maryut” & “al-Iskan dariyyah”. Survey of Egypt/Ministry of Finances, Cairo. Timm, S., 1984-1992, Das christlich-koptische Ägypten in arabischer Z eit: Eine Sammlung christlicher Syät ten in Ägypten in arabischer Z eit, 6 Vols. Wiesbaden. Toussoun, O., 1922, Mémoire sur les anciennes branches du Nil. Mémoires de l’Institut d’É gypte 4, Premier fascicle, Cairo Wilson, P., & Grigoropolous, D., 2009, The West Nile Delta Regional Survey, Beheira and K afr esh-Sheikh Provinces. London.

Of the sites surveyed here, Kom Trugi was the largest and most important, perhaps straddling the distributary of the Canopic Branch from Naukratis and providing a secondary direct link from the south to Alexandria and from the Western Delta area to Alexandria. Kom el-Qadi seems to have been a typical Early Roman town with a bath-house, similar to other new towns in the Roman delta, and a substantial mound used as a cemetery. It may have been founded in the Early Roman period at around the same time as Schedia and part of the cluster around BaslaqunLuqin, once suggested to be the location of Menelais Daressy 1929). Further detailed work in any or all of the Mareotis catchment area would be highly desirable to explore riverine, lacustrine and human interaction during periods of profound cultural change. Acknowledgements I thank Dr Dimitrios Grigoropoulos of the Deutsches Archä ologisches Institut – Athens who studied the pottery from the sites and Dr Roger Dickinson for the mapping. I am also grateful to the Supreme Council of Antiquities in Damanhur under Fawzy Fawzy el-Kholi and Naama Suleiman, as well as Zahi Hawass and Magdy Ghandour in Cairo. or the fieldwork, I thank Dr oanne Rowland, Mohamed Osman, Ashraf abd el Rahman and Ahmed Bilal. The Department of Archaeology, Durham University and Egypt Exploration Society also supported the work. Funding from the Arts and Humanities Research Council, UK is acknowledged. The survey is now published by the Egypt Exploration Society’s Delta Survey Project under the title The West Nile Delta Regional Survey, Beheira and K afr esh Sheikh Provinces.

125

LAKE MAREOTIS: RECONSTRUCTING THE PAST

126

D. DZIERZBICKA: WINERIES OF THE MAREOTIC REGION

Wineries of the Mareotic Region Dorota Dzierzbicka

Sea, it was by far the best in the Nile Valley. Sources from the Pharaonic and Persian periods testify to wine production in this area long before the coming of the Greeks (cf. Forbes 1955: 70-124). The wineries date from the Roman and Byzantine periods. Some are of considerable size and testify to extensive cultivation of grapes and large scale wine production in this area, especially in correlation with the local pottery kilns which indicate a mass production of amphorae (Empereur 1993: 39-47).

Wineries or remains of wine making installations have been found in various parts of the Nile Valley: the coastal areas and the Delta, the Fayum and Bahariya Oasis, as well as Upper Egypt (Dzierzbicka 2005: 11-24). However, their concentration in the Mareotic region, especially along the southern shore of the lake and around Abu Mina, is unmatched in other parts of the country (Fig. 1). In addition to the identified wineries, there are also elements of wine-making installations that are no longer in situ – primarily lion headed spouts through which freshly pressed or trodden must flowed into the vat. umerous examples, including those found around Lake Maryut, can be seen in Alexandria in the Greco-Roman Museum or at the Kom el-Shukaqa necropolis (Rodziewicz 1998: 28, n. 6).

Literary Sources In literary sources wine produced in the Mareotic region during the Greco-Roman Period is well attested. It was known to Horace (O des I 37.14), who wrote that the mind of Cleopatra was made mad by Mareotic wine (mens lymphata Mareotico). According to Strabo (XVII 1,14) the Mareotic region produced good wine and in such quantity that it was racked to be matured. The white Mareotic grape variety (Mareotides albæ ) is mentioned by Virgil in G eorgica (G eor. ii. 91). Athenaeus (Deip. I 33 d-e), in turn

Many factors contributed to the large concentration of wineries in the vicinity of Lake Maryut. Although in antiquity the climate in the Delta may have not been as beneficial for vineyards as on the northern shores of the Mediterranean

Fig.1: A map of the Mareotic region with marked locations of wineries.P lotted on the map are excavated and published structures, as well as unexcavated ones mentioned by scholars in other works or featured on SCA maps as protected archaeological sites.T he wineries designated as NN were plotted by Rodziewicz ( 198: 28,F ig.1) ,but the name of their location was not provided ( D.D zierzbicka) .

127

LAKE MAREOTIS: RECONSTRUCTING THE PAST claims that Mareotic wine was excellent it was white and pleasant, fragrant, easily assimilated, thin, did not go to the head and was diuretic. According to this author better yet was the wine from Taenia, from the same region. This was pale yellow and had an oily quality, which disappeared after it had been diluted with water. Besides being sweet, this wine was rather aromatic and mildly astringent.

There are three other wine making installations in the vicinity of the modern village of Huwariya. One of them is a simple Early Roman Type I unit built of dressed stone blocks bound with mortar unexplored, mentioned by Rodziewicz 1998: 31). A winery in southern Huwariya (reported by Rodziewicz 1998: 34), built of irregular stone blocks, clay and waterproof plaster, consisted of one press unit a treading platform and vat connected by two flights of steps. The round base of a mechanical press was set in a niche in the wall of the treading platform. A channel in the floor of the platform led from the press directly to the vat. A third installation, located 3 km southeast of Huwariya (Rodziewicz 1998: 35), consisted of a treading platform, a vat and a mechanical press separated from the treading floor by a low, thin wall. A channel under the floor led from the mechanical press to the vat. The walls of the structure were built of irregular stone blocks bonded with mortar and covered with plaster reinforced with potsherds.

Wineries1 The South Coast of Lake Maryut The wineries in the immediate vicinity of Lake Mareotis are located on the strip of land between the lake and the Mediterranean Sea, and along the southern shore of the lake. The majority are found on the southern shore of the lake. A large winery near Burg el-Arab (30o55’37”N 29o 31’48.50”E),2 is located in the northeast corner of a large building (El-Ashmawi 1998: 55-73). A mechanical press was set up on the treading platform and closed off from the surrounding area by a thin wall. The must from the mechanical press enclosure and from the treading floor flowed to the adjacent collection basin through two separate channels.

A small, presumably Roman winery of George Nesim (30° 57’18.30”N 29° 36’15.20”E) was found by the Lake Mareotis Research Project (Blue & Khalil forthcoming). It appears to have consisted of a treading platform and vat Type I Rodziewicz 1998: 31 , but the presence of a mechanical press unit with a separate, small collection basin within the same complex cannot be excluded, although it requires further investigation.

The winery located at Taher el-Masry (30o56’59.40”N 29o34’58.20”E)3 was built of stone blocks and stone rubble embedded in mortar. The uncovered press unit consisted of a square treading platform 3.75 m x 3.75 m and a vat 2.37 m x 2.00 m , 1.40 m deep. utlines of structures surrounding the unit suggest that the winery may have been part of some kind of a larger complex.

The North Coast of Lake Maryut To the north of the lake, at Abu Talaat (unpublished, mentioned by Rodziewicz 1998: 29), an Egyptian archaeological mission uncovered a winery with two connected treading platforms and one collecting vat. The wine-making unit was entirely hewn out of the bedrock except for one wall, which was built of stone blocks.

A restored wine-making unit near Marea (30o58’49.10”N 29o40’12.90”E) consisted of a large treading platform, a smaller room with a base for a mechanical press, and a collection vat (El-Fakharani 1983: 182-184). Two separate channels led from the two rooms to the vat. The rooms where the pressing took place were raised higher than the room of the vat and two flights of steps connected the two levels. Nearby, on the same site another unpublished winemaking unit was found.4

The winery at Site 215 recorded by the Lake Mareotis Research Project (30° 53’30”N 29° 25’22”E) has two parallel treading platforms visible on the surface. The channels in their western sides indicate the presence of two collection basins, which have not yet been uncovered. Thus, this winery represents a compound of two Type I wineries according to the typology established by Rodziewicz (1998: 31-35). Further investigation is needed to identify its other features.

1. This part of the paper includes elements previously published in Dzierzbicka 2005. 2. GPS coordinates of the wineries as well as some additional information –di mensions, architectural details –w ere collected on my study tour of the Mareotic wineries in November 2005 during my scholarship at the Polish Centre of Mediterranean Archaeology in Cairo, Egypt. The study tour included the wineries near Burg el-Arab (also called Abu Sir), at Taher el-Masry and Marea, two of the installations at Abu Mina, and the wineries at Karm el-Shewelhy and Karm el-Baraasi. I am thankful to Mr. Adli Rushdy from the West Delta Inspectorate and to Mr. Saber Selim from the Islamic Inspectorate of the Supreme Council of Antiquities for allowing me to see the sites. For GPS locations and photographs of the wineries at George Nesim, Kom Trouga and Site 215 I am greatly indebted to Dr. Lucy Blue and Dr. Emad Khalil of the Lake Mareotis Research Project. 3. Explored by Mr. Mustafa Rushdy of the Egyptian Antiquities rganisation in 1998/99 (as it was communicated to me by Mr. Maged Ahmed, Inspector of the SCA), unpublished. The survey I conducted in 2005 yielded the GPS coordinates of the site, as well as some preliminary observations based on the remains visible on the surface. 4. I am thankful to Prof. Mona Haggag for bringing it to my attention.

Area to the East of Lake Maryut Much fewer wineries have been located in the eastern part of the Mareotis region. A winery was uncovered in 1917 by Daninos Pasha near Abu ir Bay east of Alexandria (published as a private bath in Breccia 1926: 47-49). The Ptolemaic dating established by Breccia on the basis of the quality of plaster used in the structure seems rather uncertain. A large treading platform 3.25 m x 5.80 m covered with waterproof plaster was raised 0.30 m above the floor level of the building. Two channels led through a wall separating the platform from a collection basin 1.60 m deep, 4.20 m long and 2.20 m wide. Twin flights of steps connected the two parts of the winery. 128

D. DZIERZBICKA: WINERIES OF THE MAREOTIC REGION There is also a Roman Period winery at Kom Trouga (30° 57’46.90”N, 30° 10’27.10”E). This unpublished winery mentioned by El-Fakharani (1983: 184) and Rodziewicz 1998: 31 was a complex that consisted of several scattered wine making units, today badly damaged due to quarrying activity in the area. Remains of basins and adjacent raised treading platforms are visible. All of the units appear to have consisted of a single treading platform and vat, covered with waterproof plaster. The best preserved one was built of stone blocks, while others were brick structures. Possible wine-making installations have also been found at Schedia.5

two treading platforms, two vats, one mechanical press, and storage rooms. The structure was built of mud-brick and re-used stone blocks. The winery located at Karm el-Shewelhy (30o49’37.70”N, 29o38’65”E) consisted of a treading platform and vat, two storage rooms, and a courtyard. The wine-making unit was hewn in rock except for the eastern wall, which was built of limestone blocks. The floor of the treading platform and the walls of the vat were covered with waterproof plaster (Abd el-Aziz Negm 1998: 73). A sizeable winery at Karm el-Baraasi (30o50’41.40”N, 29o40’49.50”E) consisted of several large treading platforms with vats and mechanical presses. The excavations published by Abd el-Aziz Negm (1998: 63-70) brought to light a large treading platform and vat. In the same room as the vat there were two small enclosures (4 m2 each) for mechanical presses, each equipped with a small vat of its own. In November 2005 another treading platform was visible semicircular in shape, with an adjacent vat. This unit was located further to the west and separated from the first platform by a room. The floor level of this room was lower than the two treading platforms and connected to them by two short flights of steps. urther to the west traces of three more treading platforms were visible not yet excavated . The complex was built of mudbrick with corners reinforced with limestone blocks. The entrance to the unit uncovered by the excavators led from a courtyard. The winery was part of a larger complex across the courtyard from the winery there was a house, and the courtyard itself was surrounded by rooms. Around 300 m to the north-west of this complex 30o50’50.70”N, 29o40’40.20”E) there are probably more installations yet to be uncovered, as one can tell from the outlines of basins covered with waterproof plaster that are discernible on the surface.

