Mochlos III: The Late Hellenistic Settlement: The Beam-Press Complex (Prehistory Monographs) 9781931534789, 1931534780

Citation preview

MOCHLOS III The Late Hellenistic Settlement The Beam-Press Complex

Mochlos III The Late Hellenistic Settlement The Beam-Press Complex

Transport amphora III.67. Watercolor D. Faulmann.

PREHISTORY MONOGRAPHS 48

Mochlos III The Late Hellenistic Settlement The Beam-Press Complex

by Natalia Vogeikoff-Brogan

contributions by Marie-Claude Boileau, Tristan Carter, Amanda Kelly, Andrew Koh, Evi Margaritis, Dimitra Mylona, Eleni Nodarou, Maria Ntinou, David S. Reese, and Ian Whitbread

edited by Jeffrey S. Soles and Costis Davaras

Published by INSTAP Academic Press Philadelphia, Pennsylvania 2014

Design and Production INSTAP Academic Press, Philadelphia, PA

Library of Congress Cataloging-in-Publication Data Mochlos III : the Late Hellenistic settlement : the beam-press complex / by Natalia Vogeikoff-Brogan ; contributions by Marie-Claude Boileau, Tristan Carter, Amanda Kelly, Andrew Koh, Evi Margaritis, Dimitra Mylona, Eleni Nodarou, Maria Ntinou, David S. Reese, and Ian Whitbread ; edited by Jeffrey S. Soles and Costis Davaras. pages cm. — (Prehistory monographs ; 48) Includes bibliographical references and index. ISBN 978-1-931534-78-9 1. Mochlos Plain (Greece)—Antiquities. 2. Excavations (Archaeology)—Greece—Mochlos Plain. 3. Material culture—Greece—Mochlos Plain. I. Soles, Jeffrey S., 1942– II. Davaras, Kostes. III. Title. IV. Title: Mochlos 3.  DF221.C8V64 2014 939’.18--dc23 2014022232

Copyright © 2014 INSTAP Academic Press Philadelphia, Pennsylvania All rights reserved Printed in the United States of America

Table of Contents

LIST OF TABLES.................................................................................................. vii LIST OF FIGURES........................ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . ix LIST OF PLATES... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii PREFACE Jeffrey S. Soles and Costis Davaras... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv ACKNOWLEDGMENTS Natalia Vogeikoff-Brogan......................................................................... xvii ABBREVIATIONS.. . ............................................................................................. xix INTRODUCTION Natalia Vogeikoff-Brogan............................................................................ 1 1. ARCHITECTURE, STRATIGRAPHY, AND HOUSEHOLD ANALYSIS Natalia Vogeikoff-Brogan with contributions by Amanda Kelly, Evi Margaritis, Dimitra Mylona, Maria Ntinou, and David S. Reese.................................................................. 5

vi

MOCHLOS III: THE LATE HELLENISTIC SETTLEMENT

2. POTTERY Natalia Vogeikoff-Brogan... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 3. STONE IMPLEMENTS Tristan Carter....................................................................................... 49 4. CERAMIC, GLASS, METAL, AND SHELL OBJECTS Natalia Vogeikoff-Brogan with contributions by Amanda Kelly and David S. Reese... ..................................... 61 5. THE LATE HELLENISTIC SETTLEMENT AT MOCHLOS AND THE POLITICAL AND ECONOMIC SOVEREIGNTY OF HIERAPYTNA Natalia Vogeikoff-Brogan... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 APPENDIX A. Petrographic Analysis of Local and Imported Transport Amphorae from Knossos, Mochlos, and Myrtos Pyrgos Marie-Claude Boileau and Ian Whitbread....................................................... 79 APPENDIX B. Petrographic Analysis of the Hellenistic Cooking Ware Eleni Nodarou.. . ..................................................................................... 103 APPENDIX C. Archaeochemical Analysis of Two Amphorae and a Cooking Vessel Andrew J. Koh.......................................................................................109 APPENDIX D. The Animal Bones Dimitra Mylona.................................................................................... 113 APPENDIX E. Marine Invertebrates and Land Snails David S. Reese.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117 APPENDIX F. The Olive Remains Evi Margaritis... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121 REFERENCES... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123 CONCORDANCE A. Field Numbers and Catalog Numbers for Mochlos III... . . . . . . . . . . . . . . . . . . . . . . . . . . . 133 CONCORDANCE B. Contexts and Catalog Numbers for Mochlos III... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137 INDEX... ........................................................................................................... 139 TABLES FIGURES PLATES

List of Tables

1.

Summary of the charcoal remains from the Beam-Press Complex.

2.

Concordance of selected sherds for petrographic analysis from Knossos, Mochlos, and Myrtos Pyrgos.

3.

Summary of the identified fabrics from Knossos, Mochlos, and Myrtos Pyrgos.

4.

Final results from the petrographic analysis of local and imported transport amphorae from Knossos, Mochlos, and Myrtos Pyrgos.

5.

Petrographic analysis of selected Hellenistic cooking ware from the Beam-Press Complex.

6.

Preservation of the animal bones, based on NISP.

7.

Taxonomic preservation of the animal bones.

List of Figures

1.

Hellenistic sites on Crete.

2.

The area of the isthmus of Ierapetra.

3.

The Late Hellenistic settlement at Mochlos.

4.

The Beam-Press Complex in relation to the LM III settlement.

5.

The Beam-Press Complex in relation to the LM IB and LM III settlements.

6.

State plan of the Beam-Press Complex.

7.

Exterior reconstruction of the Beam-Press Complex.

8.

Architectural section A–A'.

9.

Architectural section B–B', part 1.

10. Architectural section B–B', part 2. 11. Floor assemblages of Rooms 1 and 2. 12. Floor assemblages of Rooms 4 and 5. 13. Room 4: stratigraphic section D–D'.

x

MOCHLOS III: THE LATE HELLENISTIC SETTLEMENT

14. Floor assemblages of Rooms 6, 7, and 8. Small finds are not to scale nor placed according to exact findspot. 15. Architectural section C–C' through Room 6. 16. Room 6: various views of the beam press’s construction. 17. Stratigraphic section E–E' outside Rooms 1, 2, and 3. 18. Room 6: (a) reconstructed interior view, looking southeast; (b) detail of beam press. 19. Echinus bowls with one handle (III.1–III.17). Scale 1:3. 20. Plates with outturned rim (III.18, III.19); plates with beveled or projecting rim (III.20–III.27); lebes(?) (III.28); kantharoi (III.29, III.30). Scale 1:3. 21. Kantharoi (III.31–III.36); moldmade bowls (III.37–III.48); and beakers (III.49–III.51). Scale 1:3. 22. Jugs (III.52–III.57); scale 1:4. Lagynos (III.58) and lekythoi (III.59–III.63); scale 1:3. 23. Filter jugs (III.64, III.65); situla (III.66); transport amphorae from East Crete (III.67–III.69). Scale 1:3 unless otherwise indicated. 24. Transport amphorae from East Crete (III.71–III.76). Scale 1:5. 25. Transport amphorae from Kos (III.77–III.81). Scale 1:5. 26. Transport amphorae from Kos (III.83–III.87). Scale 1:5. 27. Transport amphorae from Kos (III.88, III.89), Rhodes (III.90, III.91), and unattributed (III.92). Scale 1:5. 28. Lekanai (III.93–III.96). Scale 1:5. 29. Chytrai with collar rim (III.98–III.102, III.104–III.106). Scale 1:4. 30. Chytrai with inverted and grooved rim (III.107–III.110); chytrai with everted rim (III.111–III.114). Scale 1:3. 31. Lopades (III.115–III.119); lids (III.120–III.123, III.125, III.126); lamps (III.127, III.128). Scale 1:3. 32. Hammerstones (III.131, III.132); pestles (III.135, III.136); handstone (III.137); grinder (III.138). Scale 1:4. 33. Saddle querns (III.140–III.142). Scale 1:4. 34. Hopper-rubber mills (III.143, III.144). Scale 1:4. 35. Press bed (III.145); whetstone (III.146); stone lamp/mortar (III.147). Scale 1:4 unless otherwise indicated. 36. Biconically perforated weight (III.152); ring-shaped weight (III.153); naturally perforated weights (III.155, III.158); counter/weight(?) (III.159). Scale 1:4 unless otherwise indicated. 37. Roof tiles (III.160–III.164). Scale 1:5 unless otherwise indicated.

LIST OF FIGURES

xi

38. Roof tiles (III.167, III.168); drain pipe (III.169). Scale 1:5. 39. Clay loomweights (III.170–III.183); spindle whorl (III.184). Scale 1:3. 40. Figurine (III.185); plastic vase (III.186); glass vessels (III.187, III.188, III.190, III.191). Scale 1:3 unless otherwise indicated. 41. Lead objects (III.193–III.205). Scale 1:3. 42. Copper alloy (III.206–III.211); iron objects (III.212, III.213). Scale 1:3. 43. Total ion chromatogram of transport amphora III.74. 44. Total ion chromatogram of transport amphora III.79. 45. Total ion chromatogram of transport amphora III.100.

List of Plates

Frontispiece. Transport amphora III.67. 1.

Aerial photo of Mochlos showing the summit of the island from the north.

2A. Beam-Press Complex, Room 1 from the east. 2B. Beam-Press Complex, Room 6 from the west. 3A. Beam-Press Complex, Room 6, press bed from the west. 3B. Beam-Press Complex, Room 6, press bed from the north. 4.

Echinus bowls with or without a triangular handle (III.1, III.6, III.11); plate with projecting or beveled rim (III.24); high-necked cup (III.36); jugs (III.52, III.53).

5.

Filter jug (III.64); situla (III.66); transport amphorae (III.67, III.69).

6.

Transport amphorae (III.79, III.92); lekane (III.93); chytrai with collar rim (III.98, III.100).

7.

Chytrai with collar rim (III.106); chytrai with everted rim (III.114); lid (III.122); lamp (III.127); ground stone Type 1a, hammerstone (III.129); ground stone Type 2, implements with pecked circumferences and one or two abraded faces (III.130–III.132).

8.

Ground stone tools: Type 2, implements with pecked circumferences and one or two abraded faces (III.133, III.134); Type 3, pestles (III.135, III.136); Type 10, differentially weathered cobble (III.139); Type 14a, saddle querns (III.140–III.142).

xiv

9.

MOCHLOS III: THE LATE HELLENISTIC SETTLEMENT  

Ground stone Type 14d, hopper-rubber mills (III.143, III.144).

10. Ground stone Type 14e, press bed (III.145). 11. Ground stone tools: Type 16, whetstones (III.146); Type 19, stone lamp/mortar (III.147); Type 21, balance-pan weight (III.148); Type 22, biconically perforated weights (III.149–III.152); Type 23, ringshaped weight (III.153). 12. Ground stone Type 25, naturally perforated weights (III.154–III.158); ground stone Type 28, miscellaneous (III.159); roof tiles (III.160, III.166). 13. Loomweights (III.170–III.172, III.175, III.180–III.182); spindle whorl (III.184); figurine (III.185). 14. Plastic vase (III.186); Charonia (III.192); lead weight (III.194); iron object (III.215). 15. Thin section analysis: (a) amphibolite end member, III.75 (NV 4), XPL, width of field 4.4 mm; (b) detail of the red streaks in the matrix of III.75; (c) weathered volcanic end member, sample JE 36, XPL, width of field 7.04 mm; (d) sand-tempered fabric sample JE 16, XPL, width of field 4.4 mm; (e) fossiliferous clay fabric sample JE 25, XPL, width of field 4.4 mm; (f) fine micaceous fabric sample JE 10, XPL, width of field 4.4 mm. 16. Thin section analysis: (a) coarse sandy micaceous fabric III.88 (NV 25), XPL, width of field 4.4 mm; (b) volcanic end member, sample JE 31, PPL, width of field 4.4 mm; (c) phyllite end member, III.80 (NV 15), PPL, width of field 4.4 mm; (d) serpentinite-calcareous fabric sample NV 23, PPL, width of field 4.4 mm; (e) fine-grained volcanic rock fabric sample NV 19, XPL, width of field 4.4 mm; (f) chert-serpentinite fabric sample JE 18, XPL, width of field 4.4 mm. 17. Thin section analysis: (a) calcareous with microfossils fabric sample NV 34, XPL, width of field 4.4 mm; (b) fine silicate fabric, III.92 (NV 5), XPL, width of field 4.4 mm; (c) fine fabric sample JE 17, XPL, width of field 4.4 mm; (d) calcareous fabric sample JE 41, XPL, width of field 4.4 mm; (e) coarse silicate fabric, III.79 (NV 12), XPL, width of field 4.4 mm; (f) fine sandy micaceous fabric sample JE 14, XPL, width of field 4.4 mm. 18A. Thin section analysis of Hellenistic cooking wares: (a) subgroup a (x50), (b) subgroup b (x50), (c) subgroup c (x50), (d) subgroup d (x50). 18B. Complete and charred crushed olives found in Room 6.

Preface

This volume publishes the first of several Late Hellenistic buildings that were uncovered on the island of Mochlos during the Greek-American excavations of 1989–1994, 2005–2006, 2009–2010, and 2012. It also provides an introduction to the Hellenistic settlement that flourished on the island for nearly a century before it was abandoned. The Hellenistic remains were by no means the only remains uncovered in the course of the excavations that dated after the Bronze Age, but they were certainly the most extensive. The remains were located at or near the surface, on the top of the island, and along its south slope where they overlay much of the Late Minoan III and Neopalatial settlement remains. They also belonged to the longest lasting phase of the later occupation, the entirety of which was relatively short in comparison to nearly 1,800 years of occupation during the Bronze Age. When the project began in 1989, many Hellenistic walls were visible at ground level, and some are still visible today. Richard Seager excavating in 1908 also encountered these walls and dated them correctly to “Late Greek and Roman times” (Seager 1909, 275). Eager to find the Minoan town that lay beneath, he did not treat these walls kindly and removed many without making a record of what he found. As a result, our picture of Hellenistic Mochlos is not as complete as it might have been, but it is still remarkably well preserved. The current Greek-American project has used the same methodology in the excavation of Hellenistic levels as in the excavation of prehistoric remains on the site, preserving a careful record of the architecture, stratigraphy, and contexts of all finds, although it has not always been possible to preserve all the architecture because of the need to excavate the Bronze Age levels beneath. As a result, Natalia VogeikoffBrogan, who has been entrusted with the publication of the Hellenistic settlement, has been able to include a great deal of paleoenvironmental material—material that is often

xvi

MOCHLOS III: THE LATE HELLENISTIC SETTLEMENT     

neglected in excavations dealing with historical remains—in her discussion. She has been able to document the way different rooms in the Beam-Press Complex were used and the way the occupants of the building ate and worked. She has also taken advantage of scientific approaches that are commonly used in the examination of prehistoric material, in particular the use of ceramic petrography in pottery analysis. This analysis has allowed her to make one of her most important discoveries, the identification of a new class of pottery that she has called East Cretan Cream Ware (ECCW), a classification that allows her to attribute fine wares and other pottery to a source near ancient Hierapytna (modern Ierapetra). As a result, she has also been able to draw important conclusions about Hierapytna and its territory in the late Hellenistic period, including its commercial specialization and trade in wine. She has been able both to place the Beam-Press Complex and the site itself in its wider geopolitical context and make a number of discoveries about the history of East Crete in an important transitional period when it lost its independence and became part of the wider Mediterranean world. Unlike earlier volumes in the Mochlos publication series, which have published multiple sites and buildings in each volume and have been divided into different books in order to cover the large amount of material from the different sites, this volume publishes only one building in its entirety. The Beam-Press Complex has been chosen as the first topic in the series dealing with Hellenistic Mochlos because it was the first to have been completely excavated and was also one of the best-preserved contexts from this period. This complex also produced the most numerous and the most informative finds from any Hellenistic building excavated to date at Mochlos. The book is divided into five chapters and accompanied by six appendices. Chapter 1 provides a room-by-room description of the building, including a description of its stratigraphy and finds, and a discussion of the way the room was probably used. As in earlier Mochlos books, the presentation of each room ends with a list of artifacts and ecofacts, which are discussed in detail in subsequent chapters and appendices. Chapter 2 describes the pottery, Chapter 3 the stone implements, and Chapter 4 the ceramic, glass, metal, and shell objects. Chapter 5, the concluding chapter of the book, expands on the subject of the role that Mochlos played in East Crete and its relations with Hierapytna during the Late Hellenistic period. Jeffrey S. Soles Costis Davaras

Acknowledgments

The Beam-Press Complex at Mochlos was excavated in 1991–1992 by the University of North Carolina at Greensboro, in collaboration with the 24th Ephorate of Prehistoric and Classical Antiquities, under the auspices of the American School of Classical Studies at Athens. I am most grateful to the directors of the excavation, Jeffrey S. Soles and Costis Davaras, for trusting me with the publication of parts of the Late Hellenistic settlement at Mochlos, including the Beam-Press Complex. I am also grateful to the staff of the INSTAP Study Center for East Crete (INSTAPSCEC), especially its Director, Thomas M. Brogan, and the Assistant to the Director, Eleanor Huffman, for facilitating my work on a variety of levels—everything from providing tables for pottery reading to employing its publication team for the conservation, drawing, and photography of the finds. I cannot thank enough the artist-inresidence, Doug Faulmann, and the Chief Conservator, Stefania Chlouveraki, for their help. Other people who helped with the architectural plans and pottery profiles of the publication were Damon Cassiano, Gianluca Cantoro, Max Kalhammer, and Kostas Chalikias. Conservators Michel Roggenbucke and Cathy Hall were always willing to contribute to the conservation of the pottery. Earlier photography of the finds was undertaken by photographers Cathy May and Erietta Attali, but the final photographs were taken by the Study Center’s photographer, Chronis Papanikolopoulos. Michael Traister was responsible for the site photography. The trench masters, responsible for the excavation of the Beam-Press Complex, were, in alphabetical order: Tom Brogan, Bridget Crowell, Evi Sikla, Tom Strasser, Hara Thliveri, and Blake Woodruff. The easy retrieval of the finds was made possible through the diligent work of catalogers Mary Ellen Soles and Ann Nicgorski.

xviii

MOCHLOS III: THE LATE HELLENISTIC SETTLEMENT

My early discussions with Jonas Eiring and our subsequent collaboration in the petrographic analysis of the transport amphorae from Knossos, Myrtos Pyrgos, and Mochlos were extremely formative for this publication. I also wish to thank the former Director of the Fitch Laboratory, Ian Whitbread, for his willingness to undertake the petrographic analysis of the transport amphorae and assign the project to petrographer-archaeologist Maria-Claude Boileau. I would also like to thank Eleni Nodarou, the INSTAP-SCEC petrographer-in-residence, who undertook the thinsection analysis of the cooking ware, for sharing with me many of her thoughts about the origin of the Cretan fabrics. Finally, this publication would not have taken place without the moral support of the then Directors of the American School of Classical Studies at Athens, James D. Muhly (1997–2002), Stephen V. Tracy (2002–2007), and Jack L. Davis (2007–2012), who were always willing to grant me summer scholarly leaves. The publication of the Beam-Press Complex, as with previous publications in the Mochlos series, is a collaborative work. I have organized the publication and written four out of five chapters. Tristan Carter has written Chapter 3 on the stone implements. The contribution of Amanda Kelly on Chapters 1 and 4 deserves special mention; my discussion on the building’s roof tiles would not have been possible without her help since she collected and drew a large number of Hellenistic tiles from the surface of the site. Finally, Chapter 1, as well as the entire publication, benefitted largely from the input of Andrew J. Koh (organic residue analysis), Evi Margaritis (olive remains), Dimitra Mylona (animal bones), Maria Ntinou (wood remains), and David S. Reese (shells). Unless otherwise credited, all figures were drawn by Doug Faulmann, and all photos were taken by Chronis Papanikolopoulos. In writing the text of this volume, I benefitted immensely from the works of Penelope Allison (1999), Lisa Nevett (1999, 2010), and Bradley Ault (2005), who have been influenced in turn by the work of Martin Schiffer (1996). Another useful tool to my research has been the edited volume that Ruth Westgate, Nick Fisher, and James Whitley produced in 2007. I should also add to this list the experience I gained from my role as co-editor for the publication of ΣΤΕΓΑ: The Archaeology of Houses and Households in Ancient Crete, which addressed, for the first time, the issue of household archaeology on a regional level with a wide chronological range (Glowacki and Vogeikoff-Brogan 2011). Last but not least, I would like to thank the two anonymous reviewers whose excellent suggestions and useful comments have improved the quality of this publication. Natalia Vogeikoff-Brogan

Abbreviations

The following abbreviations are used in this volume: a angular

Fe

ACF

amorphous concentration features

g grams

C

clay object (not a vessel)

GS

ca.

circa

h. height

CA

copper alloy object

int.

interior

cf.

compares favorably with

KCF

crystalline concentration features

c:f:v

coarse:fine:void ratio

kg kilograms

d.

diameter

L. length

dim(s).

dimensions

L left

iron object ground stone object

E ethanol

LM

ECCW

East Cretan Cream Ware

m meters

EM

Early Minoan

M methanol

ESA

Eastern Sigillata A

max.

maximum

est.

estimated

ml

milliliters

ext.

exterior

mm

millimeters

MM

Middle Minoan

F fabric

Late Minoan

xx

MOCHLOS III: THE LATE HELLENISTIC SETTLEMENT

MNI

minimum number of individuals

r rounded

MOC

Mochlos

RT

roof tile

MS

Museum of Siteia

S

stone object

mw

molecular weight

sa subangular

NISP

number of identified specimens

Sh

no.

number

sr subrounded

P

pottery object (vessel)

TCF

textural concentration feature

Pb

lead object

TF

type of fabric

pers. comm.

personal communication

UM

Unexplored Mansion, Knossos

pers. obs.

personal observation

w.

width

pres.

preserved

wr

well-rounded

p.L.

preserved length

ws

water-sieved

PPL

plane-polarized light

XPL or XP

cross-polarized light

p.w.

preserved width

-/- complete/fragment(s)

shell

Introduction Natalia Vogeikoff-Brogan

History of the Mochlos Excavation The island of Mochlos lies at the east end of the Mirabello Bay in East Crete (Figs. 1, 2; Pl. 1). Until sometime in late antiquity, a narrow isthmus connected the small island to the mainland, thus offering to passing ships a shelter on both sides of the isthmus (Leatham and Hood 1958–1959, 273; Soles 2008, fig. 1). The land along the coast to the east of Mochlos consists of a narrow plain, which runs for a distance of ca. 4 km (Soles and Davaras 1992, fig. 1; Soles 2003, fig. 1). Although isolated from the interior of Crete by the steepness of the Ornos Mountains, the area was inhabited with few breaks from the Early Bronze Age until modern times. In 1908, Richard Seager was the first to explore archaeologically the site of Mochlos. After participating in the exploration of the Minoan town of Gournia in the early 1900s, Seager extended his research interests to the east of Gournia, opening trenches at the site of Vasiliki on the isthmus of Ierapetra (1905) and on the island of Pseira (1907). In the midst of his digging season at Pseira, Seager

sailed to Mochlos to explore the archaeological potential of the little island. He returned the following season to conduct a large-scale excavation in search of Minoan remains (Becker and Betancourt 1997, 78–79, 85–99). Seager makes only a brief mention of the Greek and Roman remains, an omission that reflects his lack of interest in these periods. In fact, he laments their existence because they “played sad havoc with the earlier remains . . . in places completely destroying them” (Seager 1909, 276). In his 1909 report, Seager refers sporadically to the Roman houses, which he swept away in order to unearth the Minoan remains, admitting that he could not date them with precision because the coins he found ranged in date from Hadrian to Constantine the Great (Seager 1909, 276). It is odd that Seager found several Roman coins in one season of excavation, while the Soles and Davaras team, over the course of 10 seasons of careful digging that employed both dry and water sieving, found fewer than a handful of coins, none of which can be identified as Roman.

2

NATALIA VOGEIKOFF-BROGAN

Of special interest, however, is Seager’s claim that a Greek/Roman town “had sprung up on the mainland about half a mile to the east, near where stands to-day the small church of Hagios Andreas,” a conclusion that he based on the large number of coins that came from that area (Seager 1909, 275). The site is preserved today, adjacent to a modern hotel now called the Club Aldiana, and a large scatter of sherds dating to the historical periods may still be seen here. Because of the proximity of the two sites, Mochlos and Hagios Andreas, it is very possible that the two were once parts of an extended settlement. In the 1950s, John Leatham and Sinclair Hood identified an underwater rock-cut construction on the mainland coast to the east of the modern village as a pair of fish tanks (Leatham and Hood 1958–1959). These fish tanks are now submerged due to the rise of the sea level in the eastern part of Crete, but they originally were located at the

coastline. Leatham and Hood dated the fish tanks to the Roman period on the basis of the testimonia of Latin authors such as Columella and Varro, who describe breeding or conserving fish in piscinae (Columella Rust. 8.16–17; Varro Rust. 3.17; Mylona 2008, 77–78). The new Greek-American excavations at Mochlos aimed at exploring the diachronic history of the settlement on the island and its adjacent coastal plain (Soles and Davaras 1992, 413–416). In addition to the rich Minoan and Mycenaean finds, the excavation revealed several structures of the late second to first century b.c. on the south slope of the island, including an eight-room complex with industrial features, the Beam-Press Complex (the focus of this book), as well as the remains of a long wall that most likely joins the curtain wall of the summit and formed a circuit wall that ran along the north, east, and south sides of the island.

Study of the Beam-Press Complex and Research Goals The present study examines the Beam-Press Complex, an eight-room building with industrial features, which is located on the south slope of the island outside the circuit wall in an area of the excavation known as E3 (Figs. 3–44; Pls. 1–18B). The building was first published as Hellenistic Building 1 in the Mochlos preliminary report of 1996 (Soles and Davaras 1996, 224–226). The complex is the first of the Hellenistic buildings to be published because it is one of the largest and best-preserved buildings in the settlement, with several floor deposits. Moreover, it is an unusual, non-residential building with a beam press and a second processing installation of uncertain character, perhaps used for crushing olives. The fact that very few beam presses from the Classical and Hellenistic period are preserved in situ is another reason for initiating the publication of the Hellenistic site with this building (Hadzi-Vallianou 2004, 78–86). Until recently, the Beam-Press Complex was thought to be the only building outside the circuit wall. In 2005 and 2010, however, excavation in the northeast of Area E2 revealed two rooms of a large building (the House of the Crocodile, named after a coin found there) that lay a short distance to

the west of the Beam-Press Complex (Fig. 3). The southern part of the building was excavated and removed by Seager, and the two rooms later exposed in 2005 and 2010 were located at its northeast corner, one used as a kitchen, and the other as a dining room. In 2009–2010, two rooms of another oblong building were excavated in Area D4 (Trenches 1100–1200), located inside the circuit wall and to the northeast of the Beam-Press Complex. Our current state plan shows an extensive Late Hellenistic settlement immediately behind the circuit wall, with the remains of several buildings all built on a linear plan and oriented along an east–west axis (Fig. 3). However, none is as well preserved as the Beam-Press Complex. When this study began nearly 15 years ago, there was no published archaeological material from East Crete for comparison, with the exception of finds from surveys in progress at Vrokastro, Kavousi, and Gournia (Hayden 2004; Haggis 2005; Watrous et al. 2012). All knowledge stemmed from historical studies and the analysis of ancient literary sources and inscriptions (van Effenterre 1948, 1991; Willets 1955; Spyridakis 1970, 1992; Chaniotis 1996, 1999). Unlike other studies that are initiated with

INTRODUCTION

certain questions in mind, our questions arose gradually as work on the Hellenistic pottery from Mochlos progressed. To begin with, the majority of the tableware and storage vessels from the Beam-Press Complex and other structures of the Late Hellenistic period from Mochlos did not find any parallels in the pottery deposits of Knossos and Eleutherna, which are the only two sites in Crete that have extensively published ceramic assemblages (Sackett et al. 1992; Kalpaxis, Furtwängler, and Schnapp 1994). At the beginning it was thought that the Mochlos tableware and storage vessels were produced locally at or near Mochlos. The first major breakthrough in the pottery study came in 1997 at Chania, at the Fifth Scientific Meeting for Hellenistic Pottery, when it became clear that similar tableware, storage, and transport pottery was found during the course of excavation at Myrtos Pyrgos, a site on the south coast of Crete (Eiring 2000; Vogeikoff 2000). A subsequent petrographic analysis, drawing samples from both sites, showed that the source of the most characteristic fabric at Mochlos was located on the south coast, near ancient Hierapytna (Vogeikoff-Brogan et al. 2004). The fabric was named East Cretan Cream Ware (ECCW). This new evidence raised a host of questions about the relationship of Mochlos—a small and marginal site on the north coast—to the city state of Hierapytna, as well about as the geographical extent of the latter’s political territory in East Crete in the Late Hellenistic period. What was the function of the Hellenistic settlement at Mochlos? If Hierapytnian workshops were responsible for the production of the ECCW, which had a wide distribution covering the north and south coasts of East Crete, was this material proof of Hierapytna’s economic and political expansion in East Crete after the middle of the second century b.c.? Since the vessels produced in ECCW included transport amphorae, another question concerned the production and trade of Cretan wine in the Hellenistic period. Most scholars have argued against any production of Cretan wine for trade before the establishment of the Roman rule on the island. Did Hierapytna’s political and economic growth in the second half of the second century b.c. elevate certain parts of her economy beyond subsistence production, like wine production, within its newly enlarged territory?

3

Finally, was the large beam press in Room 6 used for the production of olive oil, or wine, or both? Was it a small-scale operation addressing the personal needs of the local community, or could it have handled the production of surplus oil for export? Could the small community of Mochlos have afforded to invest time and labor in intensive olive cultivation? Chapter 1 examines the architecture, stratigraphy, and spatial distribution of the artifactual and organic remains of the Beam-Press Complex, taking into consideration the many processes that could have influenced the formation of the assemblage (e.g., whether it could have been part of primary or secondary refuse or if it had suffered from post-depositional processes like depletion). The aim of this detailed examination is to understand the function of each room as well as that of the entire structure. Methodologically, my approach follows recent advances in the field of household archaeology and the application of its main principles to classical archaeology (Schiffer 1996; Allison, ed., 1999; Nevett 1999, 2010; Ault 2005; Westgate, Fisher, and Whitley, eds., 2007; Glowacki and Vogeikoff-Brogan 2011; VogeikoffBrogan 2011). Chapter 2 is dedicated to the detailed presentation of the pottery found on the floor assemblages of the Beam-Press Complex and discusses issues of methodology, typology, morphology, and chronology. Most importantly, through thin section analysis, a new type of ware—the East Cretan Cream Ware—was identified and attributed to a source near Hierapytna. Appendix A provides the results of the petrographic analysis of the ECCW. Likewise, thin section analysis has identified the provenance of the cooking ware, the majority of which was made of the so-called Mirabello fabric. The data from the petrographic analysis are included in Appendix B. In Chapter 3, Tristan Carter examines the stone implements, artifacts that are rarely included in Classical publications, organizing them by function. His study shows that the majority date to the Bronze Age, an indication that the inhabitants of Hellenistic Mochlos reused them. Chapter 4 covers a range of other materials found in the BeamPress Complex, including roof tiles, loomweights, and metal objects. The discussion of the roof tiles is supplemented by a hypothetical representation

4

NATALIA VOGEIKOFF-BROGAN

of the complex’s roofing system. As with the stone implements, the study of the loomweights shows a certain degree of recycling since a large number of them date to the Bronze Age. After providing a comprehensive overview of the current state of knowledge concerning the history of Hierapytna in the Late Hellenistic period, the final chapter draws important conclusions about the political and economic sovereignty of the city in this period, using Mochlos as a case study. The archaeological evidence from Mochlos not only supplements written evidence that attests to Hierapytna’s expansion to the north, but it also provides, in some cases, new and more nuanced evidence about its economic growth and interests. Based on the consumption of large quantities of ECCW at the site, it is argued that the occupation of a strategic peninsula like Mochlos, located at the eastern end of the Mirabello Bay, should reflect the rise of Hierapytna to power in East Crete in the

second half of the second century b.c. Although Mochlos could have facilitated Hierapytna’s incoming and outgoing trade, the lack of any settlements along the isthmus and the Kavousi plains would have made the transshipment of goods from Mochlos to Hierapytna very difficult. The first and foremost role of Mochlos was as a coastal station— on behalf of Hierapytna—for collecting harborage fees from sea traffic along the north coast of Crete. The evidence from Mochlos also supports arguments for commercial specialization and trade in wine in East Crete in the late second and early first centuries b.c. The production of ECCW transport amphorae suggests that during the period of her largest expansion and growth, Hierapytna had moved beyond subsistence economy, trading wine for commercial profit. Along the same line, the existence of the press in Room 6 suggests the processing of oil on a scale considerably beyond that of a small-scale household economy.

Work to Follow It has been suggested in previous studies that the Mochlos settlement was a short-lived habitation, which was abandoned sometime in the course of the second quarter of the first century b.c. for unknown reasons—an abandonment perhaps connected with the conquest of Hierapytna by the Romans in 67 b.c. (Vogeikoff 2000; VogeikoffBrogan 2004; Vogeikoff-Brogan et al. 2004). This dating is now valid for the Beam-Press Complex, but not for the entire settlement. After several years of excavation, the general picture of the Late Hellenistic settlement that has emerged is one of buildings either largely depleted (i.e., their objects were removed either by their inhabitants at the time of their abandonment or were looted by scavengers after the abandonment) or with a small number of restorable objects remaining (on depletion processes, Schiffer 1985, 26–29). The time and nature of Mochlos’s abandonment in

the first century b.c. remains open; it is something that needs to be addressed in a publication that deals with the rest of the settlement. The excavation of the so-called House of the Crocodile in 2005 revealed the first building whose final abandonment can be dated with some certainty by numismatic evidence. The upper floor deposit in its kitchen yielded a bronze coin with a crocodile on the obverse and a ship’s prow with the letters CRAS written above on the reverse. It is unclear to which of Marcus Antonius’s generals it refers, P. Canidius Crassus or M. Licinius Crassus, but in either case the coin should be associated with Cleopatra and dated to the period 41–37 b.c., when Crete was ceded to Egypt as a wedding gift from Mark Antony to Cleopatra (Burnett, Amandry, and Ripollés 1992, 219; Brogan and Vogeikoff 2006, 426, fig. 658). The coin places the abandonment of the House of the Crocodile to a date after 37 b.c.

1

Architecture, Stratigraphy, and Household Analysis Natalia Vogeikoff-Brogan with contributions by Amanda Kelly, Evi Margaritis, Dimitra Mylona, Maria Ntinou, and David S. Reese

The Beam-Press Complex is located on the south slope of the island of Mochlos (Fig. 3). It was dug in two seasons, in 1991 and 1992, and uncovered by 14 trenches (Fig. 4: Area E3, Trenches 6200– 6700, 7200–7700, 8600–8700). Each trench, measuring 5 x 5 m2, was excavated leaving a 1 m wide balk on the northern and eastern sides of the trench. The balks were subsequently removed after they were drawn. Each feature encountered during the course of excavation was dug separately and given a locus number based on the trench numbering system. For example, Locus 7600 in Room 1 represents the unexcavated surface. The surface was dug in three passes (7601.1–3), uncovering three walls of Room 1. They were identified as Loci 7602 (west wall), 7603 (north wall), and 7604 (south wall). With the discovery of the walls, we changed the area between into Locus 7605, which represented the roof and wall collapse. Locus 7605.1 would refer to the first pass, 7605.2 to the second, and so on. All finds within any locus were identified by locus

number. Finds included in the publication also received an inventory number prefaced by the initial letter of the material that the object is made of or by its chemical abbreviation if it is metallic (e.g., P for pottery, C for ceramic objects, GS for ground stone, Pb for lead, Fe for iron); if cataloged, an object’s inventory number can be found in the catalog entry. The methodology of the excavation is explained further in previous volumes (Soles 2003, 8–9; 2008, 6). The Beam-Press Complex is made of two units, the East Unit and the West Unit (Fig. 6). The overall east–west length of the building, including both units, is ca. 23.50 m. The ground level slopes from north to south and from east to west, creating different floor levels (e.g., Rooms 1 and 2 of the East Unit sit at a higher level than Rooms 4 and 5 of the West Unit; Fig. 9). The East Unit, which measures ca. 10.90 x 7.10 m, is composed of three rectangular rooms with their doors opening to the east. Rooms 1 and 2 were built in a single construction

6

NATALIA VOGEIKOFF-BROGAN

phase. Room 3, which lies at a higher level than Rooms 1 and 2 (Fig. 8), abuts Room 1, but it could have been part of the same building phase. The entryways to the three rooms are off-centered. The West Unit is composed of five areas: a suite of three rooms (Rooms 4–6) and two auxiliary spaces (Rooms 7 and 8) at the western end of the unit. One enters the three-room suite from Room 6, which is a rectangular space with several builtin features, the most important being the beam press set against the east wall between the two entryways that lead to Rooms 4 and 5. The West Unit appears to have been built later than the East Unit since the north wall of Room 5 and the south wall of Room 4 abut the west wall of the East Unit. The study of the pottery and the depositional patterns have shown that the abandonment of Rooms 1 and 2 followed some kind of destruction that left several of the contents of the rooms in restorable condition. The rest of the rooms, except perhaps for Room 4, which follows the pattern of Rooms 1 and 2, went through a different depositional process: their high level of depletion shows that their abandonment was gradual. It is not clear, however, if the two depositional patterns are connected with different times of abandonment. If they are, they are not far apart. Finally, the discovery of two conjoining fragments of a ring-shaped weight (III.153; Fig. 36; Pl. 11) in two entirely different locations of the Beam-Press Complex (in Room 6 and outside Rooms 1 and 3) is an obvious example of post-depositional disturbance. It should be mentioned that both the floor and the roof/wall collapse deposits yielded considerable quantities of Minoan pottery. In the case of the floor deposits, the Minoan pottery must have been part of the floor packing of the Hellenistic structures. The presence of a large number of small and worn Late Bronze Age sherds (Late Minoan [LM] I and IIIA–B) in the ceramic assemblages from the wall and roof collapse suggests that soil containing prehistoric sherds was used in the packing of the stone-built walls and the ceilings. Directly below the foundations of the BeamPress Complex lay the LM IB ceremonial building, Building B.2 (Trenches 6500, 6600, 7600, 7700, 8600; Fig. 5). In addition, the excavations unearthed several LM IIIA–B structures beneath and in the vicinity of the Hellenistic beam press (e.g., House E beneath Rooms 6 and 7 in Trenches 6300

and 6400, and House Λ in Trench 6600; Soles 2008, 101–105, 111–112; Fig. 5). In all likelihood, the construction of the Beam-Press Complex erased several LM IIIA–B structures that were built in place of the LM IB ceremonial center (Soles 2008, 7, fig. 2). All the walls of the building are oriented north– south and east–west, which is also true of the other Late Hellenistic buildings at Mochlos (Fig. 3). The walls were constructed of rubble stones, frequently interrupted by larger blocks in the lower courses (see the south walls of Rooms 4 and 6 and the north wall of Room 5; Fig. 6). For the most part the stones are laid horizontally, but they are often placed vertically alongside doorways (see the entrance to Room 6; Pl. 2B). There is no difference in the thickness of interior and exterior walls, which varies from 0.60 to 0.70 m. The walls were preserved up to 0.80 m in height. Although there is no secure evidence, except for the amount of fallen stones found on the surface before the excavation, the upper part of the walls is restored with rubble, in the style of other Hellenistic houses in Crete (e.g., Trypitos, Lato, Eleutherna). There is no evidence for a second story. The slabs in the middle of Rooms 4 and 6 were probably used for the placement of a vertical beam to support the roof rather than an upper floor (Fig. 6; Pl. 2B). Minoan ashlar sandstone blocks were sometimes reused, and sometimes recut, for certain features in the building; examples include the steps in Room 6 leading to Rooms 4 and 5, the column base in the middle of Room 4, and the installation in the southwest corner of Room 1 (Pls. 2A, 2B). Green schist slabs used for the floors also may have been recycled from the floors of Minoan buildings in the area (Soles and Davaras 1996, 224). A large number of Corinthian-type pan tiles were recovered during the excavation (Fig. 37). The tiles, which have a ridged upper end, a hooked lower end to overlap the ridge of the tile below, and flanged sides to fit the cover tiles, conform to Örjan Wikander’s Corithian pan-tile type C2a (Wikander 1988, 208). This type of tile was popular in Greece, Asia Minor, and Cyprus from the sixth century b.c. throughout the Hellenistic period (Wikander 1988, 208 n. 59). The cover tiles found in the excavation are gable shaped in section (Fig. 38). Their width and length are equivalent to that of the pan tiles and conform to Wikander’s C1 type (Wikander 1988, 210). In addition, the cover

ARCHITECTURE, STRATIGRAPHY, AND HOUSEHOLD ANALYSIS

tiles from Mochlos are marked with U-shaped indentations on both planes. The lack of any ridge tiles or antefixes argues against the existence of a gabled roof. According to Wikander (1993, 127), “it is only from Hellenistic times onwards that we find indications that people tried consciously to avoid using ridge-tiles by building houses without a saddle roof, by replacing them with ordinary cover-tiles . . .”. The different floor elevations in the Beam-Press Complex suggest the reconstruction of

7

four units of single sloping roofs that create a system of one-slope roofs at different levels: one unit for Rooms 1 and 2; one for Room 3; one for Rooms 4 and 5; and one for Rooms 6–8 (Fig. 7). The complex’s room numbers were assigned according to the order in which the rooms were excavated. The description below does not follow the numerical order of the rooms but instead follows the way they were entered, which makes the circulation pattern through the building more comprehensible.

East Unit (Rooms 1–3) Exterior Space, East of Rooms 1–3 (Figs. 4, 6, 8, 17) The excavation of the area outside Rooms 1–3 fell within the east part of trenches 6700, 7700, and 8700 (Fig. 4). Except for a series of floating stones including an ashlar block found near the surface (+8.46 m), there were no other architectural features in the area. Because of the sloping ground, the Hellenistic levels were located at different elevations, as high as +8.35 m outside Room 3 and as low as +7.58 outside Room 2 (Fig. 17). Some of the pottery fragments found outside Room 1 joined with pottery found inside Rooms 1 and 2, including water jug III.53 (Fig. 22) from Room 1, and a transport amphora (III.92; Fig. 27; Pl. 6) and a black-glaze plastic vase in the shape of a seated satyr (III.186; Fig. 40; Pl. 14) from Room 2. Other finds include a small terracotta figurine that was discovered at +8.07 m (III.185; Fig. 40; Pl. 13; see also Soles and Davaras 1994, 434, fig. 22). Additional fragments of plastic vase III.186 were found in the fill outside Rooms 2 and 3. Seager also mentioned the discovery of “a curious rhyton, which may be late Greek, showing the bust of a satyr holding both hands to his head” (Seager 1909, 276). Similar plastic vases have been found on Delos in contexts of the late second to early first century b.c. (Hatzidakis 2004a, 374, pl. 169:ε, στ). The function of a vase like this, whether a rhyton or not, in Room 2 is unclear. The highly fragmentary condition, as well as the fact that it was found both inside and outside the Beam-Press Complex,

suggests that it might have been residual debris or the result of post-depositional disturbances.

Fill (7708.1–5) POTTERY Moldmade bowl: III.41. Lopas: P 2731 (7708.5). CERAMIC OBJECTS Loomweight: C 91 (7708.5). Loomweight: C 113 (7708.5). Human figurine: III.185. Plastic vase: III.186. MINERALS AND ROCKS Pumice (7708.1, 3–5). MAMMALS 12 indeterminate bone fragments; 1 Bos taurus; 4 Sus scrofa; 5 Ovies aries; 1 Capra hircus; 19 ovicaprids (7708.1, 3–5). MARINE SHELLS (71 MNI [6 Collected Dead]) 29 Patella (5 very large); 11 Monodonta (9 have apex, 2 open apex, 2 large); 2/17 Hexaplex (1 open body, large; 1 vermetids inside, medium; apex/ columella/distal, open body, medium; 2 columella/ distal/lip [1 medium, 1 large], columella/distal/ body [medium], 7 columella/distal [2 waterworn, 1 large], 0/5 siphon [2 waterworn, 1 large], 11 fresh MNI); 0/5 Charonia (2 body, 1 lower body,

8

NATALIA VOGEIKOFF-BROGAN

2 columella, 1 waterworn columella, 2 MNI); 4 Euthria (1 open above mouth, 1 very large); 2 Cerithium (1 open lip, 2 very large); 3 Pisania; 2 Erosaria (both L. 0.027 m); Luria (lip, L. 0.034

m); Conus; Glycymeris (waterworn); Ostrea (L. 0.036, w. 0.0505 m); Cerastoderma (waterworn, distal) (7708.1, 3–5).

Room 2 (Figs. 4, 6, 8, 11, 17) Room 2 is the south room in the East Unit. All three rooms of the East Unit share a rectangular plan of similar dimensions (Room 2 is a little bigger) and have off-centered doors facing east. Their easy access from the outside makes them suitable for shops. The excavation of Room 2 began at a surface elevation of +7.77 m at the northwest corner, +8.14 m at the northeast, +6.84 m at the southwest, and +7.12 m at the southeast corner. It was excavated in three trenches: 6700 near its entrance, 6600 over most of the room, and 7600 on its northern part (Figs. 4, 6, 8). The west and south walls of the room were already visible before the excavation. The surface layers contained roof tiles and scattered stones from the collapsed walls (Loci 6601, 6700, 6701, 6703, 7601). Two fragments of burned roof tiles were found just above the floor in the eastern half of the room (Locus 6705.1). Worth mentioning is Locus 6707, an area of collapsed wall debris in front of the room that yielded pottery of the seventh century b.c. (Soles 2008, 71, nos. IIB.1080– IIB.1085; Soles and Nicgorski 2010, 161–163). The north and east balks (Loci 6600 North and 6621) were removed after the discovery of the room and the excavation of its floor. The finds that came from the balks are listed separately unless there were joins with the pottery from the floor deposit, in which case they are included in the floor assemblage. Room 2 is a rectangular space similar to Room 1, but slightly wider. It measures 5.90 x 3.60 m. The entrance is located at the southern part of the east wall, with a doorway ca. 1.70 m wide and a pivot stone at the northern side of the doorway. The stone-built walls were preserved to a height of approximately 0.50 m. Most of the southern half of the room was covered with slabs (Loci 6605, 6705), but there is evidence that the slabs extended to the rest of the room (the notebook records slabs during the removal of the North Balk). The slabs

appeared at an elevation of +7.62–7.58 m. The large quantity of tableware (Fig. 11), more than in any other room of the complex, suggests consumption of food although very few animal bones and shells were recovered. The majority of the pottery was found in a fragmentary but restorable condition. The floor deposit yielded a considerable number of vessels used for food service, including thirteen echinus bowls with incurved rim and one handle (III.1, III.3– III.8, III.10). Another eight bowls were found during the removal of the east and north balks (III.9, III.11–III.17). Three plates with beveled rim (III.24, III.26, III.27; Fig. 20; Pl. 4) were also found on the floor, and one more plate (uncataloged P 5724) was found during the removal of the balks. The floor assemblage also included fragments from one highnecked cup (III.35; Fig. 21); another cup (III.34; Fig. 21) was found in the North Balk. Other vessels included a thin-walled beaker (III.49; Fig. 21), a possible lebes (III.28; Fig. 20), and small fragments from two moldmade bowls (III.42, III.44; Fig. 21). The fragments from the high necked cup and the moldmade bowls are small and very fine, which suggests that they are primary or secondary refuse; the vessels were no longer in use when the building was abandoned (Ault 2005, 10). On the other hand, the many and relatively large fragments of lebes III.28 suggest that the vessel was probably used before the destruction of the room (Ault 2005, 9). Room 2 also yielded four small and one large lekythoi (III.59, III.61–III.63; and see III.60; Fig. 22); two of these (III.59, III.60) were found in the north and south balks, respectively. Room 2 is the only room in the Beam-Press Complex where oil vessels were found. The storing and pouring of oil in Room 2 could also be associated with a spouted situla (III.66; Fig. 23; Pl. 5) whose pieces were found dispersed outside Rooms 1 and 3.

ARCHITECTURE, STRATIGRAPHY, AND HOUSEHOLD ANALYSIS

The excavation also uncovered several transport amphorae in the floor assemblage of Room 2 (Fig. 11), including fragments from three Koan amphorae (III.84, III.86, III.87; Fig. 26), one amphora made in the ECCW (III.74; Fig. 24), and a significant portion of an unattributed amphora (III.92; Fig. 27; Pl. 6). The amphorae would have been used for the storage of wine or some other liquid, probably water, as was jug III.52 (Fig. 22; Pl. 4). Provisional archaeochemical analysis of body sherds from amphora III.74, which has not been replicated, identified trace amounts of compounds associated with wine, although the weak strength of the chromatogram peaks suggests that the amphora may have stored other liquids as well, like water, during its life (see App. C). The presence of four chytrai (III.98, III.105, III.109, III.110; Figs. 29, 30; Pl. 6) in the floor deposit and one in the wall collapse (III.107; Fig. 30) suggests food preparation although other evidence for cooking is scanty in Room 2, just as it is in Room 1. The stone objects included a perforated weight (III.149; Pl. 11) and a hand/hammerstone (III.130; Pl. 7), both of Minoan date, which could have been reused by the Hellenistic inhabitants. The review of the pottery from Room 2 indicates food preparation and consumption, as well as activities associated with pouring/storing oil, unless the echinus bowls and the plates were stored there for other reasons. One wonders whether Room 2 functioned as a sort of a dining facility for the people who worked in Room 1.

Roof and Wall Collapse (Loci 6601, 6604.1, 6700.1, 6701.1–2, 6703, 7601.1–2, 7606.1) POTTERY Bowl with incurved rim: III.13. Chytra: III.107. Amphora, unattributed: III.76. Amphora, Koan: III.85. Amphora, Rhodian: P 3246 (6604.1/6621.1). CERAMIC OBJECTS Pyramidal loomweight: III.179. Loomweight: C 39 (7601.2). COPPER ALLOY Nail (CA 211; 7601.4). Square sheet: III.210.

9

STONES Obsidian: (6604.1). Possible balance-pan weight: III.159. Perforated weight: GS 1302 (6703.2). MARINE SHELLS (7 MNI [1 Collected Dead]) Charonia (upper body) (6700.1). 0/1 Monodonta; Hexaplex (columella/lip, large); 0/2 Charonia (2 columella, 1 large with distal siphon, 1 MNI); Fasciolaria; 0/1 Barbatia; Glycymeris (waterworn) (6701.2). LAND SNAIL Rumina (small) (6701.2). CHARCOAL 3 Olea europaea (6604.1).

Floor Deposit (Loci 6605, 6606.1–3, 6704.1–2, 6621, 6705) POTTERY 14 bowls with incurved rim: III.1, III.3–III.10 + uncataloged P 2886, P 2887, P 2898 (6606.1), uncataloged P 2907 (6601.1/6605), P 2874 (6606.2). 3 plates with beveled rim: III.24, III.26, III.27. Lebes: III.28. High-necked cup: III.35. Moldmade bowl: III.42, III.44. Thin-walled beaker: III.49. Jug: III.52. Lagynos: III.58. 3 lekythoi: III.61–III.63. Amphora, East Cretan: III.74. Amphora, unattributed: III.92. 3 Amphorae, Koan: III.84, III.86, III.87. 3 lekanai: III.95–III.97. 2 chytrai with collar rim: III.98, III.105. Chytra with inverted, grooved rim: III.109. Lamp: III.127. CERAMIC OBJECTS Loomweight: III.173. STONES Biconically perforated weight: III.149. Hand/hammerstone: III.130. Balance-pan weight: III.148.

10

NATALIA VOGEIKOFF-BROGAN

MAMMALS 6 ovicaprids (6606.1–3). MARINE SHELLS (15 MNI) 9 Patella (1 very large); Monodonta (has apex); Cerithium (open mouth, broken apex, very large) (6606.1). Patella; 0/2 Hexaplex (columella/distal/ lip, large; columella/distal, very large, 2 MNI); 0/3 Charonia (columella/ siphon; lip [very large], body, 1 MNI) (6704.2).

Plate with beveled rim: P 5724 (6621.1). High-necked cup or kantharos: III.34. 2 lekythoi: III.59, III.60. Lekane: P 5740 (6621.1). Chytra with inverted, grooved rim: III.110. 2 chytrai with collar rim: P 5729 (6621.1), P 5733 (6621.1/6600 North Balk). Lopas: P 5734 (6600 North Balk).

CERAMIC OBJECT Water pipe: III.169.

LAND SNAIL Eobania (6606.1).

LEAD Rectangular sheet: III.205. Sheet and peg: III.203. Lid: III.193.

CHARCOAL 1 Olea europaea (6704.1).

From Removal of East and North Balks (6621.1, 6600 North Balk) POTTERY 7 bowls with incurved rim: III.11, III.12, III.14– III.17 + uncataloged P 5722; 6621.1.

MARINE SHELL Charonia (lip, worn, pitted exterior) (6621.1). CHARCOAL 1 Olea europaea; 1 Pistacia lentiscus (6621.1).

Room 1 (Figs. 4, 6, 8, 11, 17) Room 1 is the middle room in the East Unit (Fig. 6). Its excavation began at a surface elevation of + 8.80–8.66 m, and it was excavated in two trenches, 7600 and 7700 (Fig. 4). The largest part of the room lay in Trench 7600, while the eastern end was in Trench 7700. It is a rectangular room measuring ca. 5.90 x 2.50 m, laid on an east–west axis, with a doorway near the southern end of the east wall. A pivot stone at the northern side of the doorway indicates that the door would swing from the northern part of the doorway and open inward. Two slabs at +7.96 and +8.08 m indicate the existence of a threshold. The first layers near the surface (Loci 7601.3, 7605.1) yielded fallen stones, broken roof tiles, pieces of pumice, and several stone tools. The excavation also revealed a rectangular paved area set against the southwest corner of the room and extending toward the middle of the south wall. The area measured 1.10 x 2.60 m and was paved with thin slabs and framed with vertical, upstanding slabs (Fig. 6; Pl. 2A). Unlike the rest of the slabs, the slab at the northwest corner of the paved area is a reused ashlar block (0.56 x 0.50 m). Of

interest is another ashlar block with a U-shaped cut on its upper surface, which has been built into the east facade of the west wall (Fig. 8; Pl. 2A). Note also that the ashlar block in the paved area is situated exactly below the ashlar block with the Ushaped cut (Figs. 8, 11). The latter was probably used to support a beam. The existence of several flat stones at the eastern part of the room, at an elevation of +8.18–7.96 m (Locus 7703.3), and a few others near the east edge of the paved area suggests that the entire floor of Room 1 was paved. The function of the paved, rectangular area with the raised edge is unknown. The U-shaped ashlar block (0.40 long x 0.30 wide x 0.25 m deep) in the west wall, if it is has not been recycled from a beam-press installation, could indicate the presence of a beam. The small distance between the U-shaped block and the surface of the paved area below (0.50 m), however, raises questions about its use and effectiveness as a beam press for olives. Presumably, there would have been more holes for the beam higher up the wall, which have not been preserved. With the beam so low, it is difficult to

ARCHITECTURE, STRATIGRAPHY, AND HOUSEHOLD ANALYSIS

imagine how any liquid would have been collected in the absence of any sunken vessel at the eastern end of the restricted paved area. If not a beam-press installation, the paved area could have served as a multi-purpose processing installation, put to different uses at different times of the year. Since no crushing equipment for olives was found in Room 6, it is possible that the crushing took place on the paved area of Room 1 (also see Ch. 5). Alternatively, the U-shaped block could have been used for anchoring the lever of a hopperrubber mill (a type of millstone found in Classical and Hellenistic sites; see Ch. 3), which would have rested on a low table or some other device, in the fashion of that found at Gamla in Israel (Frankel 2003, fig. 6). In fact, the floor deposit of Room 1 yielded two hopper-rubbers (III.143, III.144; Fig. 34; Pl. 9) and two pestles (III.135, III.136; Fig. 32; Pl. 8). Room 1 yielded other grinding equipment, including two hand/hammerstones (III.131, III.134; Fig. 32; Pls. 7, 8), an unfinished Minoan lamp probably used as a mortar (III.147; Fig. 35; Pl. 11), a grinder (III.139; Pl. 8), and part of a saddle quern (III.140; Fig. 33; Pl. 8). Except for the hopper-rubber mills, all of the other stone tools date to the Bronze Age, but this is not surprising since the Late Hellenistic inhabitants of Mochlos built their settlement in an area with extensive Bronze Age occupation. Naturally, they would have taken advantage of well-preserved stone tools or perforated stone weights lying on the surface. The same phenomenon occurs at Mycenae, where Mycenaean clay loomweights and stone tools were reused in the Hellenistic period (Bowkett 1995, 36, 39). The hopper-rubbers, the pestle, and the rest of the stone tools from Room 1 probably were used for grinding grain. The hopper-rubber mill would have stood inside the paved surface with its lever inserted into the U-shaped niche of the west wall. Whether this interpretation is plausible or not, the paved area in Room 1 must have been some kind of processing installation. Room 1 also yielded six clay loomweights (III.170–III.172, III.180–III.182; Fig. 39; Pl. 13) and four stone perforated weights (III.150, III.154, III.156, III.157; Figs. 11, 36; Pls. 11, 12). Another perforated stone weight (GS 290) was found at the surface level. The small fragment of a mortar found on the surface may not necessarily belong to Room 1. Of particular interest is a lead weight

11

(III.194; Fig. 41; Pl. 14). Most of the loomweights were found along the western half of the paved surface, probably fallen from a shelf set against the north wall of the room. The weight of the clay weights varies from 30 g to 110 g; the stone weights are much heavier, weighing from 139 g to 214 g. The lead weight weighs 166 g. The presence of so many loomweights in the room suggested to the excavators that, whatever else the room was used for, it was also used for textile production (Soles and Davaras 1996, 224). The study of the pottery from the floor deposit shows that not all of it was in use at the time of room’s abandonment, implying a combination of formation processes (Schiffer 1996). For example, transport amphora III.67 (Frontispiece; Fig. 23; Pl. 5), which was found almost fully preserved, clearly belongs to the de facto refuse of the household; amphora III.69 (Fig. 23), which is contemporary with III.67, but was found partially preserved (neck and stubs of handles), appears to have suffered from post-abandonment processes like vandalism (Schiffer 1985, 29). On the other hand, Rhodian amphora III.91 (Fig. 27), preserving part of the rim and neck, could have belonged to the primary refuse of the room (Schiffer 1985, 24–25). Jug III.56 and filter jug III.64 (Fig. 23; Pl. 5), which were found built into the north facade of the south wall, predate the occupation of the East Unit of the complex. Small fragments of the filter jug were also found in the floor deposit of Room 2, and they very likely are the product of secondary refuse, having been brought in from elsewhere and used for the construction of the floors and walls (Ault 2005, 10). Most of the pottery was found in three concentrations: (1) near the northwest corner of the room and along the east wall (Locus 7607.1A); (2) near the northeast corner of the paved area (Locus 7607.1B); and (3) along the eastern half of the south wall (Fig. 11). A complete transport amphora (III.67; Frontispiece; Fig. 23; Pl. 5, restored from many fragments) and fragments from three others (III.69, III.91, and uncataloged P 3247; Figs. 23, 27), as well as a water jug (III.53; Fig. 22), were found at the northwest corner (Fig. 11). There were also several fragments of large cooking pots (six chytrai and one lid; III.102, III.103, III.106, III.108, III.112, III.113, III.120; Figs. 29–31; Pl. 7)—some restorable, like chytra III.106—from

12

NATALIA VOGEIKOFF-BROGAN

the floor deposit, which might suggest cooking activities in the room, although the evidence for food preparation (fire) or consumption at the time of its abandonment is scanty. In addition, very few animal bones were found in Room 1 compared to other rooms of the complex. The study of the bones identified four specimens (one from a pig, two from ovicaprids, and one from a medium-sized mammal), of which only the pig bone carries traces of burning (see App. D), and over 20 edible marine shells. Finally, the lack of any drinking or eating vessels in the floor assemblage excludes the consumption of food in this room, confirming its industrial character. The large size of the chytrai also attests to the non-domestic character of Room 1. Furthermore, the easy access to all the rooms of the East Unit, without the screening of a vestibule or a corridor, also adds to the non-residential character of the three rooms (for rooms with low openness value, see Nevett 1999, 66–67, 178).