The Region of Abu Mina A number of wineries were found in the vicinity of Abu Mina, not far from Lake Mareotis. The region was a thriving grape-growing and wine-making centre. There are three wineries within the town area. A large winery uncovered in 1907 by Kauffmann to the east of the double bath (Grossmann, et al. 1984: 148 Grossmann, et al. 1997: 87-90 behind the portico of the colonnaded street (30o50’52.90”N, 29o39 76.10 E was expanded in five phases until it became a complex of five treading floors with four mechanical presses. It was built of mud-brick and stone blocks covered with waterproof plaster. Another winery was located behind the apse of the basilica (30o50’45.80”N 29o39 82.30 E . It consisted of two rooms, the first housing one large treading platform flanked by two mechanical presses, a collection vat, and one smaller raised platform. In the second room there was another treading floor and vat Grossmann 1999: 82-83, fig. 9 . The third winery (Mül ler-Wiener & Grossmann 1967: 468-473) within the area of Abu Mina was uncovered in the so-called K umring A, in three rooms of Building 17, entered from a courtyard. Two of these rooms each contained a set of one large and one small treading platform and vat. The third room on the opposite side of the courtyard was used for storage.

Elements of Mareotic Wineries The Mareotic evidence provides an abundant and diverse range of evidence for wineries, a brief overview of the basic elements of a winery (the treading platform, the collection basin, and the mechanical press), and their function, are provided here.

The winery at Izbat Mohamed Farid, 1 km north of Abu Mina, had two phases of use (Grossmann 2002: 31). In the first phase, at the beginning or first half of the 5th century AD it was a simple pressing unit consisting of a treading platform and vat, with walls built of baked brick and covered with waterproof plaster. When it was no longer in use, another winery was built (later part of 6th century AD) partly on top of the old one. This new structure comprised a treading platform, a vat, and two mechanical presses. During both phases of use the winery stood next to a multistory country house.

Taking the winery at Marea as an example, it can be seen that a basic press unit was a compound consisting of a raised treading platform and a collection basin, built within an enclosed and most probably roofed, space (El-Fakharani 1983: 182-184 . The treading floor was separated from the vat by a low wall. It was covered with waterproof plaster, preserved, for instance, in the Burg el-Arab winery (El-Ashmawi 1998: 62), and was often quite large, as in the case of the Karm el-Baraasi winery (Abd el-Aziz Negm 1998: 66). The room in which the press unit was located was sometimes entered from a courtyard, or had a large window giving easy access to the treading floor. Grapes were thrown through this window, or simply carried through the door, to the treading platform, where they were crushed.

Three wineries near Abu Mina were explored in 1986 by the Supreme Council of Antiquities (Abd el-Aziz Negm 1998: 65-73). The winery at Karm Gadoura consisted of

5. For this information I am indebted to Prof. Dr. Marianne Bergmann and Prof. Dr. Michael Heinzelmann of the Swiss-German mission excavating at Schedia.

129

LAKE MAREOTIS: RECONSTRUCTING THE PAST Fig.2: A schematic drawing of a Type 4 w inery with names of its elements used in G reek papyri from Egypt ( D. Dzierzbicka) .

Wineries in Papyri Archaeological finds are not the only available sources on wineries (for more details on wine production in papyri see Dzierzbicka 2005: 24-89). Greek papyri of the Hellenistic and Roman period allow for the identification of words used to call the elements of a winery. The reconstruction drawing of Type 4 winery (Fig. 2) shows the basic parts of a winery and gives the terms by which they were called in papyrus documents. Matching the architectural features with their Greek names in papyri makes it possible for the researcher to use hundreds of texts that provide invaluable information on the function of the winery as a whole and on its elements, permitting a better understanding of the way these structures were used and provoking more questions to ask oneself in the field.

Most wineries were also equipped with a mechanical press either a movable bag press, or a more sophisticated, fixed screw press. Their preserved remains were found in Abu Mina (Grossmann, et al. 1997: 87-90) and in Burg el-Arab (El-Ashmawi 1998: 60-64), where one can still clearly see a raised, circular base of the screw press. These devices were used to squeeze the remaining liquid out of the trodden grape pulp either directly on the treading floor, or in an area separated from it, as in Burg el-Arab. In the latter case, a channel connected the place of the mechanical press with the main collection basin, or with a smaller, separate vat. The must flowed into a plastered vat or vats dug into the ground. It remained there through the first turbulent phase of fermentation, during which the basins were covered with wooden planks for protection against contaminants.

One of these questions, especially important in the case of the Mareotic wineries, concerns their surroundings. Papyrus documents containing references to buildings and other immovable property found in vineyards, provide interesting information on other installations forming part of a wine-making complex Table 1 . The table below is a comparative chart of the facilities mentioned in extant texts, showing how frequently they appear and in what combinations. The documents mention elements already discussed: the treading platform (lê nos) and collection basin (pithos), in a later period referred to by a common term – lê nopithos, a mechanical press (stemphylourgikon organon or mê chanê ), as well as a sunning yard (hê liastê rion), a storeroom (thesauros) and a cellar (oinothê kê ). In fact, spaces that may have been storerooms were uncovered adjacent to wineries, for instance in Abu Mina (Grossmann, et al. 1997: 89). Large wineries where mass production took place even had a potter’s workshop (keramikon ergastê rion) on the premises. There was indeed a large pottery kiln in the vicinity of the winery near Burg elArab, suggesting this may have been such a more developed production estate (El-Ashmawi 1998: 62).

Generally speaking, wineries exhibit a rather uniform structure, dictated by practical needs, that hardly changed regardless of whether the winery was located on the Mareotis lakeshore, in the Bahariya asis unpublished wine press photo in Hawass 2000: 164), in Nubia (Gardberg 1970: 41-44), or even on the Crimean Peninsula (Michalowski 1958: 49 . Mareotic wineries provide an excellent chance to study the characteristic traits of these structures. What is striking about the typology elaborated by Rodziewicz (1998: 31-36) is the diversity of wine making installations which led to the distinguishing of as many as six simple and two composite types in just over a dozen wineries. However, the distinguished types are in fact very useful as they have strong similarities with wineries in other parts of Egypt and beyond (e.g. Michalowski 1958: 49). As it was noted by Rodziewicz (1998), the wineries consist of two or three essential elements: the treading platform, the collection basin, and, in types 3-8, a mechanical press. The convenient uniformity of this category of structures makes them fairly easy to recognise in an archaeological excavation. 130

P.T ebt. III.1 814, Sale, after 227 BC, Tebtynis

X

P.L ond. II 401, Petition, 115–1 10 BC, Thebaid

X

P.de m.G ieben 2, Sale, 107–3 0 BC, Sebennytos

X

PSI VIII 918, Sale, AD 38/39, Tebtynis

X

X

X

P.R oss.G eorg. II 28, Lease, after AD 163/164, Arsinoites

X

P.F lor. III 385, 2nd–3 rd century AD, Hermopolites

O ikopedon

Pyrgos

K ella

Epoikion

Epaulis

X

X X X

PSI XIII 1328, Petition, AD 201, xyrhynchos

X

P.O xy. LI 3638, Cession, AD 220, Sinary, xyrhynchites

X

P.F lor. I 50, Division of property, AD 269, Hermopolis

X

P.O xy. XXXIV 2723, Sale, 3rd century AD, xyrhynchos

X

X

X

SPP XX 58 (7) Col. II, 3rd century AD, Hermopolis

X

X

X

SB XX 14291, Lease, 3rd century AD, xyrhynchos

X

X X

X

X

X

X

X

X

X

X

X

X

X

X

X X

X

P.V ind.Sal . 12, Lease, AD 334/335, Hermopolis

X X

X

X X

P.V ind.T and. 28, Lease, AD 576/577 Hermopolis

SPP XX 218, Lease, early 7th century AD, Hermopolites

X

X X

P.C airo Masp. I 67097 (r), Contract, AD 571/572 (?), Aphrodites Kome (Antaiopolites)

Thesauros

X

P.O xy. XLIX 3491, Marriage contract, AD 157/158, xyrhynchos

P. Vat. Aphrod. 25 fr. A, Division of property, 6th century AD, Aphrodito

O inothê kê

Hê liastê rion

K eramikon ergastê rion

Pithos

Stem. organon

Mê chanê

Lê nopithos

DOCUMENT

Lê nô n; lê nos

D. DZIERZBICKA: WINERIES OF THE MAREOTIC REGION

X X

X X

X

X

X X

X

Table 1: Installations forming part of a wine-making complex mentioned in G reek papyri containing references to immovable property found in vineyards or to the appurtenances of a winery. 131