Surface (7601.1–2, 7700.1–2) POTTERY Bowl with incurved rim: P 2604 (7601.2). CERAMIC OBJECT Lentoid loomweight: C 39 (7601.2). COPPER ALLOY Nail: CA 212 (7700.1). STONES Biconically perforated weight: GS 290 (7601.2). Basin (gourna): GS 26 (7700.1). MARINE SHELL 0/2 Charonia (columella/siphon; lip fragment, 1 MNI) (7601.2).

Roof and Wall Collapse (7601.3–5, 7605.1– 2, 7703.1–2) POTTERY Amphora, Cretan: III.75. Amphora, Koan: III.82. 2 basins: P 2603 (7601.3), P 2667 (7601.4). Chytra with inverted, grooved rim: P 2602 (7601.3). Jug: P 2732 (7601.5). MINERALS AND ROCKS Pumice (7703.2).

GLASS Bowl: III.191. COPPER ALLOY Nail: CA 211 (7601.4). LEAD Sheet: PB 2 (7703.1). MARINE SHELLS (9 MNI) 2/1 Hexaplex (2 medium; 0/1 body, very large, 3 MNI); Charonia (apex and upper body) (7601.4– 5). 0/2 Hexaplex (distal/columella, medium; lip/ body [large], 2 MNI by size); 0/2 Charonia (columella/siphon, body, 1 MNI) (7605.1). Patella; 2 Monodonta (2 have apex, 2 large MNI) (7703.1).

Wall Fabric (7604A) POTTERY Jug: III.56. Filter Jug: III.64.

Floor Deposit (7607.1–2, 7608.1, 7610.1, 7703.3–6) POTTERY Plate with beveled rim: III.25. 2 moldmade bowls: III.43, III.47. Jug: III.53. 2 bases of jugs? III.55, III.57. Lekythos: P 2726 (7703.5–6). Amphora, Koan: P 3247 (7601.4–5/7605.1/7607.1). 2 amphorae, Cretan: III.67, III.69. Amphora, Rhodian: III.91. 2 chytrai with everted rim: III.112, III.113. 2 chytrai with collar rim: III.102, III.103. Chytra with inverted, grooved rim: III.108. Chytra: III.106. Lid for cooking pot: III.120. CERAMIC OBJECTS 3 lentoid loomweights: III.170–III.172. Pyramidal loomweight with incision: III.180. Pyramidal loomweight: III.181. Disk loomweight: III.172. Small biconical weight: III.182. STONES 2 hopper mills: III.143, III.144. Grinder: III.139.

ARCHITECTURE, STRATIGRAPHY, AND HOUSEHOLD ANALYSIS

2 hand/hammerstones: III.131, III.134. Saddle quern: III.140. Stone lamp/mortar: III.147. 2 pestles: III.135, III.136. 3 naturally perforated weights: III.154, III.156, III.157. Biconically perforated weight: III.152.

13

MAMMALS (7703.3–4, 7607.1) 1 indeterminate bone fragment 1 Sus scrofa 2 ovicaprids

IRON Sheet fragments: III.215.

MARINE SHELLS (24 MNI [1 Collected Dead]) Patella; 2/1 Hexaplex (1 body, 3 MNI) (7607.2); Patella (large); Monodonta (has apex); Hexaplex (apex) (7608.1). Hexaplex (very large, L. 08075 m) (7610.1). 10 Patella (3 large); 0/2 Monodonta (1 MNI); 1/3 Hexaplex (1 medium/large; 1 columella/distal/lip, 1 siphon, 1 body, 3 MNI); Charonia (columella, pitted, collected dead); Conus (worn, broken lip) (7703.3–4).

LEAD Weight: III.194.

LAND SNAIL 0/1 Eobania (7703.4).

MINERALS AND ROCKS Pumice (7703.3–5).

CHARCOAL 1 Olea europaea (7608.1).

Room 3 (Figs. 4, 6, 8, 17) The excavation of Room 3 began at a surface elevation of +10.12 m at the east and +9.54 m at the west. It was excavated as three trenches: 8700 covered the area near the entrance on the east; 7700 covered the area at the southeast corner of the room including the doorway; and 8600 exposed most of the room to the west (Figs. 4, 6, 8). After the excavation of the upper levels reached the elevation of +8.57 m, the two 8600 trenches were combined and dug as one unit (Locus 86/8716.1–5). Room 3 is a rectangular room measuring ca. 5 x 2.40 m, and it is oriented on an east–west axis, with an entrance at the southern part of the east wall. That its west wall abuts the north wall of Room 1 probably indicates that it postdates the construction of Rooms 1 and 2. A small semicircular construction (Locus 8618), very likely a bin, is located at its southwest corner (Fig. 6). According to the excavation notebooks, the upper layers (Loci 8601.1– 5, 8610.1–2, 8613.1) yielded pottery of Minoan and Hellenistic date, indicating that this is probably part of the backfill from Seager’s excavations immediately to the north (Trench 9600; see also Soles and Davaras 1996, fig. 2). The only evidence for paving was found at the southern part of the room at +8.25 m (Locus 86/8716.4). The paucity of finds

shows that its inhabitants had planned its abandonment. The stone tools (fragments from a hand/ hammerstone, saddle quern, and mortar) are Minoan. Of interest are a spindle whorl (III.176; Fig. 39) and a lead sheet with incised lines (III.195; Fig. 41) found by the doorway at +8.45–8.41 m.

Surface (7701.1–3, 8601.1–5, 8610.1–2, 8613.1) POTTERY Amphora, East Cretan: P 3509 (8601.4). Amphora, Koan: III.88. Amphora, transport: P 3158 (8601.4). Lopas: III.119. STONES Hand/hammerstone: GS 451 (8601.5). Saddle quern: GS 452 (8601.2). Mortar: GS 898 (8601.2). GLASS Bowl: III.189. LEAD Folded sheet: III.195.

14

NATALIA VOGEIKOFF-BROGAN

MARINE SHELLS (46 MNI [3 Collected Dead]) 6 Patella (1 very large); Monodonta (has apex); 1/11 Hexaplex (complete, open body, L. 0.04075 m, 1 columella/distal/part lip, 4 columella/distal [1 very large], 1 siphon/lip, 2 lip [1 large], 2 body, 1 waterworn columella/distal, 6 fresh MNI); 0/2 Charonia (2 body, 1 collected dead, 2 MNI); 2 Euthria; Cerithium (open mouth, large); Conus (slightly broken distal and lip, L. 02975 m); Erosaria (yellowish, L. 0.028 m); Spondylus (upper, hinge, medium) (7701.1–3). 6 Patella (1 large); 2/1 Monodonta (1 has apex, 1 open apex, 3 MNI); 3/7 Hexaplex (L. 0.04025, 0.05525, 0.0635 m; 1 apex [large], 1 columella/distal/lip/body [medium], 2 columella/distal/lip [2 large], 3 distal/columella [2 large, 1 very large], 9 MNI); Charonia (body, bored and collected dead); 2 Euthria; Gibbula (has apex); Fasciolaria (L. 0.04825 m); Bolinus (missing distal); Thais (broken mouth, very large, L. 0.0705 m) (8601.1–3).

LAND SNAILS (4 MNI) Eobania (7701.3). 2 Eobania (2 have some color); Rumina (8601.2–3).

Floor Deposit (86/8716.1–5) CERAMIC OBJECT Spindle whorl: III.184. MAMMALS 10 indeterminate bone fragments (86/8716.4–5). MARINE SHELL Charonia (columella/siphon, worn, medium) (86/ 8716). CHARCOAL 1 Quercus evergreen type (86/8716.5).

West Unit (Rooms 4–8) Room 7 (Figs. 4, 6) One entered the suite of Rooms 4 to 6 through Room 7. It is an oblong room measuring 3.40 x 1.50 m, opening to the south, with no remains of a doorway (Fig. 6). The west wall may be incompletely preserved at its southern end, and it is unclear whether it extended farther to the south or turned to the east, creating a narrow doorway. The northern section of the room was excavated as Trench 7300, while the southern section formed part of Trench 6300 (Fig. 4). The excavation began at an elevation of +7.67 m (northeast corner of 7300) to 6.47 m (southwest corner of 7300). The excavation of the surface layer (Locus 7302.1) yielded flat slabs and roof tiles and exposed the north wall of Room 7, which is the continuation of the north wall of Room 6. After the discovery of the north wall, the excavation concentrated on the area to the south of the wall (Locus 7304.1–3), uncovering more broken roof tiles (at +6.40 m) and the west wall of the room. The southern part of the room was first excavated as Trench 6300. The excavation began at an elevation +7.03 m (northeast corner) to +6.46

m (northwest corner). The surface layer (Locus 6301.1) revealed more roof tiles (especially in the northern part of the trench), the doorway of Room 6, the low stone construction to the south of the doorway in Room 6, and the southwest corner of Room 6. To avoid confusion, the two trenches (6300 and 7300) were combined into one trench, 63/7300, and the excavation of Room 7 was completed with Locus 63/7313.1, which unearthed remains of a paved floor (+6.26–6.18 m) that extended to the south outside the room. Although the excavation did not reveal south walls for Rooms 7 and 8, a block with a pivot hole was found outside the southwest corner of Room 6. The block suggests the existence of a gate or a wooden barrier at the southern end of Room 7. From that opening, one entered an oblong paved space, covered or uncovered, which led to Room 6. The finds from Room 7 include an intact saddle quern (III.142; Fig. 33; Pl. 8) made of granodiorite found in the floor deposit (see Fig. 14: GS 756). Querns of this type are common in Minoan

ARCHITECTURE, STRATIGRAPHY, AND HOUSEHOLD ANALYSIS

contexts, but they were obviously reused by the Hellenistic inhabitants of Mochlos. Please note that for convenience, the finds from Locus 6301.1, although they may contain some material from Room 6, are all itemized under Room 7.

Surface (6301.1, 7301.1, 7302.1) POTTERY Amphora, Koan: III.83. Amphora, East Cretan: III.73. 2 chytrai with collar rim: P 4663, P 4665 (6301.1). Chytra with inverted and grooved rim: P 4664 (6301.1). STONE Hand/hammerstone: GS 582 (6301.1). MARINE SHELLS (3 MNI [1 Collected Dead]) Hexaplex (columella/distal/lip, medium); 0/3 Charonia (apex, worn; 2 columella/siphon, 1 very large, 2 MNI) (7302.1).

Roof and Wall Collapse (7304.1–3, 63/ 7312.1) STONE Grinder: III.138. MAMMALS 6 indeterminate bone fragments; 2 Sus scrofa; 1 Ovis aries; 1 ovicaprid (63/7312.1).

15

MARINE SHELLS (12 MNI [2 Collected Dead]) 2 Patella; 4 Monodonta (4 have apex, 3 large); 2 Hexaplex (waterworn exterior but collected alive, medium; broken body, medium); Cerithium (open mouth); Conus (broken lip, L. 0.045 m); Spondylus (waterworn, upper, proximal, small); 0/1 Arca (waterworn, large) (63/7312.1).

Floor Deposit (63/7313.1) STONES 2 handstones: III.137. Saddle quern: III.142. MAMMALS 2 indeterminate bone fragments; 2 Capra hircus; 3 ovicaprids (63/7313.1). MARINE SHELLS (41 MNI [4 Collected Dead]) 13/3 Patella (3 large, largest L. 0.03625 m, 13 MNI); 10 Monodonta (7 with apex, 3 open [2 ancient, 1 recent], 4 large); 3/1 Hexaplex (1 open mouth, largest L. 0.063 m; 1 waterworn siphon/body); 0/4 Charonia (1 apex, 2 upper body [1 thin], 2 body); 7 Cerithium (5 open lip, 2 with lip [1 has hole opposite mouth], largest L. 0.0475 m; 1 worn apex and open lip); Astraea (columella/part distal); 1/2 Spondylus (waterworn, lower, proximal, small/medium; upper [rather complete], pitted exterior and part of interior, collected dead, L. 0.0875 m, w. 0.06875 m; upper, distal, small); Pinna (umbo); Chama (waterworn, medium, L. 0.03575 m) (63/7313.1). LAND SNAIL Eobania (has color) (63/7313.1).

Room 8 (Figs. 4, 6) Room 8 is located immediately to the west of Room 7. Its west wall was uncovered while excavating the surface of Trench 7200 (Fig. 4). Since most of Room 8 fell within Trench 63/7300, the latter was arbitrarily expanded to the west in order to encompass the entirety of Room 8. The open areas to the west and south of Room 8 were dug as Trenches 7200 and 6200 and are discussed below. Room 8 should be restored as an oblong room, although it is missing its southern half. Except for a few fragments of roof tiles, the area was poor in finds. The excavators were not able to locate a floor

deposit in this room. The elevation of the bottom of the west wall at +6.24 m is the best indication for the floor. The discovery of broken tiles to the west of Room 8 suggests that it was probably roofed.

Roof and Wall Collapse (63/7314.1) CERAMIC OBJECTS 2 loomweights: III.174, III.177. COPPER ALLOY Nail: III.206.

16

NATALIA VOGEIKOFF-BROGAN

STONES Sub-cuboid grinder/weight: GS 839 (63/7314.1). Quern: GS 840 (63/7314.1). MINERALS AND ROCKS Fossil oyster fragment (63/7314.1). MAMMALS 6 indeterminate bone fragments; 1 Sus scrofa; 1 ovicaprid (63/7314.1).

MARINE INVERTEBRATES (44 MNI [3 Collected Dead]) 8 Patella (2 large); 11 Monodonta (2 open apex [1 recent], 3 very large); 1/3 Hexaplex (medium; 3 distal/columella [1 waterworn], 4 MNI); 0/3 Charonia (2 columella, 1 lip, 1 MNI); 11 Euthria (1 a bit worn, 1 worn and gastropod-bored near lip, 1 very large); 5 Cerithium (2 broken mouth, 4 holed opposite mouth); Pisania; 0/3 Erosaria (1 MNI); Spondylus (waterworn, upper, small/medium); Eriphia claw (large) (63/7314.1).

Area South and West of Room 8 (Figs. 4, 6) The area to the west of Room 8 was dug as Trench 7200 (Fig. 4). The excavation began at +6.36 m and continued until +5.96 m (Locus 7202.1–3), unearthing stone tumble, probably from the west wall of Room 8, and broken roof tiles, which indicate that Room 8 was roofed. The area to the south of Room 8 was dug as Trench 62/6300 (Fig. 4). The excavation of the surface (62/6300.1) yielded large quantities of pottery, mostly dating to LM III (Soles 2008, 109) with the exception of a small number of Hellenistic sherds that included several fragments from moldmade bowls. The concentration of moldmade bowls at a surface level, compared to the very small number found in the excavation of the Beam-Press Complex, suggests that they were probably unearthed elsewhere and dumped in the area of 62/6300 by Seager. During his excavation of Block B in the area of Trenches 52/5300 and 42/4300 to the south, Seager mentions that the core of the Minoan house (now identified as House B.1) “was torn away by Roman buildings,” which he subsequently proceeded to destroy (Seager 1909, 281; for the location of House B.1 in relationship to the BeamPress Complex, see Soles and Davaras 1996, figs. 2, 27).

Surface (62/6300.1) POTTERY 7 moldmade bowls: III.37–III.40, III.45, III.46, III.48. Bowl with incurved rim: P 3487 (62/6300.1). Bowl with outturned rim: III.18. 3 high-necked cups: III.31–III.33. Kantharos(?): III.29. 2 water jugs: III.54 and uncataloged P 3488 + P 3489 (62/6300.1). Lekane: P 3484 (62/6300.1). CERAMIC OBJECTS Human figurine: C 576 (62/6300.1). Animal figurine: C 702 (62/6300.1). MARINE SHELLS (54 MNI [4 Collected Dead]) 33/4 Patella (10 large); 5 Monodonta (5 with apex, 3 large); 4/7 Hexaplex (1 very large, L. 0.081 m; 1 distal/columella/lip, 4 distal/columella, 2 body, 9 MNI); 0/6 Charonia (1 distal siphon, 5 body [some attach]); Fasciolaria; 0/3 Spondylus (3 upper proximal, 2 waterworn, 1 small and fresher); 2 Barbatia (waterworn, R, small; waterworn, L, medium) (62/6300.1). Some of these were published as coming from the surface over the area between Houses E and Z (Soles 2008, 110).

Room 6 (Figs. 4, 6, 10, 14, 15, 16, 18; Pls. 2B, 3) From Room 7 one entered Room 6, the main room of a three-room suite (Rooms 4–6), which led in turn to Rooms 4 and 5 through two doorways in its east wall. The excavation of Room 6

began at a surface elevation of +7.40 m (northern part) to 7.07 m (southern part). It was excavated in four trenches: 7400, 6400, 7300, and 6300 (Fig. 4). The largest part of the room lay in trenches

ARCHITECTURE, STRATIGRAPHY, AND HOUSEHOLD ANALYSIS

7400 and 6400, which were eventually dug as one trench, 64/7400. According to the excavators, the upper layers were loaded with rocks and very few finds. Therefore, it was thought that they might have included material from Seager’s dump since he had dug in the area north of the BeamPress Complex (see Soles and Davaras 1996, fig. 2). The removal of the rocks revealed fragments of roof tiles to the northeast and southwest of the room (Locus 7410.1–2 at +6.84–6.42 m; Locus 64/7412.1–2 at +6.36–6.17 m) and all along the south facade of the north wall (Locus 7404). Room 6 is an almost-square room measuring 5.20 x 5.05 m, laid on an east–west axis. It was provided with a doorway, 0.80 m wide, in the middle of the west wall. The floor deposit was reached at +6.24–6.17 m. The paved threshold of the doorway stood higher (+6.39 m) than the floor deposit (Fig. 10; Pl. 2B). The discovery of several schist slabs likely indicates the existence of a paved floor, which was only partly preserved. The two flat slabs in the middle of the room indicate the existence of a column used to support a central beam in the ceiling (Fig. 18:a; Pl. 2B). It is almost certain that Room 6 would have carried a high ceiling because of the higher floor levels of Rooms 4 and 5. The most important architectural feature of Room 6 is the beam press that was found built against the east wall of the room (Figs. 14, 15; Pls. 2B, 3). The beam press is made of two parts: a platform and a back support located at the southern end of the platform. The total length of the construction is 2.05 m. The platform, which stood 0.50 m above the floor, is built of large boulders in the lower part and large, flat schist slabs in the upper part. Its surface measures 1.55 x 1.06 m, with the press bed placed in the middle of the northern end. The press bed is rectangular (0.517 x 0.376 m) and spouted, and it is made of nonlocal white limestone (III.145; Fig. 35; Pl. 10). The back support is built with large and small stones at the eastern end, while the western end is made of one large block (0.90 x 0.60 x 0.31–0.38 m). The wooden beam would have rested either on the back support, making a short beam (Fig. 18), or it would have been anchored in a niche at the south wall of Room 6. Large perforated stone weights would have been hung on the other end of the beam to create a downward pressure (Brun 2004, 13–14, type A1). One of the reviewers has suggested that the ashlar block found in Room 4

17

(see below) might have served originally as a counter weight in Room 6. In either case, how the beam would sit on the back support remains problematic because it would never be possible for the beam to exert full pressure on the frails (flat baskets made of woven fiber) containing pulp. Furthermore, the anchoring of the beam on the south wall would have hindered access to Room 4 during the pressing period, unless Room 4 had gone out of use. Another possibility is that the press beam ran east–west and was anchored into the east wall. A beam of about 3.7 m would run to the stones in the center of the room, which have been reconstructed as a base for a column. In this case, the stones could have served as the base for a capstan or a similar arrangement for exerting pressure by pulling downward on the beam (Brun 2004, 13–14, type A2). There are three problems with this reconstruction: first, the beam would still have to sit on top of the so-called back support (which would become the “front support”), thus not exerting full pressure on the frails; second, the height and size of the room requires a central support; and finally, it does not explain the north–south orientation of the platform and the press bed. Half of a saddle quern (III.141; Fig. 33; Pl. 8) dating to the Late Bronze Age was found on top of the press, although it is unclear whether it was found in situ or had been placed there after the abandonment of the building. The remains of about 50 complete and crushed charred olive stones were collected from the floor level (see App. F). The operation and function of the beam press is discussed in Chapter 5. Other features in the room included the remains of a low hearth (Locus 7311) in the northwest corner of the room. Lined with stones on three sides, it formed a rectangular space ca. 0.55 m wide, which opened toward the room, and contained much ash (Fig. 6). The hearth was probably used for food preparation and for boiling water, which was necessary during the pressing procedure. The remains of prickly oak and olive charcoal (Table 1) in the room might have been used as firewood for the hearth (Blitzer 2004, 195). Most of the bone collected from the area near the open end of the hearth belongs to an ovicaprid and comes from the animal’s extremities, which are bones rich in marrow and fat; the meat-bearing bones are missing (App. D). The study of the bone fragments also show that they were not burned, which might indicate a special

18

NATALIA VOGEIKOFF-BROGAN

kind of cooking and diet. In the late 19th/early 20th centuries, for example, the inhabitants of the Mesara exploited the meat of pigs in many ways in order to stretch meals and food supplies. To prepare a special dish called tsiladia or pychte, they boiled the feet and head of the pig separately from the meat and allowed the mixture to sit for hours before removing the meat and sieving the liquid. They then cooked the head and feet of the pig and the sieved liquid in a stewpot, or tsoukali, and ate the dish when it was cold, after the fat had congealed (App. D; see also Blitzer 2004, 138). The presence of a large number of edible shells (including 66 Patella, 97 Monodonta, and 30 Hexaplex), as well as a single fish bone, also suggest consumption of food in Room 6. The discovery of a small area full of plaster (at +6.24 m, Locus 64/7416) to the east of the hearth and against the north wall is puzzling. It could have formed part of a bench. While digging the eastern part of Trench 6300, the excavators came across a low rectangular construction measuring 0.77 x 1.08 m, built of two courses (pres. height 0.13 m) and abutting the west wall of the room, immediately to the south of the doorway (Fig. 14; Pl. 2B). It is difficult to explain the function of this construction, although it may have served as a table of some sort. In olive beam presses of the late 19th/early 20th centuries in the Mesara, a flat square or rectangular stone next to the beam press served as a platform where the frails or baskets (boxas) were filled with the crushed olives and folded over before being stacked in piles on the press (Watrous, Hadzi-Vallianou, and Blitzer 2004, 155, pls. 6.34, 6.35). The pottery from the floor assemblage of Room 6 was extremely fragmentary and unable to be restored. This situation is the opposite of that from Rooms 1 and 2, where a small number of restorable vessels were found. (Note that the pottery from 6301.1 is discussed above with the pottery from Room 7, the southern part of which falls largely within Trench 6300.) The floor of Room 6 (Fig. 14) yielded fragments from three transport amphorae (III.78, III.80, III.81; Fig. 25), the handle of a coarse lekane (III.94; Fig. 28) very similar to the one found in Room 4 (III.93; Fig. 28; Pl. 6), and the rims of a few cooking pots and a lid (III.104, III.111, III.115, III.118, III.122; Figs. 29–31; Pl. 7). The dearth of ceramic finds suggests that the abandonment of Room

6 was planned, although it remains odd that the press bed was not taken away. Other finds included folded straps and sheets of lead (III.199–III.202; Fig. 41), two stone perforated weights (III.150, III.158; Pls. 11, 12), a hammerstone (III.132; Fig. 32; Pl. 7), and a grinder made of bluish-gray andesite (III.138; Fig. 32). The andesite grinder (III.138) almost certainly is Hellenistic, made of the same material (bluishgray andesite) as the hopper-rubber mills (III.143, III.144; Fig. 34) found in Room 1. Note that the two stone weights (III.153, III.155) were not heavy enough to have served as counter weights for the beam press (see also Ch. 3). The ring-shaped stone weight III.153 (Fig. 36; Pl. 11), with a preserved weight of 1.5 kg, is an unusual find. Two fragments were found on the floor deposit of Room 6, while a third (conjoining) was recovered from the area outside Rooms 1 and 3 (Locus 7807.1). The conditions of its discovery are a good example of post-abandonment processes, where objects are moved away from the locations where they were used. The weight should be restored with two holes like its parallels from Delos, which were once connected with fishing activities (Deonna 1938, 202– 203, fig. 238). Although the main function of Room 6 was connected with pressing, the “fishing ring” (III.153), the perforated stone weights (III.151, III.155), and the large amount of marine shells found in the floor deposit, as well as the unique fish bone from the roof and wall collapse, may suggest that the people involved in the operation of the beam press or the people who frequented Room 6 were also occupied with fishing activities.

Surface (64/7400, 7401.1–2) POTTERY Amphora, East Cretan: III.70. Amphora, Koan: III.89. Amphora, Koan(?): P 3530 (7401.2). MINERALS AND ROCKS Fossil scallop fragment (64/7400). MARINE SHELLS (33 MNI [1 Collected Dead]) 11 Patella (1 waterworn); 13/1 Monodonta (13 have apex, 13 MNI); 0/8 Hexaplex (1 apex, 2 lips [1 large], 3 columella/distal, 3 MNI); 0/7 Charonia (7 body, 2 have siphon notches, 2 worn, 1 MNI);

ARCHITECTURE, STRATIGRAPHY, AND HOUSEHOLD ANALYSIS

Euthria (open body); Tonna (body, large individual); 0/11 Pinna (1 valve) (64/7400). 0/2 Charonia (large columella; distal body, 1 MNI); Euthria (bit worn) (7401.2).

19

Spondylus (upper, lacks distal, small); Paracentrotus (test fragment, 27 spines) (7410.2).

LAND SNAILS (5 MNI) 5 Rumina (5 lack color or gloss) (7401.2).

LAND SNAILS (24 MNI) 12 Eobania (12 lack color and gloss); 6 Rumina (6 lack color and gloss) (7405.1–2). 5 Helicella; 0/4 Eobania (7410.2).

Roof and Wall Collapse (6401.1–3, 7405.1– 2, 7410.1–2)

Floor Deposit (6407, 64/7412.1–2, 64/7413.1, 7310.1, 7411.1)

POTTERY Amphora, East Cretan: III.72.

POTTERY Plate with beveled rim: III.22. Thin-walled beaker: III.50. Filter jug(?): III.65. Amphora, East Cretan: III.68. 3 amphora, Koan: III.78, III.80, III.81. 2 lekanai: III.19, III.94. Chytra with everted rim: III.111. Chytra with collar rim: III.104. 4 lopades: III.115–III.118. 1 lopas lid: III.122. Lamp: III.128.

STONE Differentially weathered cobble: GS 865 (7405.1). MINERALS AND ROCKS Fossil scallop fragment (7410.2). MAMMALS 1 Sus scrofa; 1 ovicaprid (7410.2). FISH 1 indeterminate bone fragment (7410.2). MARINE SHELLS (24 MNI [3 Collected Dead]) 2 Patella; 2 Monodonta (2 have apex, 1 large); 3/3 Hexaplex (lacks upper body, small/medium; medium, L. 0.05875 m; ancient irregular hole on body, large, L. 7375 m; 1 columella/distal/lip [medium], 2 columella/distal, 6 MNI); 0/15 Charonia (2 apex [1 rather thin, small], 1 columella/distal/ lip, 2 columella/distal, 2 lips [1 large; small may have a cut edge about 0.031 m in from lip edge], 2 body with siphon notches, 6 body, 2 MNI); 2 Euthria (1 broken lip); 5 Cerithium (3 open lips, 3 very large); 0/2 Pinna; 3 Chama (3 waterworn); Glycymeris pilosus (worn, hinge) (6401.1).

LEAD 2 folded straps: III.199, III.200. 2 sheets: III.201, III.202.

MARINE INVERTEBRATES (32 MNI [1 Collected Dead]) 2 Patella; Monodonta (has apex); Hexaplex (lip); 0/2 Charonia (body with siphon notch, lip); Cerithium (open mouth); 0/2 Spondylus (2 upper, 2 collected alive: broken distal, medium; hinge, bit bored exterior, medium, 2 MNI) (7405.1–2). 8/19 Patella (8 MNI); 3/13 Monodonta (2 lips, 3 MNI); 0/29 Hexaplex (2 apex, 3 distal/columella [1 medium, 2 large], 3 lips (1 worn, 3 MNI); 0/4 Charonia (1 apex, 1 siphon, 2 body, 1 MNI); 2/2 Cerithium (1 broken apex, lacks distal; 1 open mouth; body; tiny, 3 MNI); Conus (waterworn, tiny); Erosaria (no color or gloss, L. 0.025 m); Gibbula (medium); Columbella (no color);

MINERALS AND ROCKS Fossil scallop fragment in matrix (64/7412.1).

STONES Naturally perforated weight: III.158. Ring-shaped weight: III.153 + GS 605 (64/7414.1) + GS 606 (7807.1). Hammerstone: III.129. Hand/hammerstone: III.132. Biconically perforated weight: III.150.

MAMMALS 174 inderterminate bone fragments; 4 Bos taurus; 14 Sus scrofa; 12 Capra hircus; 2 Ovis aries; 37 ovicaprids; 2 medium-sized mammals (64/7412.1–2). MARINE INVERTEBRATES (254 MNI [9 Collected Dead]) 2 Patella (1 very large); Monodonta (has apex, large); 1/2 Hexaplex (complete, broken lip, medium; 2 distal/columella, 3 MNI); 0/3 Charonia (3 body, 1 has part of lip, 1 MNI); Cerithium (open mouth);

20

NATALIA VOGEIKOFF-BROGAN

2 Conus (1 small); 0/2 Pinna (7310.1). 41/5 Patella (1 waterworn, 3 large, 1 very large: L. 0.03875 m); 51/3 Monodonta (7 open apex [6 recent, 1 ancient], 7 large, 2 very large, 5 M. articulata [2 large], 52 MNI); 6/20 Hexaplex (1 a bit worn, 3 apex, 4 lip, 5 distal/columella, 1 distal/columella/body, 1 siphon; 3 very large, 1 large, 10 MNI); 0/19 Charonia (4 siphon, 1 siphon notch, 14 body [1 thick and worn exterior]; one may have two cut parallel sides, L. 0.03625 m, w. 0.01575 m, 4 MNI); 4 Euthria (3 gastropod-bored holes on mid body; lip recently broken; open body; worn, open body); 3/1 Cerithium (waterworn, distal; 3 fresh MNI); 3 Conus (2 waterworn, 3 small/medium); 4 Pisania (1 worn apex, 1 open body/side, 1 broken lip, max. L. 0.02575 m); Erosaria (recently broken, medium); 2 Gibbula (1 small, 1 medium); 3 Bittium (3 small); Fasciolaria (recently broken lip); Mitra (has gloss, no color); gastropod (waterworn, could be strung); Spondylus (upper, hinge, medium); 0/6 Pinna; 2 Chama (1 upper, 1 lower, 1 large, 2 MNI); Eriphia pincer (lower part, large, 2 pieces) (64/7412.1–2). 23/2 Patella (5 very large, largest L. 0.0545 m); 44 Monodonta (9 open apex [5 ancient, 4 recent], 2 very large, 13 large, 2 small/ medium, 2 small, 3 M. articulata); 2/9 Hexaplex (2 complete [L. 0.04175 m, L. 0.051 m]; 1 apex, 1 lip/body, 5 columella/distal, 1 columella, 1 distal, 1 large, 1 very large, 9 MNI); 0/10 Charonia (1 apex [worn], 1 lip [thick, very large], 1 body/ siphon [very large], 1 internal columella, 4 body [1 has siphon notches, large, 2 pieces], 1 distal/ columella, 2 MNI); 5 Euthria (1 worn white, 1

recently broken, 1 with gastropod-bored hole on body, 1 with 2 attempted gastropod borings); 2 Cerithium (2 open lip); Conus (waterworn, slightly broken lip); 2 Gibbula; 2 Columbella; 0/2 Tonna (large); 0/2 Phalium (waterworn siphon/body; lip [recent break, thick], 2 MNI); Bolinus (broken distal, L. 0.0515); Astraea; Spondylus (lower, L. 0.04175 m, w. 0.04225 m) (64/7413.1). Patella; 0/12 Hexaplex (9 distal/columella [2 waterworn], 2 body [2 waterworn], 7 fresh MNI); 0/3 Charonia (2 body [1 worn but collected alive], 1 columella); Conus (worn, medium) (7412.1).

LAND SNAILS (20 MNI) 4 Eobania (3 broken, 2 with color); 8 Helicella; 6 Rumina; Enidae (64/7412.2). Helicella (slight color, large) (64/7413.1). CHARCOAL 3 Olea europaea; 4 Quercus evergreen type (64/ 7412.1). 2 Olea europaea (64/7412.1–2). BOTANICAL REMAINS Over 50 Olea europaea stones (64/7412.2). Olea europaea stone (7310.1).

Beam Press (7411.1) STONES Saddle quern: III.141. Press bed: III.145.

Room 4 (Figs. 4, 6, 12, 13) Two steps in the southeast corner of Room 6 led up to the doorway that provided access to Room 4, which was set at a higher elevation. The excavation of the room began at a surface elevation of +7.77– 7.33 m and was carried out in two trenches, 6500 and 6500N (Fig. 4). Trench 6500N is technically the southern part of Trench 7700. The largest part of the room lay in Trench 6500. Room 4 is a rectangular room that is slightly wider at its western end and measures ca. 5.10 x 3.50/3.80 m on an east–west axis. A doorway, ca. 1.30 m wide, was located at the southern half of the west wall flanked by an upright orthostate block on its northern side (Fig. 6). Two steps made of a single ashlar block measuring 0.72 x 0.50 m led up to the threshold from Room 6. They once were part of a Minoan structure and are in secondary use here.

The lower of the steps lay at +6.55 m and the higher at +6.72 m, about 0.50 m above the floor level of Room 6. The stone threshold at the top of the steps was found broken in several pieces, at +6.72 m. Its eastern side was faced with small stones. Since most of the finds in Room 4 were found slightly higher (+6.82 m) than the threshold level, it is likely that the row of small stones against the eastern face of the threshold delineated the higher elevation of the earthen floor. The walls of the room were preserved to a height of 0.65–0.70 m. The layer of destruction (Loci 6501.2–6, 6510.1– 3) revealed large rocks, broken roof tiles, fragments of thin red and green schist slabs, fragments of plaster, and large patches of charcoal (+7.39– 6.95 m; Locus 6510.2–3), especially in the northern half of the room. It is not clear whether the

ARCHITECTURE, STRATIGRAPHY, AND HOUSEHOLD ANALYSIS

schist slabs supplemented the roofing system or were part of a floor that did not survive. An intact pan tile was found near the northwest corner of the room at +7.01 m with the neck of transport amphora III.79 resting on top of it. The removal of the destruction layer in the middle of the room revealed a Minoan ashlar block (0.50 x 0.040 x 0.30 m) at +7.29 m. The block carried a square hole (0.09 x 0.08) on its western side. The block was probably used as a base for a central column, although its bottom stood approximately 0.20 m higher than the floor (Fig. 13). Of special interest is the fact that after the removal of the ashlar block, the top of a Minoan pillar emerged at +6.88 m, and subsequent excavations unearthed a LM IB pillar crypt below near where the first ashlar block stood (Fig. 5; for the western Minoan pillar crypt of Building B.2, see Soles and Davaras 1996, 187, fig. 7). The excavators reached the floor deposit at +6.89–6.82 m (Loci 6504.1–2, 6505, 6511.1–2, 6512.1). Several small finds are associated with Loci 6504.1 and 6505, like a bronze ring (III.211; Fig. 42), three bronze nails (III.207–III.209; Fig. 42), two loomweights (III.176, III.183; Fig. 39), and two lead sheets (III.197, III.198; Fig. 41). At the end of Locus 6504.1, at +6.82 m, the excavator, having removed all the finds from the floor, including the ashlar block, dug another thin pass (Locus 6504.2) and found a small stone weight (III.151; Pl. 13) below the ashlar block (+6.81 m) and a loomweight (III.178; Fig. 39) near the western side of the room (+6.78 m). Finally, at the northeast corner the excavation uncovered at +7.14 m the top of the north wall of the west LM IB pillar crypt of Building B.2 running obliquely to the room (Fig. 4; Soles and Davaras 1996, 187, fig. 7). It is very likely that in the Hellenistic period the tops of both the Minoan pillar and the Minoan wall had been incorporated into the use of Room 4. Although not identifiable during the excavation or through pottery analysis, the stratigraphic evidence suggests that Room 4 was refurbished during its occupation. In the first phase the top of the Minoan pillar was probably used as a base for a central beam. In the next phase the floor was raised ca. 0.10 m, covering the Minoan pillar, which was replaced by the ashlar block (Fig. 13). During this phase a row of small stones was placed against the threshold to ease the entrance to the room. The LM IB wall

21

that runs obliquely and isolates the northeast corner of the room could have been used as a bin area. Although most of the pottery found in Room 4 is fragmentary (Fig. 12), three of the vessels were found partially restorable: the largest part of a Koan transport amphora (III.79; Fig. 25; Pl. 6) was unearthed in the northwest corner of the room (+7.11– 6.89 m, Locus 6511.1) and the upper half of a large basin with horizontal handles (III.93; Fig. 28; Pl. 6) and most of a large chytra (III.114; Fig. 30; Pl. 7) came from the northern half of the room (at +7.08– 6.95 m, Locus 6511.1; see Fig. 12). The depositional pattern of Room 4 differs from that of the rest of the rooms in the West Unit, but it could be that the three aforementioned vessels were not worth recovering after the abandonment of the building. The excavation also recovered fragments from a bowl with an incurved rim (III.2; Fig. 19), two plates with beveled rim (III.20, III.21; Fig. 20), a thin-walled vessel (III.51; Fig. 21), three more chytrai (III.99– III.101; Fig. 29; Pl. 6), and several lids (III.121, III.123–III.126; Fig. 31). The handle of kantharos III.30 (Fig. 20) could have formed part of the primary refuse in the room. The mixed nature of the pottery (amphorae, cooking pots, and tableware), in association with a moderate amount of animal bones and a considerable quantity of marine shells found in the floor deposits, suggest preparation and consumption of food (App. D). The provisional archaeochemical analysis of the Koan amphora (III.79) appears to have produced trace amounts of compounds associated with wine, tartaric acid in particular. The low content of tartaric acid may be explained by the subsequent reuse of the amphora to store water (see App. C). Similar analysis of fragments from chytra III.100 has identified cholesterol, which is associated with cooking meat products (see App. C). The four loomweights—three clay (III.176, III.178, III.183; Fig. 39) and one stone (III.151; Pl. 11)—do not constitute enough evidence for weaving activities. A large quantity of plaster fragments found throughout the room suggests that Room 4 might have had plastered walls and been used more as a living quarter and less as a work space. Room 4 also yielded several metal objects, including a bronze ring (III.211; Fig. 42), bronze nails (III.207– III.209; Fig. 42), and lead sheets, some with perforations (III.196–III.198, III.204; Fig. 41).

22

NATALIA VOGEIKOFF-BROGAN

Roof and Wall Collapse (6501.1–6, 6510.1–2) POTTERY Lopas: P 2537 (6501.6). CERAMIC OBJECTS 2 loomweights: C 58 (6501.3), C 112 (6501.4). GLASS Bowl: III.188 (G 2; 6501.5). STONES Whetstone: III.146. Grinder: GS 880 (6501.4). Hammerstone/severed cobble: GS 1142 (6501.5). Naturally perforated weight: GS 550 (6501.5). Obsidian (6501.4–5). IRON Nail: Fe 10 (6501.1). MAMMALS 1 Capra hircus; 10 ovicaprids; 1 medium-sized mammal (6501.2, 4). MARINE SHELLS (21 MNI [4 Collected Dead]) 5 Patella (1 small, 3 large); 1/4 Hexaplex (complete is waterworn; 2 columella/distal/lip, 2 columella/ distal, 1 lip, 1 waterworn body, 4 MNI); 0/4 Charonia (1 upper spire, 2 body, 1 columella, 1 MNI); Euthria (large); 2 Cerithium (2 open mouth; 1 open apex); Pisania; Spondylus (upper, hinge, bored exterior, fresh interior, small/medium); 0/1 Barbatia (waterworn, R) (6501.2,4–5). 2 Patella (1 large); Charonia (upper body); 2 Euthria (waterworn, complete, gastropod-bored in center of body; broken lip) (6510.2). LAND SNAIL Rumina (no color or gloss) (6501.2).

Floor Deposit (6504.1–2, 6505.1, 6510.3, 6511.1–2, 6512.1) POTTERY Bowl with incurved rim: III.2. 2 plates with projecting rim: III.20, III.21. Plate with beveled rim(?): III.23. Black-gloss plate/bowl: P 4218 (6504.1). Kantharos(?): III.30. Amphora, Koan: III.79. Amphora, Rhodian: III.90.

Large lekane: III.93. 3 chytrai with collar rim: III.99–III.101. Chytra with everted rim: III.114. 5 cooking lids: III.121, III.123–III.126. Thin-walled vessel: III.51. Thin-walled plate or lid: P 2853 (6510.3).

CERAMIC OBJECTS 3 loomweights: III.176, III.178, III.183. GLASS Bowl or skyphos: III.190. COPPER ALLOY Ring: III.211. 3 nails: III.207–III.209. LEAD 4 sheets: III.196–III.198, III.204. IRON 2 nails: III.212, III.213. Slag: III.214. STONE Biconically perforated weight: III.151. MINERALS AND ROCKS Fossil scallop fragment (distal, large); oyster fragment (6511.1–2). MAMMALS 21 indeterminate bone fragments; 1 Sus scrofa; 3 Ovis aries; 4 Capra hircus; 7 ovicaprids; 2 medium-sized mammals; 1 Lepus (6504.2, 6510.3, 6511.1–2, 6512.1). MARINE INVERTEBRATES (106 MNI [5 Collected Dead]) 2 Patella (1 large); Hexaplex (open body, waterworn); 0/2 Charonia (body; siphon notch [large]); Spondylus (waterworn, lower, very large and very heavy, L. 0.09175 m, w. 0.07625 m); 0/1 Pinna (6504.1). 16/2 Patella (2 large); 23/9 Monodonta (4 open apex, 1 apex fragment, 5 lip, 4 small, 4 large, 27 MNI); 1/11 Hexaplex (broken mouth; 1 columella/ distal/lip, 1 siphon, 1 lip, 3 body, 7 MNI); Gibbula; 0/3 Charonia (3 body, 1 MNI); 2 Euthria (1 broken apex [recent], 1 broken lip [recent]); Conus (slightly broken lip, small, L. 0.01575 m); 2 Pisania; Columbella (color and gloss); Eriphia claw (lower part) (6505.1). 14/2 Patella (2 large); 10/4 Monodonta (8 have apex, 1 open apex [recent], 1 worn apex, 1 large, 2 M. articulata [1 large], 10 MNI);

ARCHITECTURE, STRATIGRAPHY, AND HOUSEHOLD ANALYSIS

1/14 Hexaplex (complete, medium; 1 apex, 1 lip, 2 siphon [1 waterworn], 5 columella/distal [2 waterworn], 1 siphon/body [very large], 1 waterworn lip, 4 fresh MNI); Charonia (body); 1/1 Euthria (broken lip; apex, 2 MNI); Erosaria (L. 0.023 m); Bittium (waterworn); Tonna (body, large); Arca (distal); Ostrea (broken distal) (6511.1–2). Patella (very large, L. 0.045); Hexaplex (body/siphon, medium); Euthria (broken mouth area) (6512.1).

23

LAND SNAILS (8 MNI) 3/2 Eobania (1 juvenile with color, 4 lack color [2 young], 5 MNI) (6505.1). 0/1 Eobania; 0/1 Rumina (adult) (6511.1). Eobania (has gloss, distal) (6512.1). CHARCOAL 2 Olea europaea (6501.4). 1 Olea europaea (6501.6). 1 Pistacia lentiscus; 1 Prunus amygdalus (6510.2).

Room 5 (Figs. 4, 6, 9) One entered Room 5 from Room 6 by means of two steps that were set inside the line of the east wall of Room 6 (Fig. 6). Room 5 was excavated in one trench, 7500 (Fig. 4). The excavation began at a surface elevation of +8.25 m (northwest) to 7.87 m (southeast) and encountered a large amount of stone tumble. Broken roof tiles and a number of objects associated with the destruction level of the room were unearthed at an elevation of +7.14 m–7.10 m (Loci 7510.3, 7511.1, 7513. 1). The doorway to the room, ca. 0.50 m wide, is located in the northeast corner of Room 6, at the southern end of the west wall of Room 5. It is provided with two steps that lead up into the room, the lower made of green schist at +6.89 m, and the upper, a reused ashlar block that forms the actual threshold of the doorway at +7.09 m. The room measures ca. 4.87 (east–west) x 2.43 m (north– south). Its floor, found at +7.14–7.10 m, is built ca. 0.80 m higher than that of Room 6 (Figs. 9, 10). Even with the help of the schist step, the top of which lies ca. 0.50 m above the floor level of Room 6, access to Room 5 would have been difficult unless there were one or two additional steps below the schist block that were removed at some point. At the southeast corner of Room 5, at floor level, the excavators unearthed the upper surface of an obliquely running LM IB wall (Fig. 5). It proved to be the west wall of Rooms 2.2 b and 2.2 c of the Minoan Hall lying underneath (Soles and Davaras 1996, 187, fig. 7), parts of which the Hellenistic inhabitants of Mochlos had incorporated into their structure, as they had done in Room 4. Moreover, a lot of wood charcoal, a moderate quantity of animal bones originating from pigs and ovicaprids, and over 25 edible marine invertebrates

were collected (App. D). A triton shell (III.192; Pl. 14) may have been used to draw liquid out of their containers (App. E). There was very little pottery in Room 5, of which only Cretan amphora III.77 (Fig. 25) has been cataloged. The floor deposit also yielded a perforated weight, weighing 1,780 g, which could have been hung together with other weights at the northern end of the beam in Room 6 (III.155; Pl. 12).

Roof and Wall Collapse (7503.1–2, 7508.1–3) STONE Grinder: GS 892 (7503.2). MINERALS AND ROCKS Pumice (7503.2, 7508.1–3). MAMMALS 5 indeterminate bone fragments; 2 Sus scrofa; 1 Capra hircus; 3 ovicaprids (7508.1–3). MARINE SHELLS (7 MNI [2 Collected Dead]) Patella (large); 0/3 Hexaplex (open body; 1 waterworn columella/distal/lip; 1 waterworn distal; 1 fresh MNI); 0/2 Charonia (attaching body fragments); Euthria; Erosaria (no color or gloss, L. 0.0225 m) (7508.1+3). CHARCOAL 1 undeterminable (7508.3).

Floor Deposit (7510.1–3, 7511.1, 7513.1) POTTERY Amphora, Cretan: III.77.

24

NATALIA VOGEIKOFF-BROGAN

CERAMIC OBJECT Loomweight: III.175.

vermetids on exterior, recent holes on body, slightly worn apex, lip a bit encrusted, L. 0.218, w. 0.1077 m); 0/2 Charonia (distal columella/siphon/ part body; lip); Euthria (large); gastropod internal columella (7510.1–3). 8/17 Patella (8 MNI); 2/15 Monodonta (1 has apex, 1 open [recent]; 3 lips, 3 MNI); 0/13 Hexaplex (1 apex, 5 body, 1 siphon [small], 1 lip; 1 waterworn, 2 fresh MNI); Gibbula (medium); Euthria; 1/1 Pisania (worn, 2 MNI); 2 Bittium; Arca (waterworn hinge); Paracentrotus (6 test fragments, 12 spines, 1 internal piece) (7511.1). 7 Patella; 2/1 Monodonta (2 have apex, 1 large, 3 MNI); 0/4 Charonia (2 apex [1 worn]; 2 body [2 worn], 2 MNI); Euthria; 0/2 Cerithium (2 body [1 worn], 2 MNI) (7513.1–3).

GLASS Bowl or skyphos: III.187. STONE Weight or hand/hammerstone: III.133. Perforated weight: III.155. MINERALS AND ROCKS Fossil scallop fragment (7511.1). MAMMALS 22 inderterminate bone fragments; 4 Sus scrofa; 3 Ovis aries; 13 ovicaprids (7510.1–3, 7511.1). MARINE INVERTEBRATES (48 MNI [2 Collected Dead]) 2 Patella (2 very large); 2/5 Hexaplex (1 very large, L. 0.06625 m; 4 distal/columella, 6 MNI); Charonia (III.192) complete, collected alive, some

LAND SNAILS (6 MNI) 6 Helicella (2 have color) (7511.1). CHARCOAL 1 Olea europaea (7510.2). 1 Prunus sp., 1 undeterminable (7510.3). 4 Olea europaea; 2 Prunus sp.; 1 Prunus amygdalus (7511.1).

North of Rooms 6–8 (Figs. 4, 6) The area to the north of Rooms 6, 7, and 8 was excavated as part of Trenches 7300N and 7400N (Fig. 4). The surface lay at an elevation of +8.03 (northeast corner of 7400N) to +7.17 m (northwest corner of 7300N). A large stone pestle (S 125) was found in the northwest corner of 7300N. Large amounts of pottery and rocks, sometimes with no filling dirt, were found in a deposit below (Locus 7403.1–3, +7.52–7.13 m). The large number of rocks found in the trench suggests that Seager may have used this area as a dump (see also Soles and Davaras 1996, 190). During the Late Hellenistic period, the area functioned as an exterior space. The fragments of a transport amphora (III.83; Fig. 26) were located in the northeast corner of 7300N (Locus 7301.2) at approximately +7.03 m. Of special interest is an amphora handle stamped with a bee (III.71; Fig. 24), which was found at an upper level of Trench 7400 (Locus 7403.3) at +7.13 m. The elevation is the same as that of the floor level in Room 5, and it is likely that the stamped amphora handle was connected with the habitation of the BeamPress Complex, although it might belong to Seager’s

excavation debris. The possible association of the amphora with the Hierapytnian ceramic workshops is explored in Chapter 2. The excavation of Trench 7400N terminated at an elevation of +7.13 m, while that of 7300N terminated at +7.03 m.

Surface (7301.1, 7400) STONE Pestle: S 125.

Tumble (7301.2, 7401.1–2, 7403.1–3) POTTERY Amphora, Koan: III.83. Amphora, Cretan: III.71. CERAMIC OBJECTS 2 loomweights: C 38 (7401.1), C 208 (7403.2). MAMMALS 2 Bos taurus; 2 ovicaprids (7403.2).

ARCHITECTURE, STRATIGRAPHY, AND HOUSEHOLD ANALYSIS

MARINE SHELLS (11 MNI) 6 Patella (3 medium/large); 2 Monodonta (2 open apex); Hexaplex (columella); 0/5 Charonia (2 apex, 2 much bored exterior; columella, lip; siphonal notch, 2 MNI) (7403.1–2).

25

LAND SNAILS (7 MNI) 4/1 Eobania (4 no color, 1 bleached white/clear, 5 MNI); 2 Rumina (2 have some gloss) (7403.2).

South of Rooms 2, 4, and 6 (Fig. 4, 6) The area to the south of Room 6 was excavated as part of Trenches 6400 (southern part) and 6300 (southern part) (Fig. 4). The surface lay at an elevation of +6.08 m. The excavation of the surface revealed rock tumble in the northeast corner of 6400 (southern part), which might have been wall collapse. Although the first pass (Locus 6404.1) stopped at the dividing line between Trenches 6300 and 6400, the second pass (Locus 6404.2) included the part of 6300 that fell outside Room 6. The bottom of the south wall of Room 6 was found in the third pass (Locus 6404.3) at approximately +5.60 m, providing a preserved height of ca. 1 m for that wall. Although the floor deposit of Room 6 lay at +6.24– 6.17 m, the south wall of Room 6 was found to be deeper and set well into the Minoan levels, unlike most of the walls of the complex, which had more shallow foundations. No distinct Hellenistic level was found. The pottery and small finds recovered from the excavation of this area are of mixed date (Minoan and Hellenistic). Most of the stone tools, although Minoan in date, were found intact (grinders GS 823 and GS 806 and hammerstone GS 594).

Tumble (6404.1–3) POTTERY Basin: P 2538 (6404.2).

CERAMIC OBJECTS 2 loomweights: C 156 (6404.2), C 201 (6404.1). STONES 2 grinders: GS 823 (6404.1), GS 806 (6404.2). 2 hammerstones: GS 824 (6404.1), GS 594 (6404.2). Handstone: GS 825 (6404.1). MAMMALS 3 inderterminate bone fragments (6404.3). MARINE SHELLS (44 MNI [3 Collected Dead]) 22 Patella (2 large); 9/2 Monodonta (2 open apex [1 ancient, 1 recent], 2 large, 10 MNI); 1/7 Hexaplex (complete, open body, no lip, L. 0.052 m; 1 columella/distal/lip; 3 distal/columella; 1 lip; waterworn, columella/distal, medium/large; 5 fresh MNI); 0/5 Charonia (3 body [1 has siphon lines], 1 lip [bored exterior and edges, collected dead], 2 distal/columella/siphon, 2 MNI); Conus (waterworn, slightly broken lip); Bittium (has lip); 2 Fasciolaria (broken lip; open body above mouth [ancient]); Arca (L, adult) (6404.1–3). Some of these were published in Mochlos IIA as coming from the fill over Rooms 1 and 2 (Soles 2008, 103).

2

Pottery Natalia Vogeikoff-Brogan

Methodology and Arrangement of the Catalog The present study is based on 128 vessels. The ceramic typology is organized by function, thus grouping together vessels made in different fabrics and techniques, both local and imported. Aside from local shapes, function and shape are described according to the terminology established by Susan Rotroff in the Athenian Agora excavation series (Rotroff 1997, 2006a). The catalog consists of pottery from floor assemblages to which matching shapes from other layers have been added (the addition of pottery from unstratified layers is always noted). Since the pottery found in the Beam-Press Complex was highly fragmentary and only partially restorable, the catalog would have been very limited if it included only vessels with full profiles. As a result, every vessel that preserved a distinguishable feature has been included. In addition to the pottery found in the floor deposits of the complex, the sherd material from the surface deposit excavated to the south of Room 8 (Locus 62/6300.1) is also included. As suggested above in

Chapter 1, this deposit seems to contain material from a dump created during Seager’s 1908 excavations of Building B.1, which lies to the south of the Beam-Press Complex. The Hellenistic pottery from the dump includes shapes (moldmade bowls III.37–III.40, III.45, III.46, III.48, and highnecked cups III.31–III.33) either not found in the floor deposits of the complex, or, if found, they are so small and fragmentary that they are probably the result of primary (or secondary) deposition (III.42–III.44). It seems likely that the Hellenistic pottery to the south of Room 8 originated from a Hellenistic building of different use sitting on top of the Minoan Building B.1 that “was swept away” by Seager (1909, 276, 281). The imported transport amphorae have been very helpful for the chronology of the floor deposits and the local pottery. More than 10 of the amphorae found in the Beam-Press Complex were identified as Koan by shape and petrographic analysis. Their shape compares well with Koan amphorae from Delos associated with the destruction of 69

28 b.c.