LAKE MAREOTIS: RECONSTRUCTING THE PAST P. dem. G ieben 2 = Vandorpe, K., & Clarysse, W., 1998, A Greek Winery for Sale in a Fayum Demotic Papyrus. In A.M.F.W. Verhoogt & S.P. Vleeming (eds.), Two Faces of G raeco-Roman Egypt. G reece and Demotic and G reek-Demotic Texts and Studies Presented to P.W. Pestman: 127-139. Leiden, Boston, Köl n. P. Flor. I 50 = Vitelli, G. (ed.), 1906, Papiri greco-egizii, Papiri Fiorentini I. Documenti pubblici e privati dell’età r omana e bizantina. Milan. P. Flor. III 385 = Vitelli, G. (ed.), 1915, Papiri grecoegizii, Papiri Fiorentini III. Documenti e testi letterari dell‘ età r omana e bizantina. Milan. P. Lond. II 401 = Kenyon, F.G. (ed.), 1898, G reek Papyri in the British Museum II, 13-14, no. 401. London. P. O xy. XXXIV 2723 = Ingrams, L., Kingston, P., Parsons, P.J., & Rea, J.R. (eds.), 1968, The O xyrhynchus Papyri XXXIV, no. 2723. London. P. O xy. XLIX 3491 = Bül ow-Jacobsen, A., & Whitehorne, J.E.G. (eds.), 1982, The O xyrhynchus Papyri XLIX, no. 3491. London. P. O xy. LI 3638 = Rea, J.R. (ed.), 1984, The O xyrhynchus Papyri LI, no. 3638. London. P. Ross. G eorg. II 28 = Krüge r, O. (ed.), 1929, Papyri russischer und georgischer Sammlungen II. Ptolemäi sche und frühr öm ische Texte, no. 28. Tiflis. PSI VIII 918 = Vitelli, G., & Norsa, M. (eds.), 1927, Papiri greci e latini VIII, no. 918. Florence. PSI XIII 1328 = Vitelli, G., & Norsa, M. (eds), 1949-1953, Papiri greci e latini XIII, no. 1328. Florence. P. Tebt. III.1 814 = Hunt, A.S., & Smyly, J.G. (eds.), 1933, The Tebtunis Papyri III.1, no. 814. London. P. Vat. Aphrod. 25 = Pintaudi, R. (ed.), 1980, I Papiri Vaticani di Aphrodito, no. 25. Rome. P. Vind. Sal. 12 = Salomons, R.P. (ed.), 1976, Einige Wiener Papyri, no. 12. Amsterdam. P. Vind. Tand. 28 = Sijpesteijn, P.J., & Worp, K.A. (eds.), 1976, Fünf unddreissig Wiener Papyri, no. 28. Zutphen. SB XX 14291 = Bilabel, F., Kiessling, E., & Rupprecht, H.-A. (eds.), 1997, Sammelbuch griechischer U rkunden aus Aegypten XX, no. 14291. Wiesbaden. SPP XX 58 = Wessely, C. (ed.), 1921, Studien zur Palaeographie und Papyruskunde X X . Catalogus Papyrorum Raineri. Series G raeca. Pars I. Textus G raeci papyrorum, qui in libro „ Papyrus Erzherzog Rainer– Führ er durch die Austellung Wien 18 94 “ descripti sunt, no. 58. Leipzig.

The winery was not just a press unit, but a building or a set of buildings, located either in the urban area or in the vicinity of a vineyard. According to some documents the complex was also equipped with farm buildings epoikion, pyrgos, epaulis), storehouses (kella) and building plots (oikopedon). Indeed, the treading platform and vat at Abu Qir Bay was part of a sizeable square structure built of limestone blocks. The layout of rooms in the building is only partly preserved (Breccia 1926: 47-49). It has not been determined where the entrance to the winery was located and whether there were passages leading to the adjacent rooms. The wine-making complex in this structure may have not been limited to only one unit, consisting of a treading platform and vat, especially that the other rooms also had an industrial function, perhaps related to wine and oil production. As we learn from papyri (Dzierzbicka 2005: 24-89), after the turbulent phase of fermentation was over, the wine was poured into vessels that had been obtained from a pottery workshop located within the complex or acquired from outside the winery. Wine that was paid for in advance was distributed straight from the vat and carried off from the winery in vessels provided by the buyer. The filled jars that remained in the winery were placed in storage rooms or in the sunning yard (hê liastê rion) for maturing. Both of these facilities were usually spaces or buildings within the wine-making complex and they were equipped with doors and locks. Maturing wine was occasionally controlled to check if it had not gone sour or spoiled. After the process of fermentation had ended, the vessels were sealed. There are still many uncertainties about the broader context of wineries. The immediate surroundings of a winemaking structure can provide information on the building that housed the installation and determine whether it was located in a rural setting or in a larger industrial or even residential area. Further archaeological investigations, as well as the study of literary and documentary sources, may provide insight into industrial buildings and structures found in the vicinity of a winery and improve our understanding of the character of wine production in the Mareotic region. Acknowledgments I am grateful to the Polish Centre of Mediterranean Archaeology of the University of Warsaw for assistance granted to me during my stay in Egypt during the conference.

Ancient Sources Athenaeus, The Learned Banqueters, books I-III.106e, S.D. Olson (transl. & ed.), 2006. Cambridge. Horace, O des I, Carpe Diem, West, D. (transl. & ed.), 1995. xford. Strabo, G eography, H.L. Jones (ed.), 1996. London, New York. Virgil, Eclogues. G eorgics. Aeneid 1- 6, Vol. 1, H.R. Fairclough (ed.), 1974. Cambridge.

Bibliography Papyri P. Cairo Masp. I 67097 (r) = Maspero, J. (ed.), 1911, Papyrus grecs d’époque byzantine. Catalogue général des antiquités égyptiennes du Musée du Caire I, no. 67097. Cairo.

132

D. DZIERZBICKA: WINERIES OF THE MAREOTIC REGION Gardberg, C.J., 1970, Scandinavian J oint Expedition to Sudanese Nubia, Vol 7: Late Nubian Sites: Churches and Settlements. Astrom. Grossmann, P., Kosciuk, J., Severin, G., & Severin, H.-G., 1984, Abu Mina. Elfter Vorl ufiger Bericht. Kampagnen 1982–1 983. Mitteilungen des Deutchen Archäol ogischen Instituts Abteilung K airo 40: 148. Grossmann, P., Arnold, ., & Kościuk, ., 1997, Excavations at Abu Mina 1995. Bulletin de la Société d’Archéologie Copte 36: 87-90. Grossmann, P., 1999, Report on the excavations at Abu Mina in spring 1998. Bulletin de la Société d’Archéologie Copte 38: 75-84. Grossmann, P., 2002, Report on the excavations at Abu Mina in spring 2001. Bulletin de la Société d’Archéologie Copte 41: 25-31. Hawass, Z., 2000, The Valley of the G olden Mummies. Cairo. Michalowski, K., 1958, Mirmeki I. Warszawa. Mül ler-Wiener, W., & Grossmann, P., 1967, Abu Mina. 6. Vorl ufiger Bericht. Archäol ogischer Anzeiger 82.4: 468-473. Rodziewicz, M.D., 1998, Classification of wineries from Mareotis. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athè nes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole franç aise d’Athè nes, 11- 12 décembre 19 8: 27-36. Athens, Paris.

Secondary Sources Abd el-Aziz egm, M., 1998, Recent excavations around Abou Mina. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athè nes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole franç aise d’Athè nes, 11- 12 décembre 198: 65-73. Athens, Paris. El-Ashmawi, F., 1998, Pottery kiln and wine-factory at Burg el-Arab. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athè nes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole franç aise d’Athè nes, 11- 12 décembre 198: 55-64. Athens, Paris. Blue, L., & Khalil, E. (eds.), forthcoming, A Multidisciplinary Approach to Alexandria’s Economic Past: the Mareotis Case Study. xford Breccia, E., 1926, Le rovine e i monumenti di Canopo, eadelfia e il tempio di Pnefer s. Bergamo. Dzierzbicka, D., 2005, Wineries in Graeco-Roman Egypt. J ournal of J uristic Papyrology 35: 9-89. Empereur, J.-Y., 1993, La production viticole dans l’ Egypte ptolémaï que et romaine. In M.-C. Maouretti & J.P. Brun (eds.), La production du vin et de l’huile en Méditerranée: 39-47. Athens. El- akharani, ., 1983, Recent Excavations at Marea, Egypt. Aegyptiaca Treverensia 2: 175-186. Forbes, R.J., 1955, Studies in Ancient Technology, Vol 3. Leiden.

133

LAKE MAREOTIS: RECONSTRUCTING THE PAST

134

E. KHALIL: MAREOTIS’ WATERFRONT

Waterfront Installations and Maritime Activities in the Mareotic Region Emad Khalil

Introduction Lake Mareotis represents one of the most distinctive geomorphic features in the north-west coastal region of Egypt. In antiquity, it was fed by means of a number of canals, which bifurcated off the Nile’s defunct Canopic Branch, and flowed into the southern and eastern sides of the lake (Fig. 1). Some of these canals were navigable, which enabled merchandise to be transported to and from the hinterland. By the Greco-Roman period the lake was also connected to the sea through a navigable canal that debouched at Alexandria (Strabo 17.1.7). Its connection to both the Nile and the sea resulted in Mareotis becoming a vital conduit of communication in Egypt’s internal transport system. Moreover, it supported around its shores various agricultural activities and embraced major production centres for different industries, which contributed significantly to the economy of Alexandria and to Egypt as a whole. Accordingly, this paper will look at the role that Lake Mareotis played in the ‘maritime’1 transport system of GrecoRoman Egypt. It will also examine the types and nature of the maritime and waterfront installations that were recorded along the shores of the lake and the possible spatial and functional relationship between the different sites.

(Strabo 17.1.14; Pliny 5.11.63). It comprised a main rectangular body of water which merged to the east and south with the Nile Delta Plain, and a narrow arm that extended westwards parallel to the northern coast. However, during the past two millennia Lake Mareotis has undergone dramatic changes which significantly affected its size and nature. Nonetheless, the western extremities of the lake reflect the original extant remains, and form an arm, known as the Mareotic Arm, that extends some 40 km west of Alexandria and is 2-3 km wide and is separated from the lake’s main body by causeways and shallows. It also contains an island, Mareotis Island, which is 3.7 km long and about 680 m at its widest point (see both Blue and Hopkinson this volume).

The Lake Mareotis Research Project Much evidence indicates that Lake Mareotis extended in antiquity for about 50 km south and west of Alexandria

Since 2004 the Centre for Maritime Archaeology (CMA) of the University of Southampton, in collaboration with the Department of Underwater Antiquities (DUA) of the

Archaeological investigation of the western Mareotic Arm has been ongoing for several decades; however, previous research has been largely limited to specific areas and specific issues such as work on the Byzantine port of Marea/ Philoxenité (Petruso & Gabel 1982; El-Fakharani 1983; Rodziewicz 2003 , and work on amphorae and wine production (Empereur & Picon 1986, 1998) and on the wineries of the Mareotic region Rodziewicz 1998b .

Fig.1: The approximate ancient and present limits of Lake Mareotis ( E.K halil) .

1. The term ‘maritime’ is used in this sense to denote all aspects of waterborne activity and communication from the sea, across the lake, along the canals and on the Nile River.

135

LAKE MAREOTIS: RECONSTRUCTING THE PAST Fig.2: The Lake Mareotis Research Proj ect survey area along the shores of the western Mareotic Arm ( Lake Mareotis Research Proj ect) .

Waterfront Sites in the Mareotic Region The waterfront sites that were recorded along the shores of the western Mareotic Arm and on Mareotis Island can be classified into four categories:

Egyptian Supreme Council of Antiquities (SCA), has conducted a comprehensive archaeological survey along the shores of western arm of Lake Mareotis (see Blue this volume). The survey revealed a wealth of archaeological sites including numerous settlements of a maritime and industrial nature, which reflect the economic activities that took place in the Mareotic region (Fig. 2). Most of the archaeological sites recorded along the shores of the Mareotic Arm are located between Marea/Philoxenit and Taposiris Magna, with a concentration on Mareotis Island. The sites identified include maritime structures, such as harbours, jetties and quays, in addition to what appear to be waterfront warehouses and storage facilities. Sites pertaining to industrial activities are also evident and include amphora kilns, ceramic slag and kiln wasters, as well as a number of structures relating to water management, such as cisterns, wells and water wheels (sakkia). This correlates with much archaeological and textual evidence for viniculture and wine production in the region Rodziewicz 1998b .

1 - Harbours in the form of harbour complexes of significant design and constitute substantial structures. However, the only two sites that fit this description are Marea/Philoxenité and Taposiris Magna. The complexity and magnitude of their structures are unrepresented elsewhere in the entire Mareotic region. The two harbours date, however, to quite different periods. Taposiris Magna is essentially Hellenistic in date (see Boussac & El Amouri this volume), while Marea/Philoxenit mostly dates to the Late Roman period see Haggag, Pichot, Rodziewicz and Babraj & Szymańska this volume . However, the two harbours are associated with relatively large towns, and much historical and archaeological evidence indicates that these two towns where probably among the largest and most active along the shores of Lake Mareotis in antiquity.

The dating of the sites discovered relies primarily on the ceramic assemblages collected during the survey. Accordingly it was realised that the majority of sites date from the Hellenistic period until the 7th century.

Taposiris Magna has extensive archaeological remains that date from the Hellenistic to the Late Roman period, including evidence for thriving maritime and commercial 136

E. KHALIL: MAREOTIS’ WATERFRONT activities (El-Fakharani 1974; Empereur 1998: 225-7; Rodziewicz 1998a Boussac & El Amouri this volume . The town which is located on the northern shore of the western sub-basin of the lake has one of the best preserved harbours in Lake Mareotis (see Boussac & El Amouri this volume). Commercial activities in Taposiris Magna were mainly focused on handling products transported across the lake as well as receiving goods arriving from the west through overland routes, and shipping them to Alexandria (Empereur 1998: 225). Under the Romans, Taposiris Magna was a customs station where duties were levied on products coming from the Mareotic region and from Cyrenaica heading east towards Alexandria or to the Nile Delta (Empereur 1998: 225-7; Vör ös 2001: 15-6). Alternatively, river vessels could have travelled on the lake through the harbour of Taposiris Magna to the west as far as the lake extended.

ings to allow water to flush away the silt and sediments that might accumulate in the harbour basin. This arrangement was supplemented by the construction of a 1,700 m long wall that extends southwards from the artificial ridge to the southern shore of the lake. It was also supplemented by the construction of a solid limestone wall that extended from the northern shores of the lake to the seashore. This wall ensured that even caravans travelling overland had to go through the town of Taposiris Magna (see Fig. 1 in Boussac & El Amouri this volume) De Cosson 1935: 111 Rodziewicz 1990: 72-4 . Another prominent structure in Taposiris Magna is a 17 m high tower that stands on the coastal ridge to the north of the harbour over looking the Mediterranean coast to the north and the lake to the south (Kadous 2001: 457-60). The function of the tower and its relation with the town and harbour of Taposiris Magna is disputed (El-Fakharani 1974; Vör ös 2001: 37). Nonetheless, it is generally accepted that the tower, which represents a 1:4 or 1:5 replica of the Pharos lighthouse of Alexandria, was in fact a funerary monument for a Hellenistic necropolis that occupied the area around and below the tower (Empereur 1998: 225). The utilisation of such a structure as a landmark by navigators on both the sea and the lake is a possibility that cannot be overlooked (Fig. 3)

The harbour of Taposiris Magna was constructed to control the movement of vessels travelling through it. This was achieved by digging a channel c. 1,700 m long and 50 m wide parallel to the northern shore of the lake. The spoil resulting from the digging was piled up to form an artificial ridge which delimits the channel from the south. At the western end of the channel stood a limestone structure that took the form of a double-opening gate or bridge, through which all boats wishing to go through Taposiris Magna had to pass Empereur 1998: 225-7 Rodziewicz 1998a: 102, n. 32; Vör ös 2001: 15-6) (see Figs. 5-7 in Boussac & El Amouri, this volume). The total width of the gate is about 8.3 m, however, it is divided by a 1.2 m thick wall into two openings; one is 4.1 m wide and the other one is 3 m wide, thus indicative of the maximum possible width of the vessels that passed through. The eastern entrance of the channel is partially obstructed by a quay which is c. 230 m long extending from north to south perpendicular to the shoreline. The distance between the southern end of the quay and the eastern end of the artificial ridge, c. 100 m, forms the eastern entrance of this semi-closed harbour basin of Taposiris Magna. The eastern quay of the harbour includes at least two de-silting open-

As a result of the recent excavation of the area, it is now believed that the digging of the channel as well as the construction of this harbour system, took place the during the Early Roman period rather than during the Hellenistic period, as was previously believed (El-Fakharani 1974; Boussac & El-Amouri this volume). Although the northern shoreline at Taposiris Magna was occupied during the Hellenistic period, as evidenced by houses and shops from the 2nd and 1st century BC, it seems that it was abandoned by the end of the Hellenistic period as a result of a rise in the lake level. The flooded area was then excavated in the Roman period to create the closed harbour system (see Boussac & El Amouri this volume). Marea/ Philoxe nité is located about 15 km east of Taposiris Magna, on the southern shore of the lake. El-Falaki 1966: 96 identified this settlement and its associated harbour as the town of Marea, the capital of the Mareotic region. According to Herodotus (2.31), Marea was a post of Egyptian soldiers guarding the Libyan border during the time of King Psammetichus of the 26th Dynasty Rodziewicz 2003: 27 . In the Ptolemaic and Roman periods, Marea functioned as a major trade centre, second only to Alexandria. In the Byzantine period, in addition to its involvement in commercial activities and internal trade, Marea flourished as a stopover for pilgrims heading to the holy Byzantine shrine of St. Minas Abu Mina , 20 km south of the lake Gabel & Petruso 1980 Kucharczyk 2002 Rodziewicz 2003 . ntil recently, most archaeological research carried out in the area has revealed no evidence earlier than the 5th century AD. However, recent archaeological investigation at Marea has revealed material

Fig.3: Taposiris Magna tower ( photo: E.K halil) .

137

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig.4:

The middle quay of the harbour of Marea ( photo: E.K halil) . Excavations on the peninsula at the easternmost part of Marea, resulted in the discovery of a large Early Roman building which consists of a courtyard surrounded by numerous rooms of relatively similar size. The building is connected to a quay to the north through stairways cut into the rock. Therefore, it has been suggested that the building could have been used for storage and trade. The remains of 1st century BC to 1st century AD workshops for metalwork were also discovered on the peninsula (see Pichot this volume). The recent archaeological discoveries at Marea, particularly of pottery and coins, would indicate that the area was thriving before the 5th century and possibly as early as the Hellenistic period

from the Hellenistic and Roman periods (see both Pichot and Babraj & Szymańska this volume . Amongst the most significant archaeological remains in Marea are four quays that extend into the lake and divide the 1.5 km long shoreline into eastern, central and western harbour basins (Figs. 4 & 5). The dimensions of the quays from west to east are: 41 m x 6.5 m, 111 m x 5 m, 125 m x 7 m, and 35 m x 4 m Szymańska & Babraj 2008: 1115). Judging from the construction technique of the quays at Marea, in which large regular limestone blocks (1 m x 0.5 m x 0.03 m) were used for their construction, it seems reasonable to suggest that the four quays were constructed earlier than the Byzantine city. However, they were probably subject to several building phases in subsequent periods since evidence of Byzantine hydraulic mortar opus signinum) can still be seen between many of the building blocks of the quays.

2 - The second category of waterfront sites that was recorded along the shores of the western Mareotic Arm consists of different types of anchorage facilities such as quays and jetties, which form the majority of maritime Fig.5: The quay of Marea harbour was made of large regular limestone blocks without evidence of the use of mortar ( photo: E. K halil) .

138

E. KHALIL: MAREOTIS’ WATERFRONT

Fig.6: The box-shaped harbour on the southern shore of Lake Mareotis ( photo: Lake Mareotis Research Proj ect) . installations in the region. More than ten different anchorage facilities were recorded on the northern and southern shores of the lake and on the northern shore of Mareotis Island. Possibly the most substantial of them is a Kibotos or box-shaped harbour (Site 09) which is located at the end of a promontory on the southern shore of the lake, about 2 km to the west of the Sidi Kerir-Borg El Arab road. The harbour, which is constructed of large limestone blocks, consists of two parallel moles enclosing an area some 60 m long (N to S) and 36 m (E to W) wide (Fig. 6). The eastern mole is 60 m long and at its northern extremity it returns to the west for a distance of 12 m, while the western mole, which is less well preserved, extends for 40 m and returns to the east at the northern end for a distance of some 6 m. A gap of 18 m between the two ends of the two moles equates to the entrance of the harbour on the north side. The moles are constructed of up to three courses of single and double breadth limestone blocks. Although the dating for this harbour is still uncertain, judging from its construction technique and from the large size of blocks used c. 1.10 x 0.7 x 0.5 m), it seems that the harbour is pre-Roman in date. However, the existence of lumps of coarse red mortar opus signinum with lime inclusions between some of the blocks indicate that it remained in use at least until the Byzantine period. ne carved mooring ring was noted on the upper course of one of the blocks, which would have helped facilitate the mooring of vessels to the outside of the harbour (El-Fakharani 1984; Blue & Ramses 2006).

They are Sites 204 (Gamal) and 208 (Quseir). The jetties at each of these sites, which are located approximately 2 km apart, are about 50 m long and 8 m wide. Evidence of red mortar opus signinum is noted between some of the blocks. Also the jetty at Site 208 (Quseir) had mooring stones extending from the upper course of blocks at the eastern side (the lee side) of the quay. Other jetty-like features, although not as substantial, are located along the southern shore of the lake and the northern shore of the Mareotis Island. However, it was realised during the survey that the anchorage facilities along the northern shore of the lake are mostly associated with civic and residential sites, while those along the southern shore of the lake, particularly on Mareotis Island, are associated with sites of a commercial nature. The dating of these sites based on ceramic collections is quite problematic since the jetties are continuously washed by water in the winter, which, in many cases, does not leave any ceramics to be dated. However, judging from the ceramics dated from adjacent sites, it was realised that most of the sites could have been used for quite a long period of time, probably from the Hellenistic to the Late Roman or Byzantine period. 3 - The third type of maritime installation identified can be described as seawalls or more accurately lake walls. Unlike the jetties, which are perpendicular to the shore, lake walls parallel the shore and they were intended to define the shores and protect them from the effects of silting and sedimentation ig. 8 . At least five lake walls were discovered in the survey region. These kinds of structures are mainly found along the southern shore of the lake and the northern shore of Mareotis Island, those shorelines most subject to silting and the deposition of sediments as a result of the prevailing northwest winds that would carry sediment from the coastal ridges and deposit it into the lake. Besides acting as a form of protection against silting, the lake walls could also have been utilised as docking facilities for merchant vessels. Another possible function for such structures was to retain rainwater for use in agricultural purposes.