NATALIA VOGEIKOFF-BROGAN

on that island, as well as Koan amphorae from the Antikythera shipwreck (see the section below on transport amphorae), whose sinking has been dated between 80 and 50 b.c. (Empereur and Hesnard 1987, pl. 4:20; Rotroff 1994, 139). The fact that the floor deposits of the Beam-Press Complex did not yield a single piece of Eastern Sigillata A (ESA) pottery can be used as an argument ex silentio about the date of the building. Although produced as early as the second century b.c., the ESA ware did not become common in Crete until the Augustan period (Coldstream, Eiring, and Forster 2001, 139). At Mochlos, the so-called House of the Crocodile (see Intro.), whose abandonment is dated after 37 b.c., yielded several examples of ESA plates. The only other site in East Crete that is contemporary to the Beam-Press Complex, and whose pottery has been published in a preliminary report, is the Sanctuary of Hermes and Aphrodite at Myrtos Pyrgos (Fig. 2). Located west of Hierapytna and built directly on top of the LM IB settlement (Eiring 2000), Hellenistic Myrtos Pyrgos is a site with a short life span. The pottery, published by Jonas Eiring, has been dated to the late second/early first century b.c. by comparing certain shapes, like the everted-rim cup, with their Knossian cousins (Eiring 2000, 59). Scott Gallimore’s dissertation on Roman Hierapytna was completed in 2011, after the submission of this manuscript. Nevertheless, since there was a good amount of Hellenistic pottery in the earlier levels of the Roman buildings that Gallimore examined, his work is referenced in the catalog whenever possible. Although based on material from rescue excavations, Gallimore’s study is very important because it presents for the first time datable contexts of Hellenistic and Roman pottery from Hierapytna (Gallimore 2011). Since there is no other East Cretan site to refer to for stylistic comparisons, Knossian deposits of the Unexplored Mansion are also referenced when applicable (Sackett et al. 1992). These deposits, however, should be used with caution since they were destroyed by deep Roman foundations. According to Eiring, the dating of the later (Hellenistic) material from Knossos relies on internal typologies supported, to some extent, by references to material from Lyttos and Hagia Pelagia (Coldstream, Eiring, and Forster 2001, 91). The Sanctuary of Demeter at Knossos has yielded useful deposits of

the third and early second centuries b.c., but later material comes from mixed deposits (Coldstream, Eiring, and Forster 2001, 92). The material from the Little Palace Well is partly useful since most of its pottery dates earlier to the third quarter of the second century b.c., except for some surface material with full profiles (from a floor deposit?) that could date to the period 125–100 b.c. (Callaghan 1981, 37, 58). The so-called Cistern at Knossos yielded a rich group of material from a pottery kiln that has been dated to the last quarter of the second century b.c., but it has very few parallels with the pottery from the Beam-Press Complex (HomannWedeking 1950). Finally, comparisons with pottery from the Athenian Agora, except for cases of imported vessels, have been limited because it does not contribute to the understanding of local Cretan productions. By the Late Hellenistic period, any direct dependence on Attic pottery had ceased (Coldstream, Eiring, and Forster 2001, 91). There are a few pieces of pottery from the floor assemblages of the Beam-Press Complex that may date later than the second quarter of the first century b.c., the time of the building’s abandonment. Chytrai with collar rim and horizontal handles (III.98, III.99, III.101) find parallels in the lower fill of Knossian well MW/58–9, which John Hayes has dated to the late first century b.c. (Hayes 1971, 258, nos. 35, 36, fig. 10). Lopas III.116, with sloping wall and recurved handles, also compares with lopades in Knossian deposits of the second half of the first century b.c. (Sackett et al. 1992, pls. 103:21, 22, 104:13). Finally, the rim treatment of thin-walled beakers III.50 and III.51 resembles that of an Italian beaker found in late first-century b.c. deposits at Knossos (Sackett et al. 1992, 187, A2, no. 37, pl. 130; for local imitation, see Sackett et al. 1992, 182, A1, no. 12, pl. 126). This “inconsistency” may be explained in two ways: (1) securely dated pottery deposits of the first half of the first century b.c. are rare in Crete (Knossian deposits H 32, H 33, and H 35 from the Unexplored Mansion), thus the development of the pottery shapes during this period is insufficiently understood; and (2) Cretan pottery of the Hellenistic period is largely regional, thus it should not be treated as one entity with Knossos as the flagship. The term “local pottery” is used broadly here and refers to pottery produced in the area of East Crete. The majority of the pottery found at Mochlos is

POTTERY

also encountered at another site on the south coast: the Sanctuary of Hermes and Aphrodite at Myrtos Pyrgos (Eiring 2000; Vogeikoff 2000). Macroscopically, the fabric is soft and powdery, with its color varying from beige to light green (Munsell 10YR 8/2). For convenience, the fabric was named East Cretan Cream Ware (ECCW) because it does not appear in the central or western parts of Crete. The petrographic analysis of samples from transport amphorae made of ECCW shows a very fine clay tempered with medium or coarse sand composed mainly of weathered volcanic rock fragments and amphibolite fragments. This composition fits well with the geology of the Myrtos Valley region to the northwest of Myrtos Pyrgos (App. A; see also Fig. 2). Based on ethnographic data, which place the distance of the workshops to the clay and temper resources at a distance less than 7 km, especially when referring to subsistence agricultural communities, it is likely that ECCW was manufactured not far from its clay sources (Arnold 1985, 39–57). The most likely candidate for the location of the workshops is ancient Hierapytna, located a few kilometers away from the Myrtos Valley. Hierapytna, which had expanded to the north coast during the late second century b.c., could also have included Mochlos in her new territory (VogeikoffBrogan 2004, 216–219). The connection between Hierapytna and the production of ECCW is explored further in the discussion of transport amphorae below. If most of the Mochlos tableware originated from Hierapytna, the cooking pots came from another nearby location. Thin-section analysis of a number of cooking pots from the Beam-Press Complex shows that they were made of the so-called Mirabello fabric located in the area of Gournia (Fig. 2). The tiles that covered the Beam-Press Complex, as well as other buildings on the island, are the only ceramic item known for sure to have been produced

29

locally, since they were made of phyllite fabric. Phyllite formations surround the entire coastal plain region of Mochlos (see the discussion about tiles in Ch. 4, below). The following catalog organizes the pottery by function into nine categories: vessels used for food service, for wine service, for drinking, for pouring wine or water, for pouring oil or other liquids, vessels used for storage, vessels used for food preparation, cooking wares, and lamps. Each shape type in the catalog is preceded by a general description of that type and its significance in the Beam-Press Complex. The description discusses the primary features of the shape, its decoration and technique of manufacture, possible variations, as well as comparanda and chronology. The information presented in the catalog entries that follow here and in subsequent chapters are presented in the following order: 1) Catalog number of the object in the volume (III.1–III.215); 2) Inventory number (prefaced by C, CA, Fe, G, GS, P, Pb, or S, according to the type of object); 3) Findspot, which is listed by room/locus/pail number (unless otherwise indicated, all loci are from Area E3); 4) Figure and plate numbers; 5) Object shape; 6) Portion of object preserved; 7) Applicable dimensions, given in meters; 8) Weight for stone implements and loomweights, given in kilograms; 9) For pottery entries, type of fabric, if specific (such as ECCW), or Munsell color and number; 10) Decoration, including color of slip, interior decoration, and exterior decoration.

I. Vessels for Food Service There are only two types of vessels dedicated to food service: the echinus bowl with or without triangular handle and the plate with projecting or beveled rim. The bowl with outturned rim is missing from the Beam-Press Complex, as are other

types of plates, like the plate with overhanging rim, which appears in contemporary Knossian deposits (Coldstream, Eiring, and Forster 2001, 104– 105, fig. 3.6:h–k).

30

NATALIA VOGEIKOFF-BROGAN

A. Echinus Bowl with or without a Triangular Handle The term “echinus bowl” typically describes a shallow or deeper bowl with ring base, curved side wall, incurved rim, and no handle (for terminology, see Rotroff 1997, 161, esp. n. 53). When it carries a handle, it is usually described as a one-handler (Coldstream, Eiring, and Forster 2001, 100). Here, since it is not clear which bowls carried handles and which did not, I have avoided the latter term, which is more specific. Although uniform at first glance, the echinus bowls display a number of differences between each other (Fig. 19). The diameter of the rim usually varies between ca. 0.11 and 0.15 m, although there are examples of larger (III.3) and smaller bowls (III.1). The average height is 0.06 m with the exception of bowl III.10, which appears to be rather shallow. The ratio between rim and base is usually 2:1. The rim can sometimes be sharply curved (III.6, III.14), creating an angle with the side wall on the interior. The foot is low with an oblique profile, and the resting surface is flat and occasionally has a slight slant (e.g., III.6). A number of bowls, like III.1, III.7, and III.9, preserve a triangular handle. Although the fragmentary condition of most of the bowls does not permit the restoration of a handle, the considerable number of broken handles found in the excavation indicates that the majority carried one handle. Almost all of them are partially slipped in black or reddish brown. The interior of the bowls is fully coated, while the exterior carries the slip only around the rim (dipping method). The slip is thin and flaky. Note that bowls III.11–III.17 were found during the excavations of the east and north balks of Room 2; it is likely that the best preserved of them belonged to the floor assemblage of the room. All of the Mochlos examples are made of the ECCW fabric (see App. A). The echinus bowl with one handle was not a common shape in the Hellenistic period. In Athens, one-handlers were popular in the fourth century, but they disappeared in the early third century b.c. (Rotroff 1997, 155– 156). In fact, they hardly are encountered outside East Crete, with one example from the Sanctuary of Demeter at Knossos and a few unpublished fragments from the Cistern at Knossos (Coldstream,

Eiring, and Forster 2001, 100, fig. 3.3:w). Echinus bowls with a triangular handle of the same fabric have been identified both in the small Hellenistic Sanctuary of Hermes and Aphrodite at Myrtos Pyrgos on the south coast of Crete (Eiring 2000, 58, pl. 29:b) and at Hierapytna in contexts of the late second/first half of the first century b.c. (Gallimore 2011, 187–188, nos. 18, 20–23, fig. 5.5). There are also at least two echinus bowls with one handle from Trypitos, found in deposits of the first half of the second century b.c. (unpublished MS 11431, 11422). I have included the echinus bowl with one handle under “vessels for food service” because of its incurved rim, which does not lend itself for pouring liquids into the mouth (Eiring 2000, 57 n. 13). The existence of the handle suggests the consumption of, possibly hot, food in liquid state (e.g., thick soups). The concentration (localization) of the shape in sites of East Crete might allude to local eating habits. The majority of the bowls with incurved rim were found in Room 2 of the Beam-Press Complex. III.1 (P 411; Room 2; 6606.1; Fig. 19; Pl. 4). Bowl with incurved rim. About half of rim and side wall. Horizontal pinched handle. Pres. h. 0.04; rim d. 0.105. TF: ECCW. Color of clay uneven. Traces of black slip on exterior. III.2 (P 2534; Room 4; 6501.6/6504.1; Fig. 19). Bowl with incurved rim. Several fragments (some joining) preserving part of rim and side wall. Pres. h. 0.045; est. rim d. 0.16. TF: ECCW. Traces of red slip on interior and exterior. III.3 (P 2872; Room 2; 6601.1/6606.1; Fig. 19). Bowl with incurved rim. About half of rim, and part of side wall. Pres. h. 0.044; est. rim d. 0.16. TF: ECCW. Traces of red slip around and below rim on exterior. III.4 (P 2873; Room 2; 6605.1; Fig. 19). Bowl with incurved rim. Three joining fragments preserving part of rim and side wall. Pres. h. 0.045; est. rim d. 0.13. TF: ECCW. Traces of black slip on interior, reddish brown around and below rim on exterior. III.5 (P 2884; Room 2; 6601.1/6606.1; Fig. 19). Bowl with incurved rim. Three joining fragments preserving part of rim and side wall. Pres. h. 0.035; est. rim d. 0.135. TF: ECCW. Traces of brown slip around and below the rim on exterior. III.6 (P 2897; Room 2; 6605.1/6606.1; Fig. 19; Pl. 4). Bowl with incurved rim. Full profile. One-quarter extant. H. 0.058; rim d. 0.13; est. base d. 0.06. TF: ECCW. Traces of brown slip on interior and exterior.

POTTERY

III.7 (P 2899; Room 2; 6606.1; Fig. 19). Bowl with incurved rim. Four joining fragments preserving part of rim, side wall, and one handle. Horizontal pinched handle. Pres. h. 0.028; est. rim d. 0.124. TF: ECCW. Traces of brown slip on interior and exterior. III.8 (P 2902; Room 2; 6604.1/6605.1/6606.1–3/ 6621.1; Fig. 19). Bowl with incurved rim. About half of rim, parts of side wall, and base. H. 0.06; rim d. 0.15; base d. 0.058. TF: ECCW. Red slip on interior, and around rim on exterior. III.9 (P 3204; Room 2; 6606.1/6600 North Balk; Fig. 19). Bowl with incurved rim. Two non-joining fragments preserving part of rim, and handle. Horizontal pinched handle. Pres. h. 0.03; est. rim d. 0.12. TF: ECCW. III.10 (P 3228; Room 2; 6701.2; Fig. 19). Bowl with incurved rim. Part of rim and side wall. Pres. h. 0.035; est. rim d. 0.15. TF: ECCW. Traces of brown slip on interior and around rim on exterior. III.11 (P 5716; Room 2; 6621.1; Fig. 19; Pl. 4). Bowl with incurved rim. Full profile. One-third extant. H. 0.062; est. rim d. 0.103; base d. 0.05. TF: ECCW. III.12 (P 5717; Room 2; 6621.1; Fig. 19). Bowl with incurved rim. Full profile. One-third extant. H.0.057; est. rim d. 0.13m; est. base d. 0.065. TF: ECCW.

31

III.13 (P 5718; Room 2; 6621.1/6600 North Balk/ 6604.1; Fig. 19). Bowl with incurved rim. Five nonjoin­ing fragments preserving part of rim, side wall, base, and part of handle. Horizontal pinched handle. Est. rim d. 0.115; est. base d. 0.05. TF: ECCW. III.14 (P 5720; Room 2; 6600 North Balk; Fig. 19) Bowl with incurved rim. Two non-joining fragments preserving part of rim and side wall. Pres. h. 0.045; est. rim d. 0.15. TF: ECCW. III.15 (P 5721; Room 2; 6621.1; Fig. 19) Bowl with incurved rim. Four joining fragments preserving base and side wall. Pres. h. 0.04; base d. 0.055. TF: ECCW. III.16 (P 5719; Room 2; 6600 North Balk; Fig. 19). Bowl with incurved rim. Three (two joining) fragments preserving part of rim and side wall. Pres. h. 0.05; est. rim d. 0.14. TF: ECCW. Traces of black slip on exterior, red on rim at exterior, rest of exterior black. III.17 (P 5723; Room 2; 6621.1/6600 North Balk; Fig. 19). Bowl with incurved rim. Two non-joining fragments preserving part of rim and base. Est. rim d. 0.11; est. base d. 0.06. TF: ECCW. Traces of black slip on interior and around rim on exterior.

B. Bowl with Outturned Rim The bowl with outturned rim and carinated body, known also as the “flaring rim bowl” in the Knossian publications (Coldstream, Eiring, and Forster 2001, 102), is one of the most common shapes of the Hellenistic period throughout the Mediterranean. It is not represented in the floor assemblages of the Beam-Press Complex. Fragments from a possible example (III.18), which carries a pair of grooves on the rim, were found in the so-called Seager’s dump (Locus 62/6300.1) located to south of Room 8. Bowls with similar treatment of the rim have been found in Knossian deposits dating to the second half of the first century b.c. (Coldstream,

Eiring, and Foster 2001, 102, fig. 3.5:r, s). Another possible example of a bowl with outturned rim is the large bowl (III.19) from Room 6. Both bowls are made of the ECCW. III.18 (P 3490; south of Room 8; 62/6300.1; Fig. 20). Bowl with outturned rim. Several fragments preserving part of rim, side wall, and base. Est. rim d. 0.17; est. base d. 0.09. TF: ECCW. Remains of red slip on interior. III.19 (P 2561; Room 6; 64/7412.1; Fig. 20). Bowl with outturned rim(?). Part of rim and side wall. Pres. h. 0.025; est. rim d. 0.28. TF: ECCW. Traces of black slip around rim.

C. Plate with Projecting or Beveled Rim This shape describes large plates with a rim diameter of ca. 0.20 m, all made in the East Cretan Cream Ware. The majority of them are shallow, with the exception of III.24 and III.26, which are deeper. The shape of the rim varies, projecting downward in III.20 and III.21, thickening out in III.26 and III.27, and projecting both upward and

downward in plates III.22, III.24, and III.25. In the case of III.25, the side wall creates a slight S curve. The plates carry red slip on both the interior and exterior. Because of their fragmentary state of preservation, it is not clear whether they were fully or partially slipped on the exterior. The closest parallels are four examples found in the Yiomelaki

32

NATALIA VOGEIKOFF-BROGAN

plot in Ierapetra, the excavation of which revealed the remains of a building probably destroyed by fire in the first half of the first century b.c. (Gallimore 2011, 168–174, 188, nos. 27–29, fig. 5.5, pl. 5.1); two more examples come from Deposits H 32 and H 33 (dated to 100–50 b.c. and 75–50 b.c., respectively) in the Unexplored Mansion at Knossos (Sackett et al. 1992, pl. 101:8, 17). Of interest is the fact that similar plates, made of different fabric, however, have been found in a well at the agora of Ephesos (Meriç 2002, pl. 2, nos. K 14–K 20); some of the plates date to the period of the use of the well (“Füllung A”), which is placed in the first century b.c., before 25 b.c. Another example comes from Tel Anafa, from a late second/early first century b.c. context. According to Kathleen Slane, the Tel Anapha plate is made of a fabric that is common in Late Hellenistic levels at Teos and Ephesos; it also forms a small but distinct group in Benghazi and Delos (Berlin and Slane 1997, 362–363, FW 520, pls. 31, 54). As for Crete, this shape “may be a form that arose around the time of the island’s conquest” (Gallimore 2011, 421). III.20 (P 2852; Room 4; 6510.3/6511.1; Fig. 20). Plate with projecting rim. Three fragments preserving part of rim and floor. Est. rim d. 0.20. TF: ECCW. Traces of red slip on interior and exterior.

III.21 (P 3239; Room 4; 6504.1; Fig. 20). Plate with projecting rim. Two joining fragments preserving part of rim and floor. Est. rim d. 0.20. TF: ECCW. Traces of red slip on interior and exterior. III.22 (P 2555; Room 6; 64/7412.1–2; Fig. 20). Plate with beveled rim. Four fragments (two joining) preserving part of rim and floor. Est. rim d. 0.20. TF: ECCW. III.23 (P 2885; Room 4; 6510.3; Fig. 20). Plate with beveled or projecting rim. Base and part of lower sidewall. Ring base with flat resting surface and slight nipple underneath. Pres. h. 0.02; est. base d. 0.065. TF: ECCW. III.24 (P 2893; Room 2; 6605.1/6606.1–2/6621.1; Fig. 20; Pl. 4). Plate with beveled rim. Several fragments (some joining) preserving part of rim, and side wall. Est. rim d. 0.20. TF: ECCW. Red slip on interior and exterior. III.25 (P 2717; Room 1; 7703.5; Fig. 20). Plate with beveled rim. Two fragments preserving part of rim and floor. Est. rim d. 0.19. TF: ECCW. Traces of red slip on interior and exterior. III.26 (P 3242; Room 2; 6604.1/6606.1/6621.1; Fig. 20). Plate with beveled rim. Several fragments (some joining) preserving part of rim, side wall, and base. Restored height: 0.055. Est. rim d. 0.21; base d. 0.06. TF: ECCW. Red slip on interior and exterior. III.27 (P 2891; Room 2; 6606.1–2/6621.1; Fig. 20). Plate with beveled rim. Three joining fragments preserving part of rim and side wall. Pres. h. 0.035; est. rim d. 0.18. TF: ECCW. Traces of brown slip on interior and exterior.

II. Vessels for Wine Service The Beam-Press Complex did not yield any pottery associated with mixing wine with water except for one possible krater. It could be that the industrial character of the complex did not encourage formal consumption of wine, an activity that would have been more appropriate in a

dedicated dining context or a religious environment. Although there is evidence for consumption of food in certain rooms (Rooms 4 and 6), only a few drinking vessels were found in the floor deposits (see the discussion below).

Krater(?) The closest parallel to vessel III.28 is an example from Hellenistic House A in Eleutherna, described as a lebes, which has been dated to the third century b.c. (Kalpaxis, Furtwängler, and Schnapp 1994, fig. 30:K 246). However, the extent of preservation of vessel III.28, which is made of a fabric macroscopically similar to the ECCW,

suggests that it must have been in use during the period before the destruction of the building. III.28. (P 5738; Room 2; 6621.1/6705.1/6600 North Balk; Fig. 20). Krater(?). Many fragments (some joining) preserving one-half of rim, part of side wall, and beginning of handle. Incurved rim, which thickens out, and angular shoulder. Est. rim d. 0.23. TF: fine, very close to ECCW (Munsell 10YR 8/3).

POTTERY

33

III. Drinking Vessels The floor deposits of the Beam-Press Complex did not yield any drinking vessels that were in use during the period prior to the abandonment of the building. The vessels described below either come from the area south of Room 8, or if they were found in the floor accumulations of the complex, they were so small and fragmentary that it is safer to consider them as the products of primary

(or secondary, related to the maintenance of the floors) refuse. Unlike the deposit from the Sanctuary of Hermes and Aphrodite at Myrtos Pyrgos that yielded many cups with everted rims together with the echinus bowls, the dearth of drinking cups in the Beam-Press Complex is indicative of its non-religious/non-domestic function.

A. Kantharos The fragments of vessel III.29 are identified as a kantharos, instead of as a cup with everted rim, because of the large diameter of the rim (ca. 0.13 m), which must have required the support of a second handle (not preserved). The vessel was found south of Room 8 in Seager’s dump. The closest parallel, but with a taller rim, is a Late Hellenistic kantharos found in the Sanctuary of Hermes and Aphrodite at Myrtos Pyrgos (Eiring 2000, 56, pl. 28:b, no. 3). The description of its fabric (“soft pale buff fabric with a greenish yellow self-slip”) suits that of the ECCW. The thumb-rest on the handle of fragment III.30 suggests that it originally belonged to a kantharos. Although it comes from the floor deposit of Room 4, its fragmentary state suggests that it does not belong to the actual refuse of the room, but it either

formed part of the room’s primary refuse or was brought in from somewhere during the construction of the Beam-Press Complex. Since Room 4 may have functioned as a living quarter (see Ch. 1), the presence of a kantharos in the primary refuse of the room is not odd. Note that its fabric is coarse and red, which is macroscopically different from the ECCW. III.29 (P 3497; south of Room 8; 62/6300.1; Fig. 20). Kantharos? Part of rim, shoulder and handle. Strap handle rising above rim. Pres. h. 0.07; est. rim d. 0.18. TF: ECCW. III.30 (P 2854; Room 4; 6511.1; Fig. 20). Kantharos. Part of handle. Strap handle in right angle with thumb-rest. Pres. h. (rim) 0.04. TF: coarse, reddish yellow 5YR 6/8.

B. Kantharos or High-Necked Jugs/Cups Fragments of five vessels are included in this category, all of which have a high neck with a pair of incisions below the straight rim (incisions are missing from III.33). One example (III.32) preserves a vertical handle of oval section. It is unclear whether these vessels carried one or two handles; in the first case, they should be described as high-necked cups/jugs and as kantharoi in the second case. Of the five Mochlos examples, III.32 certainly subscribes to the ECCW. The other four are made of a fine reddish-yellow or pale brown fabric (7.5YR 7/6 or 7.5YR 8/4, 8/6). All cups are slipped on the interior and partially slipped on the exterior. None of

the cups comes from a secure context. Cups III.31– III.33 were found south of Room 8. The remaining two (III.34, III.35) were found in Room 2, but only III.35 comes from the floor deposit. The fact that it is a single and fine fragment suggests that it was probably residual. In 2009, during a cleaning operation in Trench 8000, the largest part of a vessel (III.36), similar to the ones above, was unearthed from a Hellenistic context. It has been included here because of its full profile. The shape is encountered only in East Crete. In addition to the Mochlos examples, the Sanctuary of Hermes and Aphrodite at Myrtos Pyrgos has yielded at least one

34

NATALIA VOGEIKOFF-BROGAN

kantharos (Eiring 2000, pl. 28:b, no. 3) whose shape corresponds to the Mochlos one; another example was recently found at Livari Cave, near Siteia (Y. Papadatos, pers. comm.). The upper wall in the Livari kantharos, however, slants inward, unlike the straight profile of the Mochlos vessels. III.31 (P 3485; south of Room 8; 62/6300.1; Fig. 21). High-necked cup. Four fragments (two joining) preserving part of rim, neck, and shoulder. Tall straight rim with pair of grooves on exterior. Pres. h. 0.059; est. rim d. 0.105. TF: fine, reddish yellow 7.5YR 7/6. Red slip on interior and exterior, drip on shoulder. III.32 (P 3486; south of Room 8; 62/6300.1; Fig. 21). High-necked cup. Four joining fragments preserving part of rim, neck, shoulder, and handle. Straight rim with pair of grooves below, carinated shoulder, strap handle rising above rim. Pres. h. 0.056; Est. rim d. 0.10. TF: ECCW. Brown slip on interior and exterior. III.33 (P 5588; south of Room 8; 62/6300.1; Fig. 21). High-necked cup. Three fragments (two joining)

preserving part of rim, neck and shoulder. Tall straight rim, round shoulder. Pres. h. 0.045; est. rim d. 0.12. TF: fine, very pale brown 10YR 8/4. Reddish-brown slip on exterior, dark red slip on interior. III.34 (P 5726; Room 2; 6621.1/6600 North Balk; Fig. 21). High-necked cup. Five fragments (four joining) preserving part of rim, neck and body. Tall straight rim, slightly everted, round body. Pair of grooves below rim on exterior. Pres. h. 0.07; est. rim d. 0.114. TF: fine, pink 7.5 YR 8/4. Red slip on interior and exterior. III.35 (P 5747; Room 2; 6705.1; Fig. 21). Highnecked cup. Single fragment preserving part of rim and neck. Straight rim slightly everted. Pair of thin grooves below rim on exterior. Pres. h. 0.032; est. rim d. 0.12. TF: fine, reddish yellow 7.5YR 8/6. Red slip on interior and exterior. III.36 (P 6550; 8012.1; Fig. 21; Pl. 4). High-necked cup. Several joining fragments preserving base, largest part of body, neck, rim, and one handle. H. 0.105; rim d. 0.95; base d. 0.05. TF: fine, reddish yellow 7.5YR 7/6. Red slip on interior and partially on exterior.

C. Moldmade Bowl The moldmade bowl first appeared in Attic ceramic deposits in the third quarter of the third century b.c. By the last quarter of the century, moldmade bowls appeared in many places, including Corinth, Argos, Delos, and Pergamon, only to become the main drinking vessel in the Mediterranean during the second and first centuries b.c. (Rotroff 1982, 6–10; see also Rotroff 2006b). Despite its wide production and distribution, the moldmade bowl is rare in Crete until the middle of the second century b.c., and even then it never became a popular drinking vessel (for moldmade bowls in Central Crete, see Egglezou 2005, 191– 193). The majority of the moldmade bowls imported to Crete were products of the Ionian workshop. The Attic examples are rather rare. There is evidence that Knossos produced its own moldmade bowls by the first century b.c. (Catling et al. 1981, 98, no. V.339; Homann-Wedeking 1950, 188–189; Egglezou 2005, 192). The majority of the Mochlos bowls are preserved in single fragments. Seven come from the area south of Room 8 (III.37–III.40, III.45, III.46, III.48), four pieces (III.42–III.44, III.47) were found in Rooms 1 and 2, and one (III.41) was discovered outside Room 1. The fragmentary preservation of the bowls found inside the complex

(due partly to the thinness of their side walls) suggests that they were part of the residual primary refuse of the rooms and were not in use at the time of the destruction of the complex. All of the bowls are products of the Ionian workshops. Bowl III.40 is the product of the workshop of a potter named Philon, as the signature at the bottom of the bowl indicates (Laumonier 1977, 247–273). The split leaf rosettes of bowl III.46 also recall decoration common in Philon’s workshop (Laumonier 1977, pl. 128:1214, 1300). In Delos, Philon’s moldmade bowls have been found in contexts of the late second to early first century b.c. (Zapheiropoulou and Hatzidakis 1994, 240). In addition to products of Philon’s workshop, bowls III.41, III.43, and III.45 also seem to be imports, to judge from their clay. Bowls III.37– III.39, III.42, III.44, III.47, and III.48, however, may be the products of Cretan workshops. III.37 (P 1454; south of Room 8; 62/6300.1; Fig. 21). Moldmade bowl. Three fragments (two joining) preserving part of side wall. Imbricated ferns. Pres. h. 0.06. TF: fine, pale brown 10YR 6/3. Black slip, mottled. III.38 (P 1455; south of Room 8; 62/6300.1; Fig. 21). Moldmade bowl. Four sherds (three joining) preserving part of rim and side wall. Relief rosettes with 11 petals each. Pres. h. 0.035; est. rim d. 0.12. TF: fine, light brown 7.5YR 6/4. Black slip, mottled at places.