With the exception of this square harbour on the southern shore of the lake, all the other anchorage facilities along the shores of the lake take the form of jetties and quays that extend into the water perpendicular to the shore (Fig. 7). The technique used for the construction of most of the quays was building two parallel single or double breadth piers of limestone and filling the distance between them with rubble. With the exception of the substantial structures already noted at Marea/Philoxenit and Taposiris Magna the two most substantial anchorages are located on the northern shore of the lake opposite Mareotis Island. 139

LAKE MAREOTIS: RECONSTRUCTING THE PAST Fig.7: O ne of the quays extending into the lake at the northern shore of Mareotis western arm ( photo: Lake Mareotis Research Proj ect) .

Fig.8: ( below) A lake wall extending parallel to the southern shore of the lake.E vidence for red mortar (opus signinum) can still be seen between the blocks ( photo: Lake Mareotis Research Proj ect) .

The longest of these lake walls is located on the north shore at the western end of Mareotis Island (Site 21). It is c. 245 m long and 1 m wide and is constructed of a series of limestone blocks laid as stretchers along the lake edge (see Hopkinson this volume). Other examples include a wall located in the middle of the island on the north shore, the extant remains of which extended some 70 m in length and was made of one course of large limestone blocks of 0.60 m x 0.30 m x 0.25 m dimension, that were arranged as headers facing the shoreline (Site 123). Similarly, on the southern shore of the lake further substantial walls of over 250 m in length were identified at both Sites 109 and 44. 4 - The fourth and final type of waterfront structures does not necessarily have a maritime function. At a number of sites in the survey region the remains of several multi room buildings were recorded very close to the present waterline. Some structures even extend into the water (Fig. 9). Examples of this type of structure are found in Sites 117, 118 and 119 which are located at the north-eastern shore of the island. At Site 117 there are the remains of a multi-roomed building that measures 12 m EW x 17 m NS which is divided on the inside into at least four smaller rooms. Site 118, about 25 m west of Site 117, contains the remains of at least two multi-roomed structures which measure 18 m EW x 15 m NS and 20 m EW x 20 m NS. Each of them contains the remains of numerous walls which belonged to a number of internal rooms of different shapes and sizes. About 40 m to the west of Site 118, a further Site 119 contains the remains of a rectangular building that extends from the shoreline southwards for about 40 m and measures about 25 m EW. Limited excavation carried out in the middle section of this building revealed the remains of two 140

E. KHALIL: MAREOTIS’ WATERFRONT

Fig.9: Some of the multi-room square buildings located at Site 118al ong the northern shore of the Mareotis Island.T he buildings could have been used for storage purposes ( image: Lake Mareotis Research Proj ect) . significant structures. The first structure is a rectangular enclosure that measures c. 9 m x 4.5 m that contained two rooms. Both rooms have almost the same dimension c. 3 m EW x 2.7 m NS. The second structure constructed on the same alignment as the first and adjacent to it to the west, is a rectangular building that measures c. 2.5 m NS and at least 5 m EW, in which remains of imported Hellenistic amphorae were discovered.

Marea and Taposiris Magna, only a distance of some 15 km, but out of a total of more than ninety archaeological sites recorded along the shores of the Mareotic Arm west of Alexandria, only four substantial sites were recorded to the west of Taposiris Magna. Thus, judging from the nature and extent of these sites, it is evident that the navigable limits of Lake Mareotis in antiquity extended west of Taposiris Magna for at least 12 km.

Accordingly, it seems reasonable to suggest that this type of waterfront structure were used as storage facilities for different merchandise and products that were traded along the Mareotic Arm.

Likewise, it is noticeable that maritime installations located on the southern shore of the lake from Marea to the eastern end of Mareotis Island, are in fact located on a ridge that extends for about 800 m from the present southern shoreline to the west. It is noteworthy that no sites were recorded on the southern shore of the lake opposite this ridge. This actually raises a question about the nature of the area between the ridge and the southern shore of the lake, and whether or not it was actually land in antiquity, that has subsequently been artificially excavated or subject to inundation due to the changes in ground water level (see Flaux forthcoming). Similarly, all the archaeological sites of a maritime nature that were recorded on Mareotis Island, where located along its northern shore, with essentially no evidence for sites either on the southern shore of the island or on the northern shore of the lake opposite. This also raises a question about the nature of this island and whether or not it was actually an island in antiquity.

By examining the remains of these structures, particularly on Maerotis Island, it becomes evident that they have undergone construction phases over successive periods. Moreover, it seems that the sections of the structures closest to the waterline were subject to accumulated sediments, and hence had to be rebuilt. In other words the different phases of building and modification of structures could be the result of adapting to changes in the waterline. a ns an S n an By looking at the distribution of archaeological sites in general and maritime sites in particular along the shores of the Mareotic Arm, it becomes evident that not only is there an apparent concentration of sites in the area between 141

LAKE MAREOTIS: RECONSTRUCTING THE PAST N avi gation in Lake Mareotis In the 1st century BC, when speaking about the water supply for Lake Mareotis, Strabo (17.1.7) mentions that it is filled by many canals from the ile, both from above and on the sides, and through these canals the imports are much larger than those from the sea, so that the harbour on the lake was in fact richer than that on the sea”. On another occasion (Strabo 17.1.22) speaks of “…s everal canals, which empty into Lake Mareotis”. As a result, it has been assumed that there was intense maritime traffic passing through the lake carrying various products and cargoes to Alexandria. Merchandise which would have been transported to Alexandria for local consumption and for transshipment to other Mediterranean harbours would have included Egyptian products such as papyrus, textile and grain (Rickman 1971: 300-6, 1980: 231-5; Lewis 1983: 165-7), as well as quarried stones from the Eastern Desert (Peacock 1992: 5-7; 2002: 426-7). It would have also included products imported via the Red Sea from India, Arabia and East Africa such as spices, tortoiseshell, frankincense, ivory, cotton, silk and gems (Strabo 2.5.12; Casson 1991: 200-212; Peacock 2002: 432-3). At the same time, Alexandria was receiving from the Mediterranean, for local consumption and for transshipment to the south, various products such as wine, oil, fine pottery, glass, timber, copper and tin. et, the role that Lake Mareotis played in this internal transport system is some what unclear.

The different nature of maritime installations located along the northern shore of the lake and those along the southern shore and on Mareotis Island, is also noteworthy. As mentioned earlier, maritime installations such as quays and jetties located the northern shore of the lake are mostly associated with large tell sites of a civic and residential nature. These tells were up to 60,000 m2 in area, and were densely occupied. They are mainly covered with building stones and the foundations of buildings, as well as the remains of several wells, cisterns and red brick basins lined with opus signinum, which could have been used in baths, houses or other urban structures. However, the situation on the southern shore is quite different since the southern shore is where most industrial and commercial sites were recorded, and hence maritime installations were mostly associates with those sites. In antiquity, almost all the amphora and wine production sites in the region, as well as sakkia installations (Empereur & Picon 1998 Rodziewicz 1998b see also the individual contributions of Blue, Hopkinson and Dzierzbicka this volume), were located on the southern shore of the lake, the focus of agricultural and industrial activities. The reason that agricultural and industrial activities where concentrated on the southern shores of the Mareotic Arm is mainly due to the difference in the topography between the northern and southern shores of the lake. The western arm of Lake Mareotis is delimited from the north and the south by two limestone carbonate ridges, of average elevation 25-35 m and average width 300 m (Said 1990: 499; Warne & Stanley 1993; El-Raey, et al. 1995: 191; Frihy, et al. 1996: 282). The northern ridge is known as the Abusir Ridge, and to the south a longitudinal depression 3-4 km wide known as Al-Alamein-Maryut Depression, extends roughly E-W, partly occupied by the western arm of Lake Mareotis. This depression is delimited to the south by another coastal ridge, known as Gebel Maryut Ridge, which is located 5-9 km south of the Abusir Ridge.

Although it is well known that Lake Mareotis was fed by means of a number of canals, which branched off the Canopic Branch, and flowed into the southern and eastern reaches of the lake, there is a considerable degree of uncertainty about the exact number, location and the routes of these canals. However, the most important of these canals was Schedia Canal see Bergmann, Heinzelmann & Martin this volume) It bifurcated off the Canopic Branch of the Nile at the town of Schedia, originally a Hellenistic foundation that was later known as Chaereu, currently located in the region of the villages of Kom El-Giza, Kom El-Nashw and Kom El-Hamam, some 30 km south-east of Alexandria Bergmann & Heinzelmann 2004 . While the Canopic Branch continued north to debouch into the Canopic Bay (Abukir Bay), the Schedia Canal turned northwest towards Alexandria and followed a course close to the present course of the Al-Mahmoudeyah Canal. In a statement by Strabo (17.1.16) in which he describes the town of Schedia, he mentions that it has “…t he station for paying duty on the goods brought down from above it and brought up from below it; and for this purpose, also, a schedia float has been laid across the river, from which the place has its name”. Accordingly, Schedia was the main Nile emporium, customs harbour and checkpoint east of Alexandria, where custom duties were imposed on imported and exported goods (Empereur 1998: 225; Bergmann & Heinzelmann 2004 . Moreover, it seems that the Canopic Branch at Schedia was obstructed by some kind of a pontoon that prevented boats from sailing past it until duties were paid on merchandise travelling both ways. Additionally, it was at Schedia where exported commodities

Accordingly, the distance between the Gebel Maryut Ridge and the southern shore of the lake is far greater than the distance between the Abusir Ridge and the northern shore of the lake. In antiquity, the area south of the lake was a fertile plain that flourished with agricultural activities and was known for the quality of its vines, olives and fruits as well as for the cultivation of flax and papyrus Athenaeus 1.33.d-e; Pliny 13.22.71; Strabo 17.1.14; Empereur & Picon 1998; Horden & Purcell 2002: 353; McGovern 2003: 121-3). Moreover, the abundance of calcareous clay, particularly suitable for amphora production, resulted in a thriving large-scale amphora industry along the southern shore of the lake (Empereur & Picon 1986: 103-9, 1992, 1998 Rodziewicz 1998 Blue & Ramses 2007 . Thus, the focus of wine and amphora production in Hellenistic and Roman times was the southern shores of the Mareotic Arm. Thus, it is reasonable to suggest that those maritime installations located along the southern shore of the lake were very much involved in commercial activities including the transport of Mareotic products to Alexandria and possibly also to the southern limits of the lake. 142

E. KHALIL: MAREOTIS’ WATERFRONT brought from upriver were transferred from large Nile boats to smaller boats that could travel easily through the canals to Alexandria (Procopius 6.1.3; Hassan 1997: 365 n. 13).

sequently, the lake’s southern and eastern shoreline was unstable and subject to constant change, and was therefore unsuitable for the establishment of substantial harbours and waterfront installations. A recent survey conducted along the ancient southern and eastern limits of the lake (Wilson 2007; see Wilson this volume) revealed that most settlements established in this area during the Ptolemaic and Roman periods, were located on high ground around the lake’s edge. Also it revealed that many settlements were involved in agriculture and industrial activities mainly in the service of Alexandria. However, as yet there is no evidence for substantial maritime structures or significant waterfront installations.