POTTERY

III.39 (P 1456; south of Room 8; 62/6300.1; Fig. 21). Moldmade bowl. Part of rim and side wall. Frieze of relief rosettes below rim. Pres. h. 0.034; est. rim d. 0.12. TF: fine, light brown 7.5YR 6/4. Black slip, mottled. III.40 (P 1457; south of Room 8; 62/6300.1; Fig. 21). Moldmade bowl. Two joining fragments preserving part of base and lower side wall. Imbricated ferns. Fabricant’s name at bottom: ΦΙΛΩΝΟΣ. Pres. h. 0.02; base d. 0.047. TF: fine, reddish yellow 5YR 6/6. Red slip on interior and exterior. III.41 (P 2729; outside Room 1; 7708.5; Fig. 21). Moldmade bowl. Part of rim and upper side wall. Frieze of relief rosettes below rim. Pres. h. 0.042; est. rim d. 0.14. TF: fine, reddish yellow 7.5YR 7/8. III.42 (P 8593; Room 2; 6606.1; Fig. 21). Moldmade bowl. Part of upper side wall, near rim. Pair of horizontal lines and part of leaf. Dims. 0.02 x 0.015. TF: fine, light brown 7.5YR 6/4. III.43 (P 3201; Room 1; 7607.1; Fig. 21). Moldmade bowl. Two joining fragments preserving part of rim and side wall. Unidentified row of patterning at junction of rim with side wall. Imbricated ferns on side wall. Pres.

35

h. 0.04; est. rim d. 0.125. TF: fine, light brownish gray 2.5YR 6/2. Traces of black slip on interior and exterior. III.44 (P 3237; Room 2; 6605.1/6621.1; Fig. 21). Moldmade bowl. Five fragments preserving part of side wall. Pres. h. 0.03. TF: fine, yellow 10YR 7/6. Brown slip on interior and exterior. III.45 (P 3491; south of Room 8; 62/6300.1; Fig. 21). Moldmade bowl. Part of side wall. Imbricated ferns. Pres. h. 0.035. TF: fine, red 2.5YR 5/8. Red slip. III.46 (P 3493; south of Room 8; 62/6300.1; Fig. 21). Moldmade bowl. Part of rim and side wall. Frieze with relief rosettes below rim. Pres. h. 0.045; est. rim d. 0.13. TF: fine, red 2.5YR 5/8. Black (red at places) slip. III.47 (P 5744; Room 1; 7608.1; Fig. 21). Moldmade bowl. Six fragments preserving part of side wall. Vine decoration(?) above and ferns below. Pres. h. 0.045. TF: fine, very pale brown 10YR 7/4. Red slip on exterior, brown on interior. III.48 (P 3492; south of Room 8; 62/6300.1; Fig. 21). Moldmade bowl. Part of base and lower side wall. Imbricated ferns. Pres. h. 0.01; base d. 0.035. TF: fine, light brown 7.5YR 6/4. Black slip.

D. Beaker Fragments from three thin-walled beakers were found in the floor deposits of the Beam-Press Complex. The rim of III.50 resembles that of an Italian beaker found in late first-century b.c. deposits at Knossos (Sackett et al. 1992, 187, A2, no. 37, pl. 130; for local imitation, 182, A1, no. 12, pl. 126). Beaker III.51 shares the same rim but the body is globular. Both beakers should be restored with a flat base. The Mochlos beakers do not need to date so late. Imported Italian thin-walled pottery of a form that dates in the early first century b.c. appears in Knossian deposit H 33, which could be dated before 50 b.c. (Callaghan suggests with much hesitation a post-Metellan date for the deposit; Sackett et al. 1992, 124, H33, no. 4, pl. 102.) The swelling rim

and the round body of III.49, as well as its gray fabric, are common in thin walled pottery from Italian contexts of the late second/early first century b.c. (Marabini Moevs 1973, 59–61, pls. 4:35, 5:47, 52). III.49 (P 2896; Room 2; 6604.1/6605.1/6606.1; Fig. 21). Beaker. Five fragments (some joining) preserving part of rim, lower side wall, and base. Est. H. 0.08; est. rim d. 0.08. TF: reddish yellow 5YR 7/8 with dark gray core. III.50 (P 2556; Room 6; 64/7412.2; Fig. 21). Beaker? Part of rim and upper side wall. Pres. h. 0.02; est. rim d. 0.08. TF: fine, reddish yellow 5YR 6/8. III.51 (P 2881; Room 4; 6505.1/6512.1; Fig. 21). Beaker? Five fragments preserving part of rim and side wall. Est. rim d. 0.09. TF: gray.

IV. Vessels for Pouring Wine or Water A. Jug Jug III.52 (P 324) finds close parallels in a number of jugs recovered at the Sanctuary of Hermes

and Aphrodite at Myrtos Pyrgos (Eiring 2000, 58, pl. 30:a, nos. 1–4). In addition to being made in

36

NATALIA VOGEIKOFF-BROGAN

East Cretan Cream Ware, the jugs found within the Beam-Press Complex display a thickened outturned rim, occasionally dipped in a red or brown slip (III.52), with a sharp ridge below the rim. The Myrtos Pyrgos examples have been identified as hydriai based on a body sherd decorated with a painted volute that recalls the decoration of the Hadra hydriai (Eiring 2000, pl. 30:a, no. 5). The thickened rim and the ridge below, however, in both the Mochlos and the Myrtos Pyrgos examples, are not typical of hydriai. Moreover, unlike Hadra hydriai, which carry a thin strap handle, jug III.52 displays a handle of triangular section. In addition to the examples from Myrtos Pyrgos, three sherd rims from similar jugs were also found in rescue excavations at Ierapetra (Gallimore 2011, 194, 196, nos. 44–46, fig. 5.6; Gallimore, however, following Eiring, has identified them as hydriai). Jug III.53, also made in ECCW, is a large water jug with the characteristic ridge below the rim and an unusually tall and slender neck; its preserved height suggests that III.53 would have been as tall as a transport amphora (probably taller than 0.50 m). The high arched handle of both jugs, as well as their tall neck, sets them apart from other jugs in Crete (Hadzi-Vallianou 2000, pls. 51:b, c, 52:30; Egglezou 2005, pls. 32:b, nos. 598–600, 93:b, nos. 588–590). Jug III.54, from south of Room 8, must also be contemporary with jugs III.52 and III.53.

Jug III.56 is also without parallels. It is short and round, with a continuous profile, a shallow groove running around the rim, and a pair of grooves on the shoulder. Jug III.56, together with filter jug III.64 (discussed below), may antedate the occupation of the East Unit of the Beam-Press Complex since they were found built into the south wall of Room 1. III.52 (P 324; Room 2; 6605.1/6606.1/6701.2; Fig. 22; Pl. 4). Water jug. Several fragments preserving almost half of rim and neck, part of body, and handle. Pres. h. 0.16; est. rim d. 0.10. TF: ECCW. Bright red slip on interior and exterior of rim, pinkish slip on some of the body sherds. III.53 (P 2682; Room 1; 7601.3/7607.3/7703.3–5/ 7708.1–3; Fig. 22; Pl. 4). Water jug. Rim, neck, part of shoulder and handle. Pres. h. 0.23; rim d. 0.125. TF: ECCW. III.54 (P 5586; south of Room 8; 62/6300.1; Fig. 22). Jug. Single fragment preserving part of rim and neck. Est. rim d. 0.11. TF: fine, reddish yellow 5YR 6/6. III.55 (P 2615; Room 1; 7608.1; Fig. 22). Jug. Three fragments (joining?) preserving most of base. Rim d. 0.12 (est.). TF: ECCW. III.56 (P 4213; Room 1; 7604A fill; Fig. 22). Jug. Three joining fragments preserving rim, neck, part of shoulder and handle. Pres. h. 0.075; rim d. 0.068. TF: fine, 10YR 7/4. Black slip around rim with dripping on neck. III.57 (P 2725; Room 1; 7703.5; Fig. 22). Jug? Base and part of side wall. Pres. h. 0.028; est. base d. 0.06. TF: fine, 7.5YR 8/4 with light slip (10YR 8/3) on surface.

B. Lagynos The term “lagynos” describes a one-handled jug with long neck used to hold wine. Made in both coarse and fine painted wares, it was produced from the third to the first century b.c. (Rotroff 1997, 226–231). Coarse lagynoi are poorly represented in Cretan deposits, and they are missing altogether from the Hellenistic deposits of the Unexplored Mansion at Knossos. Neither Eiring nor Egglezou cite any coarse lagynoi in their respective overviews of Hellenistic pottery from Knossos and Central Crete (Coldstream, Eiring, and Forster 2001; Egglezou 2005, 221–223). Unlike Central Crete, the tombs at Viennos in southwest Crete have yielded a considerable number of coarse lagynoi with a variety of profiles (Markoulaki and Niniou-Kindeli 1994, 218–219, figs. 13, 14). Likewise, another lagynos comes from a funerary

context at Tarra (Tzanakaki 2000, 24, pl. 4:d). The only coarse lagynos currently known in East Crete is one from the site of Trypitos (MS 11873), which dates to the late third/early second century b.c. The Mochlos example (III.58), despite its fragmentary preservation, is interesting because its fabric bears all the characteristics of the ECCW (although it has not been tested through petrographic analysis), which suggests local production. The distribution of the coarse lagynos shows a preference for eastern Aegean sites. This points to a north–south trade route running from Egypt, the southern Levant, and Cyprus to the Dodecanese, the islands of the northeast Aegean (Chios, Lemnos), and the west coast of Asia Minor, with western offshoots to Delos and Athens, and other locations as far away as the Crimean Chersonesos (for findspots

POTTERY

of coarse lagynoi, see Rotroff 2006a, 82–84). Because of their many fabrics, Rotroff rightly suggests that different centers of manufacture were responsible for the coarse lagynoi (Rotroff 2006a, 83). Lagynoi also tend to appear in places under Ptolemaic control, like Cyprus, the southern Levant, and the military camp of Koroni in Attica (for the origin of the shape, see Berlin and Slane 1997, 42–43). Their presence in southwest Crete should not be surprising as there is strong evidence for Ptolemaic influence in West Crete attested in the funerary architecture of some of the tombs at Chania and Polyrrhenia (Karetsou and Andreadaki-Vlazaki 2000, 370–371, 393–395). Coarse lagynoi are rare at the opposite end of Crete, but because of the long Ptolemaic presence at Itanos, more may appear in future publications from East Crete. Epigraphic and literary evidence confirms the

37

presence of a Ptolemaic garrison at Itanos during the third and second centuries b.c., most likely at the request of the Itanians, who felt threatened by neighboring Praisos. Having a friendly harbor at the eastern end of Crete also suited the Ptolemies’ agenda to be the overlords of the waters between Alexandria, Cyprus, Rhodes, and Crete. The close relationship between Itanos and Egypt continued until the death of Ptolemy IV Philometor in 145 b.c., when the Egyptian troops finally withdrew from Itanos for unclear reasons (the Egyptian presence at Itanos and possibly in other parts of Crete during the Hellenistic period is discussed in depth in Spyridakis 1970, esp. 69–103). III.58 (P 5735; Room 2; 6600 North Balk/6605.1/ 6701.2). Coarse lagynos. Several fragments preserving half of rim, part of neck and side wall. Pres. h. 0.10; Rim d.: 0.046. TF: possibly ECCW.

V. Vessels for Pouring Oil or Other Liquids A. Lekythos All five lekythoi were found in Room 2. Of interest is large lekythos III.62, which must have been used for short-term storage. The shape for both small and large lekythoi can be restored as follows: a funnel-shaped mouth, a thin neck in continuous profile with the body, an ovoid body, a ring base, and a strap handle starting below the rim. This type of lekythos does not conform with either the bellshaped or the side-pouring lekythoi that were popular in Athens, Delos, Amorgos, and other places in the second century b.c. (Rotroff 1997, 169–170, figs. 69:1114–1116, 70:1120, 1122–1125; Hatzidakis 2000, pl. 70:c; Pappa 2000, pl. 60). The funnelshaped mouth of the Mochlos lekythoi is shallow and probably without a spout. The closest parallels are a number of lekythoi from the island of Amorgos, found in a late second-/early first-century b.c. deposit together with side-pouring lekythoi (Pappa 2000, pl. 61:b). Note that lekythoi III.59 and III.60 were found during the excavation of the north and south balks of Room 2, but they have been included here because of the affinity of their shape with lekythoi III.61 and III.62, which were retrieved from the floor assemblage of Room 2.

III.59 (P 5589; Room 2; 6600 North Balk.1A; Fig. 22). Lekythos. Several joining fragments preserving base and body up to the bottom of the neck. Pres. h. 0.135; base d: 0.06. TF: ECCW. Traces of black slip on exterior. III.60 (P 5725; Room 2; 6621.1; Fig. 22). Lekythos. Single fragment preserving part of mouth and beginning of handle. Pres. h. 0.027; est. rim d. 0.066. TF: fine, vary pale brown 10YR 8/4. III.61 (P 3232; Room 2; 6606.1; Fig. 22). Lekythos. Two non-joining fragments preserving part of funnelshaped mouth. Pres. h. 0.035; est. rim d. 0.06. TF: fine, pink 7.5 YR 8/4. Red slip on exterior and around rim on interior. III.62 (P 5728; Room 2; 6605.1/6606.1/6621.1; Fig. 22). Lekythos. Several fragments (many joining) preserving part of mouth, neck, shoulder, body, and base. H. 0.33; est. rim d. 0.075; est. base d. 0.10. TF: ECCW. III.63 (P 2901; Room 2; 6604.1; Fig. 22). Lekythos. Two joining fragments preserving part of mouth, neck, and part of handle. Pres. h. 0.045; rim d. 0.038. TF: fine, pink 7.5YR 8/4. Red slip on exterior and around mouth on interior.

38

NATALIA VOGEIKOFF-BROGAN

B. Filter Jug The Beam-Press Complex yielded two filter jugs (III.64, III.65). A filter jug is a pitcher characterized by a strainer in the neck and a funnel-shaped mouth (Rotroff 1997, 180–182). The largest part of jug III.64, which is made of a fine but brittle and pale fabric, was found together with jug III.56 built into the north facade of the south wall of Room 1; smaller fragments of III.64 were also recovered from the floor deposit of Room 2. The evidence suggests that both III.56 and III.64 do not belong to the period of use of the East Unit of the complex, but were part of the building material brought in for the construction of the walls and floors of

the building. Filter jug III.65, which comes from Room 6, is preserved very fragmentarily; its hard gray fabric suggests a non-Cretan origin. III.64 (P 4214; Rooms 1 and 2; 7604A/6606.1; Fig. 23; Pl. 5). Filter Jug. Many joining fragments preserving one-quarter of mouth, one-third of body, part of spout, and handle. Funnel-shaped mouth, strainer with five holes, ribbed round body, strap handle. Pres. h. 0.125; rim d. 0.075. TF: fine brittle, very pale brown 10YR 8/4. Traces of black slip on exterior. III.65 (P 4667; Room 6; 64/7413.1; Fig. 23). Filter jug. Two joining fragments preserving strainer with six holes. TF: gray with hard surface.

C. Situla There is no close parallel for vessel III.66, which has been called a situla because of its basket handle (Rotroff 1997, 134–135). The small spout just below the rim suggests a function related to the subsidence of heavier liquids and the collection of lighter ones, like oil, through the spout. The fragments of III.66 were not found inside the BeamPress Complex but were spread out in the area to the east of Rooms 1 and 3 (Trench 7800) in a locus that contained roof tiles (7805.1) and obscure architectural features. The macroscopic examination

of the fabric suggests affinities with ECCW. Since Room 2 yielded a number of vessels associated with oil storage and pouring (III.59–III.63), it is possible that situla III.66 may originally have belonged to the oil apparatus of the room. III.66 (P 2152; east of Rooms 1, 2, and 3; 7805.1; Fig. 23; Pl. 5). Situla. Several fragments preserving threequarters of rim, large part of side wall, three-quarters of base, spout, and handle. Rim d. 0.21; base d. 0.12. TF: fine, very pale brown 7.5YR 7/4.

VI. Storage Vessels Transport Amphora The results of the thin-section petrographic analysis have confirmed that a considerable number of transport amphorae found in the Beam-Press Complex, as well as in the Sanctuary of Hermes and Aphrodite at Myrtos Pyrgos, were made in the East Cretan Cream Ware (see App. A). At least, four types of amphorae were produced in this fabric. The shape of Type 1 is characterized by a folded, triangular rim, short cylindrical neck, ovoid body, piriform toe, and short elliptical handles, slightly arched (III.67–III.69). The average height is

0.65 m. Although fragmentarily preserved, Type 2 appears to imitate Rhodian amphorae to judge from the right angle and the round section of handle III.72; in the case of example III.73, the handle is also arched. Most importantly, Type 2 amphorae were stamped (III.71, III.72). Handle III.71 is stamped with a bee, a device used by the polis of Hierapytna to strike her coins. Note that none of Type 2 examples comes from a secure context (i.e., floor assemblages), but they have been included in the discussion of Cretan amphora production

POTTERY

because they add valuable information to the subject. Type 3 appears to imitate the rim and the double-barreled handles of the Koan amphorae; the fragmentary preservation of III.74 does not allow us to confirm whether this type also imitated the shape of the body and toe of the Koan amphorae. Finally, Type 4 is represented by one example (III.75) that displays an outturned rim with straight end, long neck, and strap handles ridged in the middle. The refiring tests for all four types of ECCW amphorae turned out cream colored because of the use of calcium-rich clays (see App. A). Amphora III.76 belongs to a different potting tradition. It bears some resemblance to the AC3 amphorae of Antigone Marangou’s typology (Marangou-Lerat 1995, 82–84, fig. 68, esp. nos. A 123–A 125, one of which has been found in the Early Roman cemetery of nearby Lato), especially in the treatment of the toe and the handles. The AC3 amphora is usually encountered in secondcentury a.d. contexts, although Hayes has dated an example from Knossos to the Augustan period. Amphora III.76 was found on an upper (surface) layer of Room 2 and should postdate the destruction of the Beam-Press Complex; it is included here because the petrographic analysis confirmed that it belongs to the ECCW, thus adding information to the life span of the fabric. The provisional archaeochemical analysis of one of the amphorae made in ECCW (III.74) appears to have identified trace amounts of compounds associated with wine (tartaric acid), although the weak strength of the chromatogram peaks shows that amphora III.74 may have stored other liquids, but not oil, during its life (see App. C). These results need to be confirmed with a larger scale analysis of samples from amphorae in the ECCW, but this is a first indication that the ECCW transport amphorae were made to carry wine (in case there is any skepticism about their contents). Moreover, as mentioned above, the shape of amphora III.74 intentionally imitates that of the Koan amphorae. The Beam-Press Complex also yielded a fragment from an amphora with a rounded rim and a double-barreled handle (III.77), the product of a workshop in Central Crete, as indicated by its sand-tempered fabric, which is encountered in amphorae from Knossos. The petrographic analysis has shown that the sand and the fossiliferous clay

39

of this fabric is consistent with the geology of the Knossos area (App. A; see also Eiring, Boileau, and Whitbread 2002). In addition to 11 Cretan amphorae, 12 examples of Koan amphorae (III.78–III.89) have been identified on the basis of stylistic analysis. No complete Koan amphora is preserved. The best-preserved amphora (III.79) displays an outturned round rim, a long neck with the characteristic offset at the junction of neck with shoulder, a steep shoulder, and long double-barreled and arched handles. Although the base is missing, a piriform toe may be restored, as the other examples indicate (III.82– III.84). The preserved height is 0.65 m, the maximum diameter 0.35 m, and the height of the handles 0.26 m. The tall and lean shape of III.79 can be compared with one of the Koan amphorae from the Antikythera shipwreck (no. 6), as well as a Koan amphora from Delos. The maximum diameter of the Antikythera amphora is 0.38 m, while the height of the handles measures 0.20 m (Davidson Weinberg et al. 1965, 15). The shipwreck has been dated between 80 and 50 b.c. (Rotroff 1994, 139). The Delian example (TD 7947), whose handles are as tall as 0.20 m and whose maximum diameter is 0.40 m, comes from the House of the Seals, which was destroyed in 69 b.c. (Empereur and Hesnard 1987, pl. 4:20). In general, the tendency in the development of the Koan amphorae is toward taller and leaner shapes. The maximum diameter of Koan amphorae found in contexts of the middle of the second century b.c., and earlier, is bigger than 0.40 m, while that of amphorae from contexts of the end of the second/early first century b.c. is smaller than 0.40 m. Virginia Grace cites a Koan amphora found in an Augustan context at the Athenian Agora (P 11880) as having a maximum diameter of 0.32 m (Davidson Weinberg et al. 1965, 10 n. 17). One, however, should allow for some flexibility in this rigid scheme of development, especially when this has been based on a very small number of samples (a handwritten sequence of the Koan amphorae produced by Grace in 1962, in preparation for the publication of the Antikythera amphorae, lists fewer than 20 examples from the second and first centuries b.c.; see Virginia R. Grace Papers, American School of Classical Studies at Athens, Archives in the Blegen Library, folder 632). For example (and contrary to Grace’s

40

NATALIA VOGEIKOFF-BROGAN

scheme of development), the maximum diameter of a Koan amphora from an Augustan context at Knossos is 0.40 m and the height of the handles 0.20 m (Sackett et al. 1992, 183, 189, A2, no. 97, pl. 135). Moreover, context can sometimes be misleading since there is evidence that transport amphorae with a chronological difference up to 40 years were stored side by side in the same room (Finkielsztejn 2004, 282). Taking this all into consideration along with the overall view of the finds from the complex, III.79 should probably be placed in the first half of the first century b.c. It is difficult to comment on the shape of the rest of the Koan amphorae from the complex because of their fragmentary preservation. Amphora III.86 shares the same rim diameter and long handles with III.79; in addition, the shape of its lower body suggests a lean amphora. Amphora III.84 is different in several aspects from III.79: the rim is wider, the handles create a right angle instead of being arched, and the lower body is wider (cf. III.84 and III.86). Amphora III.83 also displays a wide lower body, which suggests a maximum diameter closer to 0.40 m. If the relative chronology of the Koan amphorae were to be based on the criterion of the maximum diameter, amphorae III.83 and III.84 should precede III.79 and III.86. The formation of the shoulder of amphora III.81 also suggests a type of Koan amphora different from III.79. Finally, III.87 differs from all the rest in the treatment of the rim, which is heavily folded. The toe is also less pronounced, but it is not clear whether this was intentional or if the toe is very worn. Compared to other types of stamped transport amphorae (e.g., Knidian and Rhodian), Koan amphorae remain poorly studied and published (with the exception of Georgopoulou 2005; Hein et al. 2008a). Neither their shape development nor the dating of their stamps is well defined. The stylistic analysis developed by Grace in 1965 remains our only published guide concerning the development of the shape (Davidson Weinberg et al. 1965, 10). For the Late Hellenistic period, there are only a dozen published Koan amphorae for comparison. Part of the problem is that the majority of the stamped Koan handles do not come from datable contexts. In addition, only a small percentage of Koan amphorae were stamped (Empereur 1982, 226–227; Georgopoulou 2005, 181). To make things more complicated, Maurice Picon and Jean-Yves Empereur have

shown that by the first century b.c., Koan-type amphorae were produced in a wide region of sites including the Bodrum and Halikarnassos peninsulas and Rhodes (Empereur and Picon 1986); Grace believed that Koan imitations were produced as early as the third century b.c. (see Davidson Weinberg et al. 1965, 10). More recently, Maria Briese has argued, on the basis of careful macroscopic examination, for Halikarnassos as a possible production center of Koan-type amphorae (Briese 2005). In the same line of thought, Hayes has suggested that a number of double-barreled Koan-type amphorae found at Paphos in late second-century b.c. contexts were produced locally (Hayes 1991, 86); confusion arises, however, because amphora OΔ 4978 from Well 21 (Hayes 1991, fig. 37:1) is listed both as Koan (Hayes 1991, 164) and a Koan imitation (Hayes 1991, 86, no. 16). The petrographic analysis of a number of stamped Koan amphora handles, mainly from the Athenian Agora, by Ian Whitbread has contributed significantly to the understanding of the production of Koan amphorae. For instance, in examining stamped Koan amphorae with double-barreled handles and Koan amphorae with single-barreled handles, Whitbread concluded that the same workshops produced both types of amphorae. Another problem that emerged from Whitbread’s study was the large diversity in the Koan fabrics: four classes in 25 samples from the Athenian Agora (Whitbread 1995, 98). The diversity in the Koan fabrics could be interpreted as evidence of several production centers, but it could also indicate different periods of production. More recently, a combined chemical and petrographic analysis of 45 samples from the site of Kardamaina (ancient Halasarna) at Kos has shown that several samples presented similarity with Whitbread’s Koan fabric class 3, providing further evidence that Halasarna was an important production center of Koan amphorae (Hein et al. 2008a, 1055–1057). The petrographic analysis of the Koan-type amphorae from Mochlos has produced similar results, with the majority of the samples (III.78, III.80– III.84, III.87, III.89) matching Whitbread’s Koan fabric class 3 and a few that match or are close to Koan fabric class 4 (III.85, III.88; and see App. A). Whitbread’s Koan fabric class 3 is based on the petrographic analysis of six double-barreled handles

POTTERY

from Athenian Agora contexts. One amphora (SS 12196) found in the building fill of the South Stoa dates before 150 b.c., and three other amphorae (SS 12700, SS 12703, SS 12975) in the building fill of the Middle Stoa date before 183 b.c. The chronology of the Halasarna samples (Halasarna fabrics I and II) that match Koan fabric class 3 presents an even wider range, spanning from the fifth century to the first century b.c. (Hein et al. 2008a, 1060). The data from the Athenian Agora, Halasarna, and Mochlos have shown that Koan fabric class 3 was used for several centuries, from the fifth to the first century b.c. Surprisingly, our best-preserved example (III.79) does not fit into any of Whitbread’s five classes of Koan fabrics. The Halasarna chemical and petrographic analysis also produced several loners (Hein et al. 2008a, 1052, 1060). This shows that Whitbread’s seminal work needs to be continued and enriched with a larger database of samples. The Beam-Press Complex yielded fragments from two Rhodian amphorae (III.90, III.91) whose fabrics match Whitbread’s Rhodian fabrics 1 and 2 (Whitbread 1995, 60–62). Their fragmentary state of preservation prevents further comments on shape and chronology. Finally, amphora III.92 cannot be attributed to any known types; in addition, the petrographic analysis was inconclusive concerning the geological origin of its fabric (see App. A). The thickened rim, the placement of the handles at the base of the neck, as well as the color and the hardness of the fabric, resemble a type of amphora (Hellenistic amphora 2) found in Benghazi (Riley 1988, 122, D 14, fig. 68). According to John Riley, this type of amphora is rare in Cyrenaica and “does not seem to have been based on any particular eastern Mediterranean type.” The fragment of the same type of amphora also appears in the Late Hellenistic deposits of Tel Anafa (Berlin and Slane 1997, 164, PW 538, pl. 68). EAST CRETE III.67 (P 2557; Room 1; 7607.1; Frontispiece; Fig. 23; Pl. 5). Transport amphora, Cretan. Nearly complete, restored from many fragments. H: 0.65; rim d. 0.11. TF: ECCW. III.68 (P 2558; Room 6; 64/7412.1; Fig. 23). Transport amphora, Cretan. Part of rim and neck. Folded, triangular rim. Pres. h. 0.037. TF: ECCW. III.69 (P 2683; Room 1; 7601.5/7703.4; Fig. 23; Pl. 5). Transport amphora, Cretan. Several joining fragments

41

preserving part of rim, neck, and stubs of two handles. Pres. h. 0.14; rim d. 0.104. TF: ECCW. III.70 (P 3531; north of Rooms 6, 7, and 8; 7401.1–2; Fig. 24). Transport amphora, Cretan. Two non-joining fragments preserving part of rim and side wall. TF: ECCW. III.71 (P 3507; north of Rooms 6, 7, and 8; 7403.2; Fig. 24). Transport amphora. Stamped handle. TF: ECCW. III.72 (P 3655; Room 6; 7405.1; Fig. 24). Transport amphora, Cretan. Part of handle. Remains of stamp? TF: ECCW. III.73 (P 4661; south of Rooms 7 and 8; 6301.1; Fig. 24). Transport amphora, Cretan. Single fragment preserving part of handle. TF: ECCW. III.74 (P 3244; Room 2; 6604.1/6605.1/6606.1; Fig. 24). Transport amphora, Cretan. Several fragments preserving two-thirds of rim, upper part of neck, part of body, and one double-barreled handle. Pres. h. 0.07; rim d. 0.012. TF: ECCW. III.75 (P 2716; Room 1; 7601.3–5; Fig. 24). Transport amphora, Cretan. Four joining fragments preserving part of rim, neck, and one handle. Est. rim d. 0.125. TF: ECCW. III.76 (P 5587; Room 2; 6604.1/6600 North Balk/ 6621.1/6701.2; Fig. 24). Transport amphora. Several fragments preserving part of body, beginning of handle, and toe. TF: ECCW. Off-white slip on exterior.