As it approached Alexandria, the Schedia Canal bifurcated into two branches in the Alexandrian suburb of Eleusis El- ozha . The first branch turned towards the north-east leading to Canopus, east of Alexandria, while the other branch continued south of Alexandria and parallel to the lake’s northern shore, until it debouched into Lake Mareotis south-east of Alexandria. According to Strabo (17.1.7), boats also sailed from the Nile to the Canopic Branch and through the network of canals that fed the lake from the south and east, then across the lake northwards to Alexandria. This indicates that navigation on Lake Mareotis was intense and operated in many directions. It also raises a point about the practicalities of sailing in Lake Mareotis from south to north against the prevailing north-westerly wind. The predominant winds along the north coast of Egypt are north-westerly and they prevail more than 40% of the time throughout the year and more than 70% of the time during the summer sailing season (El-Zouka 1979: 125-7; El-Gindy 1999: 17). Accordingly, merchant vessels sailing in Lake Mareotis from south to north would have faced a direct headwind, which meant that the boats had to tack in order to reach Alexandria. Tacking in Lake Mareotis was possible considering the large area of the water body; however, tacking from the southern limits of the lake to Alexandria would have meant that boats would have to travel several times the direct distance across a water body full of shallows and marshlands and against prevailing winds. In the 5th century St. Palladius (7.1) mentioned that he sailed across Lake Mareotis from north to south, from Alexandria to the monastic settlement of Mount Nitria, a distance of about 50 km, in a day and half. Accordingly, sailing in the lake in the opposite direction would have taken much longer, possibly as long as four to five days.

C onclusion There were two ways for river vessels to travel to and from Alexandria, either across the lake, or along the Schedia Canal. Considering the arguments outlined above, particularly in relation to the prevailing winds, it seems reasonable to suggest that the main northbound traffic probably went via the Canopic Branch and the Schedia Canal, rather than across the lake. However, sailing south across the lake would have been conducted with considerable ease. Along those stretches of the canal where boats had to maneouver against the wind, they could have been towed along from the shore, a standard procedure for moving river boats in rivers and canals around the world. In that respect, Strabo’s statement (17.1.7) about the lake harbour south of Alexandria being richer than the seaport of Alexandria, would still be valid. At the time of Strabo, the Schedia Canal debouched into Lake Mareotis, so all the canal traffic had to pass through the lake. Moreover, boats coming from the western arm of the lake also arrived at the lake harbour. Therefore, it is possible that the lake harbour was quite busy receiving river vessels from the south as well as from the west. At the same time, it is not unreasonable to suggest that east–w est commercial traffic along the western Mareotic Arm was probably more intense and more regular than the north south traffic that passed through the main body of the lake. Recent archaeological investigation in the Mareotic region have revealed that the number, nature and extent of archaeological sites along the shores of the Mareotic Arm, is unparalleled any where else in the Western Deltaic Region (see Blue this volume). Settlements in this region were located far from the silting effects of the Nile sediments, the coastline was more stable, prevailing winds were more favourable for east-west movement, and settlements were in close proximity to Alexandria. Therefore, it is understandable why so many shoreline settlements and associated maritime installations were established along its shores. Thus, the contribution of the western Mareotic Arm to the economy of ancient Alexandria and hence of Egypt as a whole, was probably far more significant than any other area along the shores of Lake Mareotis. Thus, the shores of the western arm of Lake Mareotis appear to have been one of the most active areas of economic activity in the Western Deltaic region during the Hellenistic, Roman and Byzantine periods.

Furthermore, the extended period of travel across the lake would have laid boats venerable to another challenge that prevailed on Lake Mareotis in antiquity. Achilles Tatius (4.12) in the 2nd century and Heliodorus (1:14) in the 3rd century, spoke of piracy and bandits on Lake Mareotis. The marshes and islands of the lake provided excellent hideouts for groups of bandits and their vessels. Also, the large size of the lake made it quite difficult to guard and control, therefore, it is possible that sailing across the lake with valuable commodities was quite risky. Moreover, settlements located on the southern and eastern shores of Lake Mareotis were far more susceptible to sedimentation from silts deposited via the nearby Canopic Branch of the ile, particularly during flood seasons, as well as sediment which had been carried by the prevailing winds across the lake from the north-west to the southeast. All this would have contributed to the build up of sediments against the southern and eastern shores of the lake, thus preventing settlements in this region from being as actively involved in across lake transportation. Con143

LAKE MAREOTIS: RECONSTRUCTING THE PAST Bibliography Achilles Tatius, The Adventures of Leucippe and Clitophon, S. Gaselee (transl.), 1969. London. Athenaeus, eipnosophistae C.B. Gulick (transl.), 1953. London. Bergmann, M., & Heinzelmann, M., 2004, Schedia ( K om el Gi ah om el amam epartment of eheira Report on the documentation and excavation season. 18 March – 18 April 203. Paper presented at conference on Alexandria in Antiquity, University of Oxford, December 2004. Blue, L., & Ramses, S., 2006, Lake Mareotis Research Proj ect. Report submitted to the Egyptian Supreme Council for Antiquities on the fieldwork and results of the September 2006 field season. Blue, L., & Ramses, S., 2007, Lake Mareotis Research Proj ect. Report submitted to the Egyptian Supreme Council for Antiquities on the fieldwork and results of the May & uly/August 2007 field seasons. Casson, L., 1991, The Ancient Mariners. Princeton. De Cosson, A., 1935, Mareotis: Being an Account of the istory and ncient Monuments of the orth est Desert of Egypt and Lake Mareotis. London El-Ashmawi, F., 1998, Pottery Kiln and Wine Factory at Burg El-Arab. In .- . Empereur ed. , Commerce et artisanat dans l le andrie hell nisti ue et ro maine. ctes du collo ue d th nes organis par le C R le Laboratoire de c ramologie de Lyon et l Ecole fran aise d th nes d cembre 55-64. Athens, Paris. El-Fakharani, F., 1974, The Lighthouse of Abusir in Egypt. Harvard Studies in Classical Philology 78: 257-272. El-Fakharani, F., 1983, Recent Excavation at Marea in Egypt. In G. Grimm, H. Heinen & E. Winter (eds.), egyptiaca re ernsia II as R misch y antin ische Ägypten. Akten des internationalen Symposions 26.- 30.Se ptember 1978i n Trier, 175-186. Mainz. El-Falaki, M.S., 1966, Ancient Alexandria. Alexandria: ), Dar Nashr Al-Thaquafa ( M.B. Astronome (transl.), 1872. Copenhague. El-Gindy, A., 1999, Meteorological and Hydrodynamic Conditions in the Eastern Harbour of Alexandria and its Vicinity. In . Halim ed. , Proceedings of the orkshop on the tatus of Pilot Pro ect for the ustain able Development of the Submarine Archaeological Sites at Q ayetbey Citadel and Eastern Harbour of le andria o ember Reports on ydro dynamics Geophysics Morphology and Geology, V. II: 1-82. Alexandria. El-Raey, M., Nasr, S., Frihy, O., Desouki, S., & Dewidar, Kh., 1995, Potential Impact of Accelerated Sea-Level Rise on Alexandria Governorate, Egypt. J ournal of Coastal Research 14: 190-204. El-Zouka, M.K., 1979, Irrigation reas in the estern Delta: A G eographical Study ( 8´ ), Alexandria (in Arabic). Empereur, .- ., & Picon, M., 1986, A La Recherche des ours d Amphores. In - . Empereur & . Garlan (eds.), Recherches Sur Les Amphores G recques. Actes

du colloque international organisé par le Centre national de la recherche scientifi ue l ni ersit de Rennes II et l cole fran aise d th nes th nes 10- 12Se ptembre 1984) : 103-126. Athens, Paris. Empereur, .- ., & Picon, M., 1998. Les Atelier d Amphores du Lac Mariout. In .- . Empereur ed. , Commerce et artisanat dans l le andrie hell nisti ue et romaine. ctes du collo ue d th nes organis par le C R le Laboratoire de c ramologie de Lyon et l Ecole fran aise d th nes d cembre 75-88. Athens, Paris. Empereur, .- ., & Picon, M., 1992, La Reconnaissance Maréotide. Extrait des Cahiers de la Céramique gyptien 3: 145-152. Empereur, .- ., 1998, Alexandria Rediscovered. New ork. Flaux, C., forthcoming, Geomorphological survey of the south-west shores of Lake Mareotis. In L. Blue & E. Khalil (eds.), A Multidisciplinary Approach to Alexandria s Economic Past the Mareotis Case tudy. Oxford. Frihy, O.E., Dewidar, Kh.M., & El-Raey, M.M., 1996, Evaluation of Coastal Problems at Alexandria, Egypt. O cean & Coastal Management 30.2-3: 281-295. Gabel, C., & Petruso, K., 1980, An Environmental and Cultural Study at Lake Maryut, Lower Egypt. Boston ni ersity frican tudies Center orking Papers No. 25. Boston. Petruso, K., & Gabel, C., 1982, Marea: A Byzantine Port in Northern Egypt. oston ni ersity frican tudies Center orking Papers No. 62. Boston. Hassan, .A., 1997, The Dynamics of Riverine Civilization: A Geoarchaeological Perspective on the Nile Valley, Egypt. orld rchaeology 29.1: 51-74. Heliodorus. Aethiopica, T. Underdowne (transl.), 1895. London. Herodotus. The Histories, A. de Sélincourt (transl.), 1971. Middlesex. Horden, P., & Purcell, N., 2002, The Corrupting Sea. Oxford. Kadous, E., 2001, Monuments of Ancient Alexandria. ( ) Alexandria (in Arabic). Kucharczyk, R., 2002, Marea 2001: Windowpanes and other Glass inds. In M. Gawlikowski & W. A. Daszewski (eds.), Polish Archaeology in the Mediterranean XIII: 65-71. Lewis, N., 1983, Life in Egypt under Roman Rule. Oxford. McGovern, P.E., 2003, ncient ine he earch for the O rigins of Viniculture. Princeton. Palladius, The Lausiac History. The Monks of Nitria. VII. I. http://www.fordham.edu/halsall/basis/palladiuslausiac.html (accessed February 2010). Peacock, D., 1992, Rome in the Desert: A Symbol of Power. An Inaugural Lecture delivered at the University of Southampton. Peacock, D., 2002, The Roman Period. In I. Shaw (ed.), The O xford History of Ancient Egypt: 422-445. Oxford. Pliny the Elder, The Natural History, J. Bostock & B.A. Riley (transl.), 1855. London. 144

E. KHALIL: MAREOTIS’ WATERFRONT le Laboratoire de c ramologie de Lyon et l Ecole fran aise d th nes d cembre 27-36. Athens, Paris. Rodziewicz, M.D., 2003, Philoxenit , Pilgrimage Harbour of Abu Mina. Bulletin de la Société Archéologi ue d le andrie 47: 27-47. Said, R., 1990, Quaternary. In R. Said (ed.), The G eology of Egypt: 487-507. Rotterdam. Strabo, G eography, H.L. Jones (transl.), 2001. London. Szymańska, H., & Babraj, K., 2008, Marea V.1 : Byzantine Marea – Excavations in 20203 and 206 . Krakow. Vör ös , G., 2001, Taposiris Magna: Port of Isis. Budapest. Warne, A.G., & Stanley, D.J., 1993, Late Quaternary Evolution of the Northwest Nile Delta and Adjacent Coast in the Alexandria Region, Egypt. J ournal of Coastal Research 9.1: 26-64. Wilson, P., 2007, estern elta Regional ur ey. http://www.dur.ac.uk/penelope.wilson/Delta/AbuMatamir.html (accessed February 2010).