CENTRAL CRETE III.77 (P 2900; Room 5; 7510.1; Fig. 25). Transport amphora, Cretan. Part of rim, neck, and handle. Pres. h. 0.10; est. rim d. 0.12. TF: sand tempered, very pale brown 10YR 8/3. KOS III.78 (P 1542; Room 6; 64/7412.1–2; Fig. 25). Transport amphora, Koan. Many fragments preserving part of neck, shoulder, body, and one handle. Pres. h. (of handle) 0.20. TF: fine Koan Fabric 3. Pale yellow slip (2.5Y 7/4). III.79 (P 1543; Room 4; 6504.2/6511.1; Fig. 25; Pl. 6). Transport amphora, Koan? Many fragments preserving rim, neck, part of shoulder, and body, and two handles. Pres. h. 0.65; max. d. 0.35; rim d. 0.11. TF: fine, coarse silicate fabric (5YR 6/8 reddish yellow). Very pale brown slip (10YR 8/3). III.80 (P 2154; Room 6; 64/7412.1–2; Fig. 25). Transport amphora, Koan. About 20 fragments preserving part of body and handle. Double-barreled handle. TF: fine Koan Fabric 3. Slip (7.5YR 8/4). III.81 (P 2155; Room 6; 64/7412.2; Fig. 25). Transport amphora, Koan. About 100 fragments preserving part of shoulder, body, and toe. TF: fine Koan fabric 3. Pale yellow slip (2.5Y 8/2).

42

NATALIA VOGEIKOFF-BROGAN

III.82 (P 2114; Room 1; 7703.1; Fig. 26). Transport amphora, Koan. Fragment preserving part of lower body and toe. Piriform toe. Pres. h. 0.07. TF: fine Koan Fabric 3. III.83 (P 2153; Room 7; 7302.1; Fig. 26). Transport amphora, Koan. Several fragments preserving part of lower body and toe. Piriform toe. Pres. h. 0.30. TF: fine Koan Fabric 3. Pale brown slip (10YR 8/3). III.84 (P 2888; Room 2; 6604.1/6606.1/6621.1/6701.2; Fig. 26). Transport amphora, Koan. Several fragments (some joining) preserving part of rim, body, one handle, and toe. Est. rim d. 0.14. TF: fine Koan Fabric 3. Off-white slip on exterior. III.85 (P 2889; Room 2; 6604.1/6605.1/6606.1; Fig. 26). Transport amphora, Koan. Five fragments preserving part of rim and neck, handle, and toe (worn). Est. rim d. 0.16. TF: fine Koan Fabric 4. Off-white slip on exterior. III.86 (P 2911; Room 2; 6605.1/6606.1–2/6600 North Balk/6621.1/6701.2; Fig. 26). Transport amphora, Koan. Many fragments preserving part of rim, body, handle, and base. Rim d. 0.12. TF: fine Koan Fabric 3 or 4. Very pale brown slip (10YR 8/3). III.87 (P 2914; Room 2; 6604.1/6605.1/6606.1/6621.1; Fig. 26). Transport amphora, Koan. Several fragments preserving one-half of rim, part of lower body with toe. Est. rim d. 0.11. TF: fine Koan Fabric 3. Traces of offwhite slip on exterior.

III.88 (P 3159; Room 3; 8601.4–5, 8610.1; Fig. 27). Transport amphora, Koan? About twenty fragments preserving part of neck, shoulder, body and handle. TF: coarse, sandy micaceous; Koan Fabric 4(?). III.89 (P 3508; north of rooms 6, 7, and 8; 7401.1–3; Fig. 27). Transport amphora, Koan. Eight fragments preserving part of rim, neck, and two handles. Est. rim d. 0.12. TF: fine Koan Fabric 3. Pale brown slip (10YR 8/3).

RHODES III.90 (P 2535; Room 4; 6501.6/6504.1/6511.1; Fig. 27). Transport amphora, Rhodian. Many fragments preserving one handle, part of side wall, and part of base. TF: Rhodian Fabric 1. III.91 (P 2606; Room 1; 7607.1; Fig. 27). Transport amphora, Rhodian. Part of rim and neck. Rounded rim, neck flares out. Pres. h. 0.03; est. rim d. 0.11. TF: Rhodian Fabric 2. UNATTRIBUTED III.92 (P 2727 + P 2913; Room 2; 6604.1, 6605.1, 6600 North Balk, 6221.1, 7708.3; Fig. 27; Pl. 6). Transport amphora, unattributed. Several fragments preserving part of rim, body, and toe. H. 0.65, est. rim d. 0.14, max. d. 0.28. TF: fine silicate, 2.5YR 6/8 (light red) with 2.5YR 4/2 (weak red) core.

VII. Food Preparation Lekane The term “lekane” describes large bowls with two handles made of plain or coarse ware. Fragments from five lekanai, which could have been used for food preparation, either solid or liquid, are cataloged below. Although very little is preserved from lekane III.94, its fabric (coarse with large inclusions) and handle shape are identical to that of III.93. The closest parallel for III.93 is a lekane from Paphos from a context of the middle (or slightly later) of the second century b.c. (Hayes 1991, 156–157, fig. LVI, no. 19). Vessel III.95 is included under lekanai even though its fine fabric suggests that it could have been used for serving as well. Two lekanai (III.96, III.97) carry traces of black or red slip on the rim or the exterior. III.93 (P 2122; Room 4; 6510.2–3/6511.1/6512.1; Fig. 28; Pl. 6). Lekane. Half preserved. Missing lower body and largest part of base. Outturned rim, flat on top, round body tapering in toward bottom, two horizontal

handles ridged in the middle, flat base. Pres. h. 0.24; rim d. 0.38. TF: coarse, reddish yellow 2.5YR 5/6. III.94 (P 5495; Room 6; 7310.1; Fig. 28). Lekane. Six fragments preserving handle and part of side wall. Wide horizontal handle grooved in the middle. TF: coarse, 5YR 6/8. III.95 (P 5736; Room 2; 6605.1/6606.1/6621.1/6600 North Balk/6700.1/6701.2; Fig. 28). Lekane. Several fragments preserving part of rim, side wall, and base. Outturned rim, flat on top, ring base. Est. rim d. 0.19; est. base d. 0.13. TF: ECCW. III.96 (P 5739; Room 2; 6621.1/6605.1; Fig. 28). Lekane. Three non-joining fragments preserving part of rim and side wall. Flaring rim, beveled at edge. Est. rim d. 0. 32. TF: coarse, reddish yellow 5YR 7/6. Traces of red slip around rim. III.97 (P 5740; Room 2; 6621.1). Lekane. Several joining fragments preserving part of rim, side wall, and one handle. Rounded outturned rim, horizontal handle. Est. rim d. 0.19. TF: coarse, reddish yellow 5YR 6/6. Traces of black slip (red at places) on exterior.

POTTERY

43

VIII. Cooking Ware The Beam-Press Complex yielded fragments from several cooking vessels, which included chytrai, lopades, and lids. Although three types of chytrai are distinguished based on their shape, the petrographic analysis has not discerned any significant difference in their fabrics (App. B). In fact, the majority of the cooking ware retrieved from the complex was made of the so-called Mirabello fabric, which is characterized by granite/granodiorite inclusions. The clay sources for this fabric should be located on the isthmus of Ierapetra near the area of Gournia (Fig. 2), where the only deposits of the main inclusion, a granitic-dioritic bedrock, have been found (Barnard and Brogan 2003, 7 n. 17). There are examples of Mirabello pottery in almost all phases of the Minoan settlement of Mochlos, especially in the Early Minoan (EM) and Middle Minoan (MM) periods. Pottery with granodiorite inclusions also appears in the Archaic deposits of Azoria (Haggis et al. 2007, 277 n. 83). As with the ECCW, we see continuity in the use of the same clay and temper sources throughout all periods. Moreover, it is almost certain that the area between Gournia and Istron had already passed into Hierapytna’s political control by the second century b.c.

(Hayden 1995, 94–96). If the eastern half of the Mirabello Bay, including Mochlos (as argued in Ch. 5) and the area around Gournia, were part of Hierapytna’s extended polity at this time, then the granodiorite cooking pots from Mochlos should not be considered “imports,” as they were in previous periods. This change has implications regarding the status of the regional economy, which is no longer insulated but fluid among the various regions of the growing polity. The chytrai are organized into three types based on the shape of their rim. Since they are all made of Mirabello fabric, they could be products of one workshop, or more likely the products of several workshops that were located in the vicinity of Gournia and Istron and had access to the same granodiorite outcrops. The vessels are thin-walled to provide better conduction of the heat from the exterior to the interior (App. B). What is interesting is the large size of the majority of the chytrai (III.106, III.114); in the case of fragmentarily preserved chytrai (e.g., III.99–III.101), we can infer the size of the chytra from the large rim diameter (0.25–0.30 m). The larger size of the chytrai is likely related to the industrial character of the Beam-Press Complex.

A. Chytra with Collar Rim, Lidded The term “chytra” is used to describe a deep, usually globular vessel with a flanged or plain rim, a round base, and two handles. The chytrai with collar rim are characterized by a straight (III.98, III.100, III.101) or concave (III.99, III.104) rim with flat top. Although no complete example is preserved, the body appears to be globular (III.98, III.101). Handles can be horizontal (III.98, III.100, III.101) or vertical (III.105). Chytrai III.98, III.99, and III.101 compare with Athenian Agora Forms 6 and 7, which are represented by a few (possibly imported) examples coming from deposits dating as early as the second half of the second century b.c., and as late as the end of the first century b.c. (Rotroff 2006a, 174, nos. 604– 606, figs. 76, 77, pl. 65). In Crete, chytrai with collar rim and horizontal handles are known from the

lower fill of Knossian well MW/58–9, which Hayes has dated to the late first century b.c. (Hayes 1971, 258, nos. 35–36, fig. 10); the Mochlos chytrai, however, suggest that the shape appeared earlier. Chytra III.100, with its P-shaped handle attached to the rim, finds its exact parallel in an imported chytra found in a Late Hellenistic (150–86 b.c.) context at the Athenian Agora (Rotroff 2006a, 177, 309– 310, no. 620, fig. 79, pl. 66). Another variation in the treatment of the handles appears on chytra III.105, which preserves a rare double-barreled vertical handle. This is also seen on an imported chytra with a double handle from a late second-century b.c. context in the Athenian Agora (Rotroff 2006a, 172, 306, no. 591, fig. 74, pl. 63).

44

NATALIA VOGEIKOFF-BROGAN

Chytra III.106 from Room 1 stands out from the rest of the chytrai with collar rim. Found in many fragments, it has been restored to a height of 0.33 m, without its bottom. Unlike the other chytrai, the rim slopes outward and lacks the usual inner flange. The vertical strap handles recall those of chytrai from Knossos and Trypiti on the southcentral coast of Crete, which broadly date to the second century b.c.; the bodies of those chytrai are different, however (Egglezou 2005, pl. 63:396, 530). We encounter similar handles in chytra III.108, which belongs to another form (see below). Chytra III.106 is most likely a hybrid between the chytrai with collar rim and inner flange and the chytrai with an inverted and flangeless rim (for the introduction of chytrai without inner flange, see below). In conclusion, the chytra with the collar rim must have been a long-lived type in the eastern Mediterranean, making its first appearance in the second half of the second century b.c., to judge from the imports in the Athenian Agora, and lasting throughout the first century b.c. III.98 (P 364; Room 2; 6604.1/6605.1/6621.1; Fig. 29; Pl. 6). Chytra with collar rim. Many fragments preserving three-quarters of rim, part of shoulder, one handle, side wall, and base. Rim d. 0.28. TF: Mirabello (red 2.5YR 4/8).

III.99 (P 2536; Room 4; 6501.6/6504.2/6511.2; Fig. 29). Chytra with collar rim. Several fragments preserving part of rim, carination between shoulder and body, and part of base. Est. rim d. 0.24. TF: Mirabello. III.100 (P 2150; Room 4; 6504.1/6511.2/6510.3; Fig. 29; Pl. 6). Chytra with collar rim. About 35 fragments preserving part of rim, side wall, one handle, and part of second handle. Est. rim d. 0.30. TF: Mirabello (light red 2.5YR 6/8). III.101 (P 2850; Room 4; 6504.1–2; Fig. 29). Chytra with collar rim. Several fragments (some joining) preserving part of rim, side wall, and handle. Pres. h. 0.11; est. d. 0.25. TF: Mirabello (red 2.5YR 5/8). III.102 (P 2733; Room 1; 7703.4–6; Fig. 29). Chytra with collar rim. Several fragments (some joining) preserving part of rim and side wall. Pres. h. 0.04; est. rim d. 0.24. TF: Mirabello (red 2.5YR 4/6). III.103 (P 5742; Room 1; 7703.3). Chytra with collar rim. Five fragments preserving part of rim and side wall. Tall rim 0.04. Pres. h. 0.047; est. rim d. 0.28. TF: coarse red 2.5 YR 6/8. III.104 (P 2569; Room 6; 64/7412.2; Fig. 29). Chytra with collar rim. Part of rim. Pres. h. 0.02; est. rim d. 0.10. TF: coarse, gray. Traces of burning. III.105 (P 5732; Room 2; 6605.1/6621.1; Fig. 29). Chytra with collar rim(?). Five fragments (some joining) from two handles. TF: Mirabello (red 2.5YR 4/8). III.106 (P 2681; Room 1; 7703.4–6; Fig. 29; Pl. 7). Chytra with collar rim. Restored from many fragments. Missing base. Pair of wheelmade grooves on shoulder. Pres. h. 0.33; rim. d. 0.27. TF: coarse, red 2.5YR 5/8.

B. Chytra with Inverted and Grooved Rim There is a second type of chytra recognizable at Mochlos. It shares with the previous type the collar rim but without an inner flange; instead, the upper end of the rim inclines inward. A pair of thin grooves runs the exterior of the rim (III.107– III.110). The lack of inner flange suggests that tight lids were no longer needed during cooking, thus permitting larger amounts of evaporation. Simmering has been connected with the cooking of legumes and beans, while lidded cook pots facilitate the preparation of stewed food (Joyner 2007, 190). Gallimore has published an example of a similar chytra from Hierapytna that was found in a context dating to 75–25 b.c. (Gallimore 2011, 228–229, no. 138, fig. 5.14). A few examples of similar chytrai have also been found on Delos and in the Black Sea region (Hatzidakis 2004b, 648, pls. 316:a, 320:6; Tolstikov and Zhuravlev 2004, 275, pl. 98:4).

Without petrographic analysis, the chytra with inverted and grooved rim would have been classified as a possible import because of its distribution pattern (Vogeikoff 2000, 73). Yet, the petrographic analysis has shown that it was made of Mirabello fabric, like the chytrai with collar rim. The distribution pattern of the chytra with inverted and grooved rim is similar to that of the plate with beveled rim, which is encountered in Ephesos and Tel Anafa (see above). Changes in the form of cooking pots have been associated with changes in diet, which could be the result of the infiltration of a new group of inhabitants (Joyner 2007; also on the issue of pottery and ethnic identity, see Peignard-Giros 2007). The fact that the chytra with inverted rim appears on Delos and in the Black Sea region, and the plate with beveled rim is also produced at Ephesos,

POTTERY

suggests that we are possibly dealing with shapes associated with the culinary needs of a larger ethnic group, most likely the Italians of the eastern Mediterranean. Prosopographic evidence suggests that several Roman gentes commercially active in the east during the early first century b.c. were also present on Crete (Baldwin Bowsky 2001, esp. 37). In fact, Rome had been involved in the political affairs of Crete for more than a century. One should not forget that Hierapytna had Rome on her side in the dispute with Itanos at the end of the second century b.c., a sign that the Romans were already focusing on Hierapytna’s land as a potential source of wealth (Stefanaki 2006, 308–309). Finally, the epigraphical documentation suggests that Hierapytna did not delay much in accepting Italian rule and influence after her conquest in 67 b.c., an indication that she was already familiar with the Italian way of living (Vogeikoff-Brogan 2012a; see also Baldwin Bowsky 1994, esp. 9–10). Whether the abandonment of the Beam-Press Complex is associated with the Roman conquest of

45

67 b.c. or is slightly later (within the first half of the first century b.c.), the new cooking pots could constitute evidence for Italian influence in the material culture of Crete as early as the first decades of the first century b.c. III.107 (P 2601; Room 1; 7606.1; Fig. 30). Chytra with inverted, grooved rim. Five fragments (three joining) preserving part of rim and side wall. Thick ridge at junction of rim with side wall. Pres. h. 0.03. Est. rim diam. 0.15. TF: coarse, red 2.5YR 5/8. III.108 (P 3186; Room 1; 7703.4–6; Fig. 30). Chytra with inverted and grooved rim. Many fragments (some joining) preserving part of rim, shoulder, body, and base. Est. rim d. 0.22. TF: Mirabello (reddish yellow 5YR 6/8). III.109 (P 3238; Room 2; 6606.1; Fig. 30). Chytra with inverted and grooved rim. Six (two joining) fragments preserving part of rim and shoulder. Pres. h. 0.04; est. rim d. 0.23. TF: Mirabello (red 2.5 YR 5/6). III.110 (P 5731; Room 2; 6621.1; Fig. 30). Chytra with inverted rim. Six fragments (five joining) preserving part of rim and shoulder, and stump of handle. Pres. h. 0.05; est. rim d. 0.24. TF: Mirabello (red 2.5 YR 5/6).

C. Chytra with Everted Rim The third type of chytra is characterized by a short neck and an everted rim, which forms an obtuse angle in the interior (III.111–III.113). The inner flange is missing here too. In addition to being thin-walled, chytra III.112 carries a thin groove around the rim. The shape of III.113, especially the ridge on the shoulder, recalls a chytra from the Hellenistic/Roman levels at Tarsus (Goldman, ed., 1950, fig. 191:G). The rim of large chytra III.114 (P 2800) differs from the rest because it carries two ridges on the exterior while missing the obtuse angle on the interior. In fact, the shape of III.114 resembles an Archaic chytra from Azoria, and it could possibly represent the continuation of an older tradition of chytrai popular in East Crete (Haggis et al. 2004, 384–385, fig. 41:6).

III.111 (P 2573; Room 6; 64/7412.2; Fig. 30). Chytra with everted rim. Part of rim and neck. Pres. h. 0.025. TF: coarse, reddish yellow 5YR 6/8. III.112 (P 2597; Room 1; 7608.1; Fig. 30). Chytra with everted rim. Single fragment preserving part of rim and neck. Pres. h. 0.024, est. rim d. 0.16. TF: hard, yellowish red 5YR 5/6. III.113 (P 2734; Room 1; 7703.3–4; Fig. 30). Chytra with everted rim. Many fragments preserving part of rim, neck, and side wall. Pres. h. 0.11; est. rim d. 0.24. TF: Mirabello (yellowish red 5YR 5/8). III.114 (P 2800; Room 4; 6504.1–2; Fig. 30; Pl. 7). Chytra with everted rim. Almost complete profile. Pres. h. 0.20; rim d. 0.26. TF: Mirabello (red 2.5YR 5/6).

D. Lopas The lopas is a shallow and wide bowl with a flanged rim, a curved or flat base, and two handles. Most of the lopades were found in the floor

deposit of Room 6. Although they are preserved very fragmentarily, there are enough body sherds in the deposit to believe that they were in use in

46

NATALIA VOGEIKOFF-BROGAN

the period immediately prior to the abandonment of the Beam-Press Complex. All examples display an outturned rim, flat with an inner flange in most cases, although concave rims are not excluded (III.118). The shoulder opens out, recalling lopades of the second century b.c., unlike the straight shoulders of the first century b.c. (Coldstream, Eiring, and Forster 2001, fig. 3.20:c). The handles are horizontal and attached to the underside of the rim, rising slightly above it. Unlike other contemporary lopades from Crete, which display handles of round section (Sackett et al. 1992, pl. 104:13–15), the handles of III.119 have been flattened. Lopas III.117 differs from the rest in the shape of both the side wall, which slopes inward, and the recurved handles. In Crete, lopades with sloping walls and recurved handles are found in Knossian deposits of the second half of the first century b.c. (Sackett et al. 1992, pls. 103:21–22, 104:13).

III.115 (P 2568; Room 6; 64/7410.1–2/7412.1–2; Fig. 31). Lopas. Part of rim. Pres. h. 0.02; est. rim d. 0.20. TF: coarse, reddish yellow 5YR 6/8. Traces of burning. III.116 (P 2572; Room 6; 64/7412.2; Fig. 31). Lopas. Part of rim and side wall. Pres. h. 0.021; est. rim d. 0.21. TF: coarse, dark gray. Traces of burning. III.117 (P 2560; Room 6; 7410.1–2 / 64/7412.1; Fig. 31). Lopas. Two joining fragments preserving part of rim, side wall, and part of handle. Horizontal, recurved handle. Pres. h. 0.036; est. rim d. 0.36. TF: Mirabello (red 2.5YR, 5/8). III.118 (P 2570+P 2571; Room 6; 64/7412.2; Fig. 31). Lopas. Part of rim and side wall. Pres. h. 0.02; est. rim d. 0.16. TF: coarse, red 2.5YR 5/8. Traces of burning. III.119 (P 4207; Room 3; 8610.1; Fig. 31). Lopas. Single fragment preserving part of rim and handle. Est. rim d. 0.31. TF: Mirabello (red 5YR 5/6).

E. Lids Most of the lids (III.121, III.123–III.126) come from Room 4, where they must have been used to cover chytrai with collar rims (e.g., III.99–III.101). III.120 (P 2600; Room 1; 7607.2; Fig. 31). Lid. Three joining fragments preserving part of rim and body. Pres. h. 0.025; est. rim d.: 0.27. TF: coarse, red 5YR 5/8. III.121 (P 4222; Room 4; 6504.1; Fig. 31). Lid. Several fragments preserving knob, part of dome, and rim. Pres. h. 0.022; est. rim d. 0.14. TF: Mirabello (red 2.5YR 5/6). III.122 (P 4660; Room 6; 64/7412.2; Fig. 31; Pl. 7). Lid. Several joining fragments preserving part of knob and largest part of dome. TF: Mirabello (red 2.5YR 5/6).

III.123 (P 4224; Room 4; 6504.1; Fig. 31). Lid. Part of knob. TF: coarse, red 2.5YR 5/6. III.124 (P 4223; Room 4; 6504.1). Lid. Several fragments (a few joining) preserving part of dome. TF: coarse, gray. III.125 (P 8594; Room 4; 6505.1/6511.2; Fig. 31). Lid. Single fragment preserving knob. TF: coarse, light red 2.5YR 6/6–6/8. III.126 (P 8595; Room 4; 6505.1 + 6511.2; Fig. 31). Lid. Single fragment preserving part of knob. TF: coarse, light red 2.5YR 6/6–6/8.

IX. Lighting The Beam-Press Complex yielded only two lamps (III.127, III.128). Both belong to longlived types, and therefore they cannot contribute to a more precise dating of the building. Although lamp III.127 was recovered from the balk of Room 2, its ECCW fabric (identified macroscopically) and its good state of preservation suggest that it was in use in the period prior to the abandonment of the building. Lamp III.127 is characterized by a collar rim, slightly concave base, and a

long nozzle with a large oval wick hole. The lamp should be restored with a horizontal handle. Similar lamps, but with shorter nozzle and smaller wick hole, are known from Central Crete. Catling has dated them to the second half of the second century b.c. (Sackett et al. 1992, 261, nos. L24, L25); in contrast, Hayes has suggested a lower chronology in the first century b.c. (Hayes 1971, no. 41). Similar lamps from Hellenistic burials near Phaistos have been dated to the second quarter of the

POTTERY

second century b.c. based on their context (HadziVallianou 2000, 94–95, pl. 48; see also Mercando 1974–1975, 121–134, nos. 16, 19; for an overview of the type, see Egglezou 2005, 282–284). Fragments of a second lamp (III.128) were found in the floor deposit of Room 6. It is made of dark gray clay and decorated with a relief ivy leaf on the shoulder. This type of lamp is an import from Knidos. Similar lamps were found at Knossos (Sackett et al. 1992, 262, L31–L35, pl. 225), the Athenian Agora (Howland 1958, 126, type 40A), the Mahdia shipwreck (Rotroff 1994, 137–139, fig.

47

13), and houses on Delos (Bruneau 1965, 33–70). This type of lamp first appears in the second century b.c. and continues into the first quarter of the first century b.c. III.127 (C 106; Room 2; 6605.1 Balk; Fig. 31; Pl. 7). Lamp, Cretan. Full profile. Missing small part of side wall and most rim. H. 0.034; base d. 0.03. TF: ECCW. III.128 (C 107; Room 6; 64/7412.2; Fig. 31). Lamp, Knidian. Two joining fragments preserving part of shoulder and nozzle. Nozzle with flukes. Relief ivy leaf on shoulder.

3

Stone Implements Tristan Carter

The study of the stone implements from the Beam-Press Complex is highly problematic as a significant proportion of the assemblage is almost certainly residual, having little relevance to the activities undertaken in this structure during the late second/early first century b.c. Most of the ground stone, and all the chipped stone, have parallels from Neo- and Postpalatial assemblages at Mochlos, whereby one could choose to ignore all of these artifacts. That said, it is not inconceivable that some of the intact prehistoric ground stone tools were being reused. In elucidating what proportion of this material was being employed in the Beam-Press Complex, we have to take into consideration the issues of fragmentation and context— that is, distinguishing those implements that were recovered in a near complete state and part of a floor deposit. At the same time, one has to be wary of assuming that all the material with Bronze Age parallels is necessarily Bronze Age in date; it is unclear as to what extent stone tools continued to be used in post–LM III Crete and to what extent one can trace the style and mode of Minoan lithic technology

into the historical period (Shaw 2000, 386). This problem is, unfortunately, largely a reflection of publication bias, whereby stone tools have tended to receive short shrift once one leaves the prehistoric period, despite the fact that technological analyses of “mundane” or household/community technologies such as stone tool production, coarseware ceramic manufacture, and weaving could provide important avenues of research to investigate the nature of social disjuncture at the end of the second millennium b.c. in Crete and elsewhere. Thus, while the literature has documented and discussed the introduction of a range of specialized agrarian lithic implements during the historical period related to the processing of grain, olives, and grapes with the Classical/Hellenistic hopperrubber mills and beam presses (Frankel 2003; Foxhall 2007), plus the Roman trapeta and mortaria (Childe 1943; Moritz 1958; Runnels and Murray 1983; Foxhall 1993; Jameson 2001), little attention has been accorded to the non-specialized type of tool that dominated Bronze Age ground stone assemblages—the hammerstones, handstones, pestles, querns, polishers, applicators, perforators, and

50

TRISTAN CARTER

weights. Curtis Runnels and others have investigated mainland historical lithic technology, including chipped stone, most famously for threshing sledges (Runnels 1982, 1985a), but little has been written pertaining to Cretan material, though the imbalance has started to be redressed through recent publications from Kommos (Blitzer 1995; Shaw 2000), Chania (Karantzali 1997), Eleutherna

(Kalpaxis 1994), and, in particular, the volume detailing the post-Minoan material stratified above the Unexplored Mansion at Knossos (Sackett and Cocking 1992). It is hoped that this report can make a small contribution to the study of stone tool technology in the historic period, though much remains to be done.

Methodology The study of the stone tools from the BeamPress Complex follows the analytical methodology and mode of presentation established in Mochlos IC (Carter 2004b), with each artifact assigned to a single category. In the case of the ground stone tools, the classification was based on the perceived function of the tool, an interpretation based on its shape and size, plus the location, form, and intensity of use-wear. Subdivisions within these classes draw on both morphology and raw material. Where implements were used for more than one task, they were categorized according to what was considered

to have been the primary function, with all other details recorded in the text. Each artifact was described through reference to a range of attributes, both qualitative: shape, raw material, color; and quantitative: length, width, and thickness, measured by calipers of 0.01 m precision. In the case of ground stone tools, the range of dimensions given for each class only relates to complete pieces. The implements were also studied in natural and artificial light, with the regular use of a small hand lens (10–20x magnification) to note finer details and record use-wear.