Procopius, The Buildings, H.B. Dewing (transl.), 1940. London. Rickman, G., 1971, Roman G ranaries and Store Buildings. Cambridge. Rickman, G., 1980, The Corn Supply of Ancient Rome. Oxford. Rodziewicz, M.D., 1990, Taenia and Mareotis: Archaeological Research West of Alexandria. Acta of the First International Colloquium of the Egyptian Society of G reek and Roman Studies 1: 62-80. Cairo. Rodziewicz, M.D., 1998a, rom Alexandria to the West by Land and by Waterways. In .- . Empereur ed. , Commerce et artisanat dans l le andrie hell n isti ue et romaine. ctes du collo ue d th nes organis par le C R le Laboratoire de c ram ologie de Lyon et l Ecole fran aise d th nes 11- 12dé cembre 198: 93-103. Athens, Paris. Rodziewicz, M.D., 1998b, Classification of Wineries from Mareotis. In .- . Empereur ed. , Commerce et artisanat dans l le andrie hell nisti ue et romaine. ctes du collo ue d th nes organis par le C R

145

LAKE MAREOTIS: RECONSTRUCTING THE PAST

146

S. RAMSES & A. OMAR: OUTRAGE AND DESTRUCTION

Lake Mareotis Research Project. Phases of Outrage and Destruction Sameh Ramses & Ahmed Omar

Interest and Aim of the Project The historical importance of the region of Mareotis during the Ptolemaic and Roman periods, either as an important wine production center, in support of other Egyptian industries (Strabo 17.1.7), or as a resort of the elite (Buttler 1989: 8-12), has attracted a great deal of archaeological attention, as this volume illustrates (El-Fakharani 1983; Empereur & Picon 1998; Rodziewicz 1998). Nevertheless, the region of Mareotis is threatened by modern urban development and many historical sites are subjected to destruction, not only due to its unique location in the vicinity of Alexandria, but also its moderate climate which has established the area as one of the most popular resorts in the region (Abd-Elhady 2008: 177-184). Therefore, the aim of the Lake Mareotis Research Project, conducted between the summer of 2004 until the summer of 2008, was to survey, record and map the platforms, harbours, and any other waterfront constructions that could be found along the northern and southern shores of the western part of Lake Mareotis, and to create a very detailed database for each site, as well as to determine its present state of preservation.

Phases of Outrage and Destruction According to the last survey of 2008 more than 70 archaeological sites have already been noted in the region (Fig. 2). Of course, not all of them had previously been registered, but some had been surveyed or documented.1 However, despite efforts by the Archaeological Department of Alexandria University to discover and protect some of these sites, the full extent of the area, and the variety of remains, both on land and in the water, still remains to be determined and much work needs to be done. This is particularly urgent considering the very immediate threats of modern urban development and various industrial and agricultural activities that are threatening the shores of the lake. Consequently, during the course of the survey, three main factors were identified that are believed to present a particular danger to the archaeology of the region. Fish Farms As a consequence of being inundated by water largely of agriculture runoff and drainage water, the lake water quality has deteriorated making it more difficult for a variety of fish to survive. Simultaneously, many fish-farmers began to appropriate quite large portions from the lake basin and along the shores. In the process of creating fish producing basins, a great deal of dredging and destruction of some of the archaeological sites, both offshore and under water, has taken place. Accordingly, the original shores topography has been deformed, and the dredging leads to

The survey project covers an area of about 40 km long and up to 3 km wide within the western arm of Lake Mareotis, to the west of Borg El-Arab airport, on the Sidi Kerir road, and extends towards El-Hammam city, it also includes Mareotis Island some 3.7 km long and 680 m wide (Fig. 1; see Blue, Hopkinson and Khalil this volume). Fig .1: G eneral plan from Alexandria to Mareotis ( Lake Mareotis Research Proj ect) .

1. It is noteworthy that only about nine sites had already been registered by the Supreme Council of Antiquities (SCA), and only six are considered to be SCA properties. Atlas Almoqea Alathariaya, SCA, 2002: No. 3: El-Beheira Governorate; No. 4: The Rest of the Lower Egypt Governorates. Egypt.

147

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig.2: Map of sites investigated during the Lake Mareotis Research Proj ect ( Lake Mareotis Research Proj ect) . Industrial Activities Although not as polluted as some parts of Lake Mareotis, the western arm determines a substantial area of the current lake littoral, and is considerably affected by a huge amount of untreated industrial waste and polluted water which flows into the lake on a daily basis. The industrial activities around the shores of the lake not only affect the water quality but in some cases have a direct impact on the

mis-interpretations of its original form. These activities particularly characterise the southern shore of the lake (Fig. 3). It is worth mentioning that some farms have appropriated quite large areas, up to nearly 4 hectares, especially in the vicinity of Mareotis Island, which embraces a great deal of archaeological remains, making access to the island increasingly difficult by cutting away the roads and destroying the routes that lead to the island. 148

S. RAMSES & A. OMAR: OUTRAGE AND DESTRUCTION

Fig.

o e amples of fish farm deformation along the southern shore of the lake Lake Mareotis Research Pro ect .

Fig.4: Lime storing and collecting at Site 4 ( Mareotis Research Proj ect) .

Lake

Fig.5: Lake wall affected at Site 4 ( Research Proj ect) .

archaeological sites located in the region. Site 44, located on the southwest shore of the lake, provides clear evidence for this negligence. It is directly adjacent to a cement factory, and is used to store the lime for the cement (Fig. 4). The cement mound is huge and it no doubt, amongst other things, covers the western end of a lake wall that extends along the lake shore (Fig. 5), and disregards the existence of additional archaeological features including a circular structure at the far west of the site near the water’s edge, and four square probable wine-basins, that were previously excavated by the SCA. Similar scenes are witnessed further west, for example at the site of Naga El-Mawalik (Site 109). This is an example of an endangered site that still displays aspects of its ancient harbour and enclosure walls, together with many other hills and mounds of pottery. Its location close to the cement factory, means that it is being used to supply the factory with limestone. Parts of this site are now totally inaccessible, with unrecorded archaeological features being surrounded by barbed wire.

Lake Mareotis

On the north shore of the lake, at the western extent of the survey area, Site 214, immediately opposite Naga El-Mawalik, has recently been purchased by the El-Nasr Salt Company. They have acquired two already excavated sites (Sites 214-215) which contain two wineries and a big house of the late Roman period and numerous mounds of pottery (Figs. 6-7). These sites are due to be dredged and leveled to be used as evaporations basins to produce salt. They have only been partially recorded. There are many other places in the area free of archaeological remains that could be used for such purposes. Other sites, particularly those on the northern shore of the survey area, are being used as rubbish tips totally swamped by rubbish from the city and isolated from the main road by means of barbed wire. These sites are distinctive by their bad smell. U rban Development Urban development is considered to be one of the most dangerous factors that affect the archaeological sites of the 149

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig.6: ( above) Site 214: the partially excavated winery ( Lake Mareotis Research Proj ect) .

Fig.7: ( right) Site 214: part of the fragmented paved oor of a Roman period aterfront building Lake Mareotis Research Proj ect) . region. Vast portions of land have been acquired, many divided into smaller units to be sold and used for housing. Both the local Bedouin and big companies that wield profit and power are responsible for destroying and leveling vast areas at a remarkably rapid rate. The Lake Mareotis Research Project has had direct experience of these actions over the last four years of survey. This can be summarised in the following four case studies:

documentation of the site, we returned to find that the three hills were totally destroyed by the locals and the Hellenistic tomb flattened ig. 12 . It is worth mentioning that this site had not been previously registered or recorded by the SCA. El-G amal ( Site 204) The situation at the site of El-Gamal was a little different as the site was quite impressive and important, to the extant that it had already been partially recorded by the SCA, and a site guard had been assigned. The destruction of this site was undertaken in stages first a soccer field was cleared to the east of the tell; the following year another one was cleared on the prohibited area of the archaeological site towards the limit of the lake shores, destroying part of the tell and site boundaries, and so the destruction continued (Fig. 13).

Site 09 A Kibotos harbour, the only well preserved example either on the sea, or on the lake. Nevertheless, the governorate has dedicated an area of about 8 hectares around the site to build a private university without leaving any access to the archaeological site (Figs. 8-9). Accordingly, over time, the site will be destroyed. Ez-Z eraa Elbahry ( Site 201) When visited in September 2004, there was a hill of about 5 m in height, which contained several mounds of pottery with traces of building plots everywhere. When we returned in May 2007, the 5 m high hill had disappeared and there were only private property signs and modern partitioning walls visible (Figs. 10-11).

Even those sites renowned for their archaeological importance are also under threat. To the southeast of Taposiris Magna, the extensive seawall and jetty associated with the site are slowly being encroached by small concrete building plot walls, purchased for urban development (see Boussac & El Amouri this volume).

Nagea Ez-Z ohorat ( Site 20) A similar pattern of destruction was noted at Nagea EzZohorat (Site 202) as was witnessed at Ez-Zeraa Elbahry (Site 201). The site comprised three hills containing amongst other remains, a cistern, a quay, and a rock-cut Hellenistic tomb. We revisited the site in August 2007 and just few days after we had commenced the recording and

Further destruction is caused at many sites by the introduction of agriculture, particularly the cultivation of figs and olives. Cattle are also allowed to graze freely around the whole region, including on established archaeological sites. Moreover, seasonal extraction of reeds in marshy lake side areas permits the access of heavy loaders and trucks to a number of the archaeological sites. 150

S. RAMSES & A. OMAR: OUTRAGE AND DESTRUCTION

Fig.8: K ibotos harbour extending from the southern shore of the lake ( Lake Mareotis Research Proj ect) .

Fig.9: O ne of the university properties on the land surrounding the K ibotos ( Lake Mareotis Research Proj ect) .

Fig.10: Archeological tell divided for private possession ( Lake Mareotis Research Proj ect) .

Fig.1 1: Private property sign and wall established on Site 201( Lake Mareotis Research Proj ect) .

Fig.12:

Flattened Hellenistic tomb at Site 20dur

ing bulldozer work in 207( 151

Lake Mareotis Research Proj ect) .