Ground Stone Tools The Beam-Press Complex produced 31 ground stone implements from its floor deposits, with just over half coming from two rooms: 13 from Room 1 and six from Room 6 (followed by three from Room 7, and two from Rooms 2, 4, and 5). Following the scheme laid out in Mochlos IC, the three tool-groups (plus “miscellaneous”) were discerned in the Hellenistic assemblages: percussive and abrasive hand-held implements, working surfaces and/or receptacles, weights, and miscellaneous. The following sub-classes are represented in the assemblage: A. Percussive and Abrasive Hand-Held Implements Type 1. Implements with pecked or battered ends Type 2. Implements with pecked circumfer ences and one or two abraded faces Type 3. Pestles

Type 4. Handstones Type 5. Implements with heavy abrasive use wear Type 10. Differentially weathered cobbles B. Working Surfaces and/or Receptacles Type 14. Querns Type 16. Whetstones Type 19. Mortars C. Weights Type 21. Balance-pan weights Type 22. Biconically perforated weights Type 23. Perforated weights Type 25. Naturally perforated weights D. Miscellaneous Type 28. Miscellaneous

STONE IMPLEMENTS

51

A. Percussive and Abrasive Hand-Held Implements (Types 1–5, 10) Type 1. Implements with Pecked or Battered Ends Defined by the occurrence of percussive damage on one or both ends, these implements are often referred to as “hammerstones” or “pounders,” though a number also have a face smoothed from abrasive use-wear. The Beam-Press Complex produced just one hammerstone (III.129). These tools were used to shape or reduce something by striking them in a downward hammer-like motion. A specific function cannot be assigned to these implements on the basis of use-wear alone, as percussive damage can occur by working hard materials such as other stones or working softer materials that are resting on a hard base (e.g., the processing of foodstuffs on a quern). TYPE 1A. HAMMERSTONES

These implements are the simplest form of hammerstone at Mochlos, almost invariably a local limestone beach cobble that has no modification aside from percussive use-wear on one or both ends. The Beam-Press Complex produced only one example from Room 6 (III.129). The hammerstone is a staple feature of Early to Late Bronze Age ground-stone assemblages in Crete, and it is common at Mochlos, with numerous examples from the late LM IB Artisans’ Quarter and the LM III houses (Warren 1972, pl. 79:B [EM II Myrtos Phournou Koryphi]; Bruun-Lundgren and Wiman 2000, 181, pl. 98 [LM IIIC Chania-Kastelli]; for further references, see Carter 2004b, 63–64). III.129 (GS 569; Room 6; 6407.1; Pl. 7). Hammerstone. Intact. Dims. 0.1366 x 0.0878 x 0.0593. Weight 1,307 g. Ovoid limestone cobble with percussive damage on both ends and a limited amount of abrasive usewear on both faces.

Type 2. Implements with Pecked Circum­ferences and One or Two Abraded Faces These implements, also known as hand/ hammerstones, are defined by having percussive use-wear around the circumference. Most also have a central depression pecked into one or both faces

(probably as an aid for handling); they also have one or both faces abraded smooth from use. This combination of percussive and abrasive use-wear indicates that these tools were a form of handstone, used to reduce materials on hard surfaces such as querns by processes of pounding and grinding. The Beam-Press Complex produced five complete examples (III.130–III.134). Each was made from a local limestone cobble and ranged between 0.0639– 0.1417 m long, 0.0605–0.1225 m wide, 0.0256– 0.0661 m thick, and 148–1,920 g in weight. These tools should be Bronze Age in date; they are one of the most common ground stone tool types at Mochlos and are known from numerous Minoan sites from EM IIA onward (Warren 1972, pl. 79:D below [Myrtos Phournou Koriphi]; for further references, see Carter 2004b, 64–65). Their completeness, combined with the fact that they were found in floor deposits, strongly suggests that they were reused by the inhabitants of the Beam-Press Complex. III.130 (GS 268; Room 2; 6606.5; Pl. 7). Hand/ hammerstone. Intact. Dims. 0.1179 x 0.0788 x 0.0661. Weight 940.5 g. Ovoid limestone cobble with a central depression pecked into one face, the opposite face slightly abraded smooth from use and percussive damage on both ends and around parts of its circumference. III.131 (GS 551; Room 1; 7703.3; Fig. 32; Pl. 7). Hand/hammerstone. Intact. Dims. 0.0639 x 0.0605 x 0.0256. Weight 148 g. Circular limestone cobble with a shallow central depression pecked into both surfaces and percussive damage around its circumference. III.132 (GS 619; Room 6; 7412.1; Fig. 32; Pl. 7). Hand/hammerstone. Intact. Dims. 0.0721 x 0.0643 x 0.0276. Weight 196.1 g. Circular limestone cobble with a shallow central depression pecked into one face and percussive damage around its circumference. III.133 (GS 760; Room 5; 7510.3; Pl. 8). Hand/ hammerstone? Intact. Dims. 0.1417 x 0.1225 x 0.0589. Weight 1,920 g. Small flat limestone boulder pecked into a circular form, with a central depression pecked into one surface and the beginning of another in the opposite face. Both faces have an amount of abrasive use-wear. Alternatively may be a large unfinished biconically perforated weight. III.134 (GS 908; Room 1; 7607.2; Pl. 8). Hand/ hammerstone. Intact. Dims. 0.0942 x 0.0783 x 0.0388. Weight 487.6 g. Flat ovoid limestone cobble with percussive damage around its circumference and abrasive wear on both faces.

52

TRISTAN CARTER

Type 3. Pestles Pestles are tools employed for both percussive and abrasive purposes. Commonly elongated or cylindrical in form, the main part of the body acts as a handle, with the base functioning as the working face. Room 1 produced two complete pestles (III.135, III.136), both of conical form and both made of nonlocal stones—one a brown granodiorite (probably from the Gournia region), the other a non-Cretan import of blue-gray diorite. Both pestles are considered to be of Bronze Age date, probably Neopalatial, but their reuse in the Hellenistic period is quite conceivable. The diorite example (III.135) has an almost identical parallel from a LM IB floor deposit in House C.2 at Mochlos (GS 415), though it is worth noting that two pestles from the post-Minoan deposits over the Unexplored Mansion at Knossos are of the same form and a related raw material, blue-gray andesite (Sackett and Cocking 1992, 392, pls. 324, 326:a, nos. 13a, 13b). Unfortunately these Knossian pieces came from contexts of uncertain date, perhaps as early as the Orientalizing period. Note, however, that the floor deposit of Room 1 also yielded large fragments of two Hellenistic hopper-rubber mills (III.143, III.144), which suggest that the two intact pestles, whether of Bronze Age date or not, would have been extremely useful to the grinding activities that took place in that room. III.135 (GS 766; Room 1; 7703.5; Fig. 32; Pl. 8). Pestle. Intact. Dims. 0.061 tall x 0.0516 x 0.0498. Weight 184.3 g. Cobble of blue-gray diorite ground into a truncated cone with a slightly rounded base smoothed from abrasive use-wear. III.136 (GS 386; Room 1; 7608.1; Fig. 32; Pl. 8). Pestle. Intact. Dims. 0.1073 tall x 0.0852 x 0.0759. Weight 754.5 g. Cobble of brown granodiorite ground into a conical form with a hollow pecked into the base, the remainder of which is smoothed from abrasive use-wear.

Type 4. Handstones Using a percussive and abrasive mode, handstones were used to crush and grind materials on a fixed working surface. Room 7 produced the Beam-Press Complex’s only regular ovoid handstone (III.137), the upper stone employed as a grinder in conjunction with a quern or other static working surface. It was made from a locally available limestone cobble and was unmodified aside

from the use-wear; the piece has numerous parallels from the Bronze Age occupation at Mochlos (and Crete more generally), including the late LM IB Artisans’ Quarter and LM III houses (for further references, see Carter 2004b, 65–66). It is thus quite likely a reused Minoan tool. III.137 (GS 754; Room 7; 63/7313.1; Fig. 32). Handstone. Intact. Dims. 0.1096 x 0.0812 x 0.0615. Weight 1,385 g. Large sub-triangular limestone cobble with one face, could conceivably have been used as a small quern or working surface rather than held as a handstone.

Type 5. Implements with Heavy Abrasive Use-Wear These implements have the same degree of abrasive use-wear as that seen on the handstones, but they are classed separately because they are less regular in form. The Beam-Press Complex assemblage contained only one example from Room 7, III.138, which is almost certainly Hellenistic since it is made of a distinctive imported blue-gray andesite, the same raw material as the hopperrubber mills (see below, Type 14d). III.138 (GS 784; Room 7; 63/7312.1; Fig. 32). Grinder. Broken. Dims. 0.1129 x 0.0771 x 0.0566. Weight 595 g. Oblong cobble of blue-gray porous andesite with rounded end and margins; abrasive use-wear on both faces. Might be reused fragment of larger implement.

Type 10. Differentially Weathered Cobbles These limestone cobbles and pebbles have a grooved or waisted form due to the differential weathering of their surfaces, the softer part eroded by water. They are a recognized form of material culture at Mochlos on the basis of being recurrently found in archaeological deposits, often in associations with other tool types. Based on context (rather than use-wear), they have been variously interpreted as small weights for fishing nets and/or looms, while occasionally a worn surface suggests their function as a small grinding implement. One differentially weathered limestone cobble (III.139) came from Room 1, with clear indications of being used as a grinder. This is another type of implement that is commonly recovered from Bronze Age contexts at Mochlos including the late LM IB Artisans’ Quarter and LM III houses, with parallels known from numerous other

STONE IMPLEMENTS

Minoan settlements, up to at least LM IIIC, as recorded at Chania-Kastelli (Bruun-Lundgren and Wiman 2000, pl. 105:d, no. 4; for further references, see Carter 2004b, 71). The piece was probably reused since it forms part of the grinding apparatus of Room 1.

53

III.139 (GS 416; Room 1; 7607.2; Pl. 8). Grinder. Intact. Dims. 0.0801 x 0.0498 x 0.0271. Weight 210.1 g. Oblong differentially weathered limestone cobble pebble with abrasive use-wear on base.

B. Working Surfaces and/or Receptacles (Types 14–19) These tools provided stationary surfaces, or hollows, on, or within which materials were reduced, usually by one of the mobile implements described above. The different categories represented in the Beam-Press Complex assemblage are querns (Type 14), whetstones (Type 16), and mortars (Type 19a).

Type 14. Querns Querns comprise the lower half of an abrasive tool kit employed in conjunction with a hand-held implement (Types 1–5) to crush and powder materials on its surface. Nine querns of varying shapes, sizes, and raw materials came from the BeamPress Complex; typologically they can be separated into three groups: saddle querns (Type 14a), hopper-rubber mills (Type 14d), and beam presses (Type 14e), of which there were three (two [a pair], and one examples, respectively). TYPE 14A. SADDLE QUERNS

Saddle querns are a Bronze Age type of quern, so named for their distinctive shape, being rectangular with rounded ends and having a curved base and slightly concave longitudinal profile. This profile is accentuated as the central part of the quern is worn, which has the benefit of helping to stop the handstone from slipping off the ends. Three (III.140–III.142) saddle querns came from the Beam-Press Complex, each made from small boulders shaped by percussion and abrasion. Room 7 produced the sole complete example (III.142); the other querns are of such a fragmentary state that they should probably be considered as residual and likely not in use at the time of this building’s occupation. These tools are of typical Bronze Age form, typologically akin to those from the Artisans’ Quarter at Mochlos, the LM III

houses, and numerous other third and second millenium sites on Crete and in the southern Aegean (for further references, see Carter 2004b, 73–74). All three querns are made of granodiorite, a stone associated with the Gournia area (Hayden, Moody, and Rackham 1992, 308–311; Haggis and Mook 1993, 290). III.140 (GS 553; Room 1; 7703.4; Fig. 33; Pl. 8). Saddle quern. Broken. Dims. 0.2339 x 0.1782 x 0.0508. Weight 2,115 g. One end of an oblong boulder of finegrained gray-brown granodiorite, worked all over by percussion and abrasion into a flat rectangular form with a rounded end. The end is slightly raised, giving the quern its distinctive longitudinal curvature, while the upper face is abraded from use. III.141 (GS 617; Room 6; 7411.1; Fig. 33; Pl. 8). Saddle quern. One-half extant. Dims. 0.2689 x 0.1532 x 0.0361. Weight 2,375 g. One end and mid-section of an oblong boulder of fine-grained gray-brown granodiorite, worked all over by percussion and abrasion into a flat rectangular form with a rounded end. The end is thickened and raised, giving the quern its distinctive longitudinal curvature, while the upper face is abraded from use. III.142 (GS 756; Room 7; 63/7313.1; Fig. 33; Pl. 8). Saddle quern. Intact. Dims. 0.4348 x 0.1439 x 0.1017. Weight 8,750 g. Oblong boulder of fine-grained graybrown granodiorite, worked on upper face by percussion and abrasion into a flat rectangular form with rounded ends. The margins have received rudimentary pecking while the underside is largely unworked. The upper face is abraded from use giving the quern a longitudinal curvature with slightly raised ends.

TYPE 14D. HOPPER-RUBBER MILLS

Type 14d comprises a specialized and distinctive type of quern/millstone referred to as “hopperrubbers” or “Olynthian mills,” named after the site in Chalkidike where they were first recognized in abundance (Robinson and Graham 1938, 328; Cahill 2002, 163–166, fig. 34). They comprise a top

54

TRISTAN CARTER

and bottom stone, the former having a slit through which grain could be fed into a cavity in the mill (thus serving as a hopper), with a linear grinding motion (back and forth) achieved by using a wooden handle that was inserted into a slit at the front of the upper stone (Moritz 1958, 38–42; Kardulias and Runnels 1995, 121–124, fig. 95; Frankel 2003). Technologically, they represent an improvement on the hand-mill in that it left one hand free to continuously feed the grist into the hopper, thus permitting longer periods of grinding without having to stop. The Beam-Press Complex produced a pair of broken upper and lower stones (III.143, III.144) from Room 1, made from an imported blue-gray porous andesite. Parallels are known from across the Greek world, including Morgantina on Sicily; Isthmia; central Laconia; the southern Argolid; Olynthus; Vergina; Delos; and Eleutherna, Knossos, Trypitos, and Azoria in Crete (for a site distribution with bibliography, see Frankel 2003, 3–5; for Vergina [not included in Frankel’s list], see Tsigarida 1992; for Crete in particular, see Kalpaxis 1994, 124–125, pl. 47 [Eleutherna]; and Sackett and Cocking 1992, 392, 394–395, pls. 324:17, 325:12, 327:57a; Sackett and Jones 1992, 12–13, pl. 29 [Knossos]). The excavations over the Unexplored Mansion at Knossos produced at least two examples, including one from the Hellenistic Southeast House and another from a secondary context of the second to third century a.d. Excavations at Trypitos (inventoried object MS 11870; N. Vogeikoff-Brogan, pers. comm.) and Azoria (pers. obs.) have produced others. The earliest securely dated hopper-rubber mills come from Athens (late fifth century b.c.) and Halieis (350–300 b.c.), and they continued to be produced until the Roman period (WilliamsThorpe and Thorpe 1993, 270). The source of these millstones is debated. Geochemichal analysis of a considerable number of samples from the eastern Mediterranean has identified the islands of Thera (modern Santorini) and Nisyros as the source of most of the Classical and Hellenistic hopper-rubber mills found in Greece (Williams-Thorpe and Thorpe 1993, 291, 293–294; Frankel [2003] unfortunately does not correlate his stylistic analysis of these mills with provenance). Other studies focusing on examples from the Argolid have identified the Saronic Gulf as a potential source for andesite, with

Aegina known to have been a major center of quern manufacture from the later Neolithic to the historical period (Runnels 1981, 117–127; 1985b; Kardulias and Runnels 1995, 121). Hopper-rubber mills made from Saronic Gulf andesite have been recovered from the nearby southern Argolid, at the Classical–Hellenistic site of Halieis and from surface collections at two late Classical/early Hellenistic farmsteads (B36 and E43) from the Argolid Exploration Project (Kardulias and Runnels 1995, 121). It has also been suggested that from the beginning of the Hellenistic period Nisyros, in the Dodecanese, had become the preferred source. Another farmstead site (B89), of purely Hellenistic date, produced three andesite hopper-rubber mills (Kardulias and Runnels 1995, 121). The procurement and exchange of querns has a long heritage in the Aegean, from the later Neolithic onward, with a number of production centers located in areas of volcanic activity exploiting local andesites and tuffs, including Aegina, Thera, and Melos, the latter having an important tradition of manufacturing and exporting millstones until the earlier part of the twentieth century (Moritz 1958, 46–48; Sparkes 1982, 234; Wagstaff 1982, 236–241). From the Hellenistic period there is direct evidence for the maritime transport of querns with the recovery of 29 Nisyrean hopper-rubber mills from the Kyrenia shipwreck located just off the north coast of Cyprus (Runnels 1981, 117–127, table 23). III.143 (GS 390; Room 1; 7607.1; Fig. 34; Pl. 9). Hopper-rubber mill. Broken. Dims. 0.2434 x 0.193 x 0.0959. Weight 4,310 g. Front and corner section of the upper stone of an Olynthian mill made from a boulder of blue-gray porous andesite, worked by percussion and abrasion. The piece retains the front dowel cutting and handle slot, plus the outline of the hopper feeding slot. The ventral surface is dressed with furrows cut in parallel lines mainly along its longitudinal axis but also shorter parallel lines at right angles along the margin. III.144 (GS 391; Room 1; 7607.2; Fig. 34; Pl. 9). Hopper-rubber mill. Three-quarters extant. Dims. 0.3394 x 0.2435 x 0.0884. Weight 8,450 g. One side, plus parts of front and back of a hopper mill’s grinding slab made from a boulder of blue-gray porous andesite, worked by percussion and abrasion. It has a hollow base, having been raised on a flange around its margins; its flat dorsal surface is smoothed from abrasive use-wear.

STONE IMPLEMENTS

TYPE 14E. PRESS BEDS

Type 14e comprises press beds—stone apparatuses employed as a base upon which to compress crushed olives (or trampled grapes) with a stone or wooden press for the extraction of oil and juice. They comprised a lower stone, or bed, onto which sacks of olives or grape pulp were laid, plus an upper crushing stone or stones, with which pressure was exerted onto the olives/grapes below. The press bed will also have a spout, and often a channel leading to it, for the oil to drain out into a container below, such as a pithos (Foxhall 2007, 133–177). The Beam-Press Complex at Mochlos produced a complete and in situ rectangular spouted press bed (III.145; GS 1487) in Room 6, made of a finely worked boulder of nonlocal white limestone. In addition to the well-known press bed from the so-called andreion at Praisos, Despina HadziVallianou, in her Cretan survey of equipment used in the production of olive oil, has recorded a number of press beds, some of which could be dated to the Hellenistic period. They are usually circular although there are some rectangular ones (HadziVallianou 2004, 93–96, 124–126, figs. 32:a–34:b). III.145 (GS 1487; Room 6; 7411.1; Fig. 35; Pl. 10). Beam-press bed. Intact. Dims. 0.517 x 0.376 x 0.115 (height of spout), 0.087 (height of bed). Weight >10,000 g. Complete rectangular spouted beam-press bed made from a boulder of white fine-grained nonlocal limestone. The bed is trapezoidal in cross section, with the base the narrowest part; the upper surface of the bed has an interior margin defined by scoring, while the spout is set into the middle of a raised front lip. The boulder was shaped by percussion and abrasion, the underside being only cursorily smoothed; all faces bear clear traces of peck marks, probably from a pick-adze. The pick marks on the upper face are accentuated, probably intentionally to give the working surface an abrasive texture; the center of this face is smoothed from use.

Type 16. Whetstones Whetstones are small mobiliary implements, usually made from dense fine-grained raw materials, that are employed as surfaces for sharpening the working edges of metal or stone tools. The building produced just one example (III.146; GS 888), from the surface of Room 4, made of fine-grained nonlocal sandstone. Although it does not come from a

55

datable context, it has been included here because of its raw material, which is not otherwise represented at Minoan Mochlos—a fact that could be taken to indicate its Hellenistic date. In turn, however, whetstones, along with saddle querns, appear to be one of the ground-stone classes at Minoan Mochlos that regularly involve the consumption of exotic raw materials and/or the procurement of nonlocal products, as viewed in the assemblages of the late LM IB Artisans’ Quarter and the LM III houses (Carter 2011, 84). Whetstones certainly are one of the types of stone tools that continue to be produced in the historical period, with a number of examples from Protogeometric to Roman Knossos, plus Geometric and Archaic Kommos (Sackett and Cocking 1992, 392–395, pls. 324, 326 [Knossos]; Shaw 2000, 387– 390 [Kommos]). III.146 (GS 888; Room 4; 6501.1; Fig. 35; Pl. 11). Whetstone. Broken. Dims. 0.0754 x 0.0466 x 0.0292. Weight 154.1 g. Broken whetstone worked into a subrectangular form with rounded end, made from a nonlocal orange, fine-grained indurated sandstone with small pebble inclusions/adherents on surface.

Type 19. Mortars Mortars are open vessels within which a variety of materials were reduced by a hand-held percussive or abrasive implement, often a pestle. Although Minoan in date, the unfinished stone lamp III.147 has been included in the publication because it was found on the floor of Room 1 together with other grinding equipment of Minoan and Hellenistic date (III.135, III.136, III.139, III.143, III.144). There is little doubt that III.147 was reused in the Late Hellenistic period as a mortar. The fact that the lamp is unfinished is a reflection of the fact that these vessels were made at Mochlos; indeed, the community manufactured a wealth of stone vases throughout the Bronze Age, with good evidence for a number of Neopalatial production loci (Carter 2004b, 2008). III.147 (S 45; Room 1; 7607.1; Fig. 35; Pl. 11). Stone lamp/mortar. Intact. Rim d. 0.175, base d. 0.10, h. 0.108. Circular depression (ca. 0.12), flat rim with two channels. Pendant handles. Pedestal base. Purple granite? Complete, except for minor chips at edges.

56

TRISTAN CARTER

C. Weights (Types 21–23, 25) On the basis of their form, scale, and context, the stone objects in this tool-group are interpreted as weights, a group of material that can be subdivided into intentionally modified pieces (Types 21– 23) and those with natural perforations (Type 25).

Type 21. Balance-Pan Weights The Beam-Press Complex produced a small piece of a possible balance-pan weight (III.148) from the floor deposit of Room 2. From its fragmentary state and parallels from elsewhere at Mochlos, we can be quite certain of its Bronze Age date. The small size of the fragment also suggests that it was probably residual. Although balancepan weights have been recovered from LM III deposits at the site, and were possibly used at this time, it is thought that most were employed by the Neopalatial community, with numerous stone and lead examples coming from secure LM I contexts in the settlement and the Artisans’ Quarters on the coast (Carter 2004b; Brogan 2006). The raw material of III.148 is interesting as it is a fine-grained nonlocal stone, but also because it may have been chosen on the basis of its color, its purple-black hue recalling (or imitating) that of hematite, the primary raw material used to make balance-pan weights in Egypt, the Near East, and Anatolia during the Bronze Age (Petruso 1992, 1). III.148 (GS 1486; Room 2; E3 6606.1; Pl. 11). Balance-pan weight. Broken. Dims. 0.0296 x 0.0276 x 0.0178. Weight 23.9 g. Corner of a finely polished and shaped implement (sub-cuboid), made from a cobble of purplish-black fine-grained shale/schist.

Type 22. Biconically Perforated Weights These weights are centrally perforated, flat, circular cobbles. The perforation has a biconical profile due to it having been initiated from both faces, the process involving direct percussion to create a central depression, which was then continued by abrasion. Some cobbles were slightly water worn, or had their faces ground smooth prior to perforation, but the margins were generally left unworked. Four examples (III.149–III.152) came from floor assemblages, each made from limestone cobbles.

Although only one was complete (III.152), it was still possible to record the variance in size, from 0.0532 to 0.1323 m in diameter. These distinctive pieces should again be Bronze Age in date, recorded throughout the sequence at Mochlos, with numerous examples from the late LM IB Artisans’ Quarter and the LM III houses (Carter 2004b). They are an extremely common tool type in Bronze Age Crete, produced from at least EM IIA onward (for EM IIA–B Myrtos Phournou Koryphi, see Warren 1972, 117). Each has an hourglass profile unless otherwise stated. The intact weight, III.152, which was included in the floor assemblage of Room 1, was found together with three naturally perforated weights (III.154, III.156, III.157), six clay loomweights (III.170–III.172, III.180– III.182), and one lead weight (III.194). III.149 (GS 375; Room 2; 6704.2; Pl. 11). Biconically perforated weight. One-half extant. Dims. 0.1323 x 0.0712 x 0.0378. Weight 644 g. Circular sandy limestone cobble, biconically perforated, broken in half. III.150 (GS 577; Room 6; 7310.1; Pl. 11). Biconically perforated weight. One-half extant. Dims. 0.1056 x 0.0653 x 0.0354. Weight 315.4 g. Circular limestone cobble, biconically perforated, broken in half and one face possibly sheared off. III.151 (GS 1280; Room 4; 6504.2; Pl. 11). Biconically perforated weight. One-half extant. Dims. 0.0532 x 0.0311 x 0.0209. Weight 39.2 g. Small, flat circular limestone cobble, biconically perforated. III.152 (GS 545; Room 1; 7703.4; Fig. 36; Pl. 11). Biconically perforated weight. Intact. Dims. 0.0671 x 0.0634 x 0.0335. Weight 214.5 g. Circular limestone cobble, biconically perforated and with smoothed surface.

Type 23. Perforated Weights (“Ring-Shaped” Weight) Room 6 produced two fragments of a ringshaped weight (III.153) of fine-grained local yellow sandstone; another conjoining fragment was found in the area outside Rooms 1 and 3. This piece should be restored with two pierced lugs like its parallels from Delos, which have been connected with fishing activities (Deonna 1938, 202–203, fig. 238). If not used for fishing, weight III.153 might have been added, together with

STONE IMPLEMENTS

other weights and stones, to the end of the beam press for extra pressure. III.153 (GS 841; Rooms 6 and 7; 64/7413.1, 64/7414.1, 7807.1; Fig. 36; Pl. 11). Perforated weight. Broken. Ext. d. 0.284; int. d. 16; thickness: 0.064. Weight 1,500.5 g. Mid-section of a circular weight, made from a boulder of fine-grained sandstone. Has a tubular cross-section and was manufactured by percussion and abrasion.

Type 25. Naturally Perforated Weights Five naturally perforated limestone pebbles and cobbles (III.154–III.158) came from the BeamPress Complex, with three from Room 1 (III.154, III.156, III.157) and single examples from Rooms 5 and 6. Their weight varies from 32 to 1,780 g; thus most are quite small and would have been most appropriate as weights for weaving and/or fishing (as net sinkers). Only III.155 is large enough to have been used as a tethering stone for an animal or small craft. While a relatively common object from Minoan levels, these pieces have also been recovered from historic contexts, as for example a group of 10 from the floor of Geometric Building Z at Kommos (Shaw 2000, 390–391, pl. 5.47). Given their unmodified form, there is no way of dating these pieces beyond contextual association. The three (III.154, III.156, III.157) from Room 1

57

were found together with stone weight III.152, six clay loomweights III.170–III.172 (Pl. 13), III.180– III.182 (Pl. 13), and one lead weight III.194 (Pl. 14). Our study of the finds from Room 1 has shown that it was used mostly for grinding activities. The presence of several small weights from clay, stone, and lead may indicate some weaving as well. III.154 (GS 1155; Room 1; 7607.2; Pl. 12). Naturally perforated weight. Broken. Dims. 0.0651 x 0.0588 x 0.0281. Weight 139.5 g. Sub-rectangular flat limestone cobble with natural perforation at one end, its interior worn smooth from use. III.155 (GS 557; Room 5; 7513.1; Fig. 36; Pl. 12). Naturally perforated weight. Intact. Dims. 0.164 x 0.112 x 0.06. Weight 1,780 g. Ovoid limestone cobble with a natural perforation near one margin. III.156 (GS 1160; Room 1; 7607.2; Pl. 12). Naturally perforated weight. Intact. Dims. 0.0593 x 0.0388 x 0.0272. Weight 82.7 g. Sub-ovoid limestone cobble with perforation at one end. III.157 (GS 1162; Room 1; 7607.2; Pl. 12). Naturally perforated weight. Intact. Dims. 0.0641 x 0.0519 x 0.036. Weight 145.4 g. Sub-triangular limestone cobble with perforation near apex. III.158 (GS 568; Room 6; 64/7412.1; Fig. 36; Pl. 12). Naturally perforated weight. Intact. Dims. 0.0529 x 0.0267 x 0.0261. Weight 31.6 g. Elongated limestone