LAKE MAREOTIS: RECONSTRUCTING THE PAST

Fig.



o soccer fields erected at the tell boundaries around ite

C onclusion and Recommendation From the above mentioned, the slow destruction of archaeological sites in Mareotis can be specified as follows: Cultivating the boundaries of archaeological tell sites and in some cases, on the tells themselves. Setting up barbed wire around many of the sites to give the impression that they are private property, and when the authorities are notified, the signs are not removed. sing bulldozers and tractors to rapidly destroy and level the archaeological hills. Selling archaeological sites, especially the unregistered sites, to private individuals, by employing unorthodox means that appropriate the names of established people, to ensure that legally the sites cannot be retrieved, thus preventing anyone, even the governmental organizations, from taking any action.

Lake Mareotis Research Pro ect .

Generally, the aim of these suggestions would be to create a project which promotes a particular concept for both the short- and long-term management of the region, which considers its archaeological and historical importance. Such a plan should also aim to promote and preserve Mareotis for the coming generations. This is a great challenge that we all face. Bibliography Abd-Elhady, D., 2008, erasat fi sasyat l sayaha. Alexandria. Atlas Almoqea Alathariaya 3: El-Beheira Governorate, Egypt. Supreme Council of Antiquities, August 2002. Cairo. Atlas Almoqea Alathariaya 4: The Rest of Lower Egypt Governorates, Egypt. Supreme Council of Antiquities, August 2002. Cairo. Buttler, A.G., Fateh Elarab Lemesr, M.F. Abo Hadid & T. El-Massrein (transl.), 1989. Cairo. El-Fakharani, F., 1983, Recent Excavation at Marea in Egypt. In G. Grimm, H. Heinen & E. Winter (eds.), Aegyptiaca Trevernsia II, Das Röm isch-Byzantinische Ägypten. Akten des internationalen Symposions 26.- 30. September 1978 in Trier, 175-186. Mainz. Empereur, J.-Y., & Picon, M., 1998, Les Ateliers d’Amphores du Lac Mariout. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athè nes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole franç aise d’Athè nes, 11- 12 décembre 198: 75-88. Athens, Paris. Rodziewicz, M.D., 1998, Fom Alexandria to the West by Land and by Waterways. In J.-Y. Empereur (ed.), Commerce et artisanat dans l’Alexandrie hellénistique et romaine. Actes du colloque d’Athè nes, organisé par le CNRS, le Laboratoire de céramologie de Lyon et l’Ecole franç aise d’Athè nes, 11- 12 décembre 198: 93-103. Athens, Paris. Strabo, G eography, H.L. Jones (transl & ed.), 1996. London, New York.

Accordingly, it has been established that the problems of the region are extensive and complex, but once identified can be resolved by the means of establishing a compromise between protecting Mareotis’ archaeological sites, whilst at the same time meeting the increasing needs of development and modernization. Therefore, the following is suggested to overcome the problems: Coordination between the different regional authorities including those that operate in the interest of archaeology, investment, irrigation, and the governorate, to establish an integrated long term plan, which recognises a satisfactory compromise that identifies the needs of all parties. nifying the mapping system of each organization, in order to get acquainted with each others properties, in order to stop the illegal acquisition of land. Support the Supreme Council of Antiquities, both to provide the facilities needed to control and protect such a vast area, and to implement archaeological legislation. By encouraging local interest in the importance of the sites both as a source of income, and as an important aspect of cultural heritage. By promoting the whole region to become an open archeological park in order not to be neglected and misused, otherwise it will soon disappear. 152

S. RAMSES & A. OMAR: OUTRAGE AND DESTRUCTION

153

LAKE MAREOTIS: RECONSTRUCTING THE PAST

154

INDEX

Select Index Places/topography Abu Mina, 41, 54, 67-73, 77-78, 81, 115, 127-130, 137 Abuqir Lake, 13, 15-16, 18 Abusir, 15, 35 Abusir Ridge, 36, 43, 142 Alexandria, 1-4, 7-8, 12-13, 15-16, 18-20, 25-28, 32-33, 35-36, 47, 49, 54-55, 57, 66-67, 69, 71-72, 75-78, 8283, 87, 89, 100, 103, 107-108, 110, 113, 115-116, 123125, 127-128, 135, 137, 141-143, 147 Beheira, 37, 107-115, 119, 122, 124-125, 147 Cairo, 13, 15, 67, 107, 123, 125, 128 Canopic Branch/Canopic Nile, 4, 7, 13, 18, 25, 36, 75, 107-108, 110, 116, 119, 122, 124-125, 135, 142-143 Chaereou, 7 Fayum, 44, 110, 127 Huwaryia/Hauwariya, 68-70, 75, 78 Kanubiye Canal, 108, 114 Kom el Giza (see also Beheira), 107-115 Lake Mareotis/Lake Mariout/Maryut Lake, 1-4, 7-8, 1119, 21-23, 25-28, 30-33, 35, 37, 44, 47-48, 57, 66-67, 72-73, 75-77, 83, 87, 95-96, 119, 121, 123-124, 128129, 135-137, 139-143, 147-152 Libya, 1, 3-4, 8, 75-77, 157 Marea/Mareia (see also Philoxenité), 1-4, 7-8, 14-15, 17, 27-28, 30-32, 35, 37, 44, 47-52, 54-55, 57-61, 66-67, 69-73, 75, 77-83, 91, 100, 128-129, 135-139, 141 Mareotic Arm, 32-33, 135-136, 138, 141-143 Mareotis Island, 28-32, 35, 37, 41-42, 44, 135-136, 139142, 147-148 Maryut Depression, 25, 35, 142 Mediterranean Sea, 11-12, 15-17, 19, 25, 27, 127-128 Naukratis Canal, 123, 125 Nile Delta, 36-37, 44, 83, 107, 110, 135, 137 Nile River, 3-4, 7-8, 11-13, 15-16, 18, 25, 27, 35-36, 39, 44, 50, 54, 66, 75, 83, 95-96, 103, 107-108, 110, 116, 125, 127, 135, 142-143 Nile Valley, 3, 27, 35, 41, 44, 54, 103, 127 Nitria/Nitraria, 8, 13-15, 143 Philoxenité/Philoxenite, 27, 35, 47, 54-55, 67, 69, 71-73, 75, 77-78, 83, 135-137, 139 Plinthine, 4, 14, 67, 87-88 Sais, 119, 124 Schedia, 7, 13, 107-110, 114-116, 125, 129, 142 Schedia Canal, 18, 142-143 Taposiris Magna/Taphosiris, 7-8, 27, 67, 71, 87-88, 92, 101, 136-137, 139, 141, 150 Western Desert, 3, 5, 123-124

El-Fakharani, F., 1-2, 27, 31, 37, 47, 49-53, 69-70, 76, 78, 129 El-Falaki, M., 1, 17-18, 37, 47, 57, 67, 75, 137 Herodotus, 3, 7, 54, 73, 75, 137 Justinian, 7-8, 76 Pliny, 4 Procopius, 7-8, 55 Sophronius, 8 St Menas, 51-52, 54, 78 Strabo, 3-4, 7-8, 18, 35-36, 39, 47, 54, 66, 75, 127, 142-143 Theodosius, 8 General amphora/amphorae, 27-29, 32, 37, 43-44, 58, 62, 72-73, 75, 78, 81-83, 87, 90, 100-103, 113, 115-116, 121, 124, 127, 135-136, 141-142 auger cores/auger survey, 30, 37-38 basilica, 51-53, 75-78, 81-83, 129 bath complex/bath facility, 28, 58, 70-72, 75, 110, 116, 128-129 bridge, 42, 89-92, 100, 107-108, 116, 137 camel, 4-5, 55 canal, 3-4, 7-8, 14-15, 18-19, 21, 27, 35-36, 50, 52, 66, 75, 91, 95-96, 107-108, 110, 114, 116, 121, 123-124, 135, 142-143 causeway, 25, 58, 60, 71, 77, 87, 89-91, 93, 95, 100, 135 cistern, 2, 28, 36-37, 42, 53, 58, 69-70, 72, 80, 136, 142, 150 coins, 51-53, 70, 79, 82, 91, 96, 100-101, 103, 107, 110, 113-115, 128 Coptic Encomium of St Menas, 54-55, 69, 72, 77 evaporation, 37, 149 finewares, 28, 121 fish farms, 19-22, 96, 100, 147, 149 fish-tank, 96-100, 103 floods/seasonal flooding, 4, 16-17, 22, 35-37, 39, 43, 54, 87, 90-91, 95, 124, 137, 143 forts, 49, 124 glass, 5, 25, 27, 52, 66, 75, 121, 142 harbours, 4, 7, 25, 27, 31, 38-39, 47-51, 53-55, 57, 66-73, 75-78, 81, 83, 87, 90-91, 93, 95-96, 98, 100-101, 103, 107-108, 110, 116, 124, 136-139, 142-143, 147, 149151 hypogea/ hypogeum, 54, 69, 71 jetty/jetties, 28-29, 32, 37, 39, 41-43, 47, 49, 53-53, 58, 60, 67, 71-72, 75, 77-78, 81, 83, 89, 91, 93-98, 100-101, 103, 136, 138-139, 142, 150 karum, 36, 43, 71 Kibotos/Kibotos harbour, 31, 47-48, 71, 75, 91, 139, 150151 kiln, 31-32, 36-37, 43-44, 70-73, 75, 77-78, 82-83, 127, 130, 136 lake port, 67, 70, 72, 75 lake wall, 30, 32, 39, 43, 139-140, 149

People Athanasios, 77 Christodorus, 8 Claudius Ptolemy/Ptolemy the Geographer, 4, 36, 47, 54 Coste, P., 87, 89-91 De Cosson, A., 1, 35, 67, 75, 87, 89

155

LAKE MAREOTIS: RECONSTRUCTING THE PAST lighthouse, 49, 58, 137 mole, 28, 31, 41, 67, 139 natron, 3, 5, 15 papyri/papyrological documents, 4-5, 8, 96, 110, 130-132 papyrus, 7, 54, 75, 142 Pharos, 3, 137 pilgrim, 54-55, 69, 71-73, 77-78, 82, 137 pilgrimage, 66, 68-72, 78 pilgrimage centre, 67, 71-72 pollution, 27-28 port, 14, 25, 27, 42, 47, 49-52, 66-67, 69-73, 75, 78, 83, 96, 123-124, 135 qadus, 28 quarry, 27, 32, 36, 43, 67, 71, 77, 91, 107, 116, 129, 143

quay, 28-29, 32, 39, 41, 43, 47, 49-50, 66-67, 72, 77, 136140, 142, 150 sakkia sakia sa iya s iyah, 28, 30, 32, 37, 41, 76, 79-81, 83, 136, 142 sebakh, 107, 110, 123 sewage, 25, 27, 77, 83 tombs, 2, 17, 32, 49, 54, 69, 71, 111, 116, 150-151 villa, 35, 43, 58, 77-78, 111, 113-114, 116, 124 vineyards, 5, 43, 75, 78, 127, 130-132 wine, 4-5, 25, 27, 43, 53-54, 70, 75, 77-78, 83, 101, 103, 121, 127-132, 142, 149 wine production, 30, 32, 70, 114, 127, 130, 132, 135-136, 142, 147 winery, 32, 43-44, 53, 69-70, 123, 127-132, 135, 149-150

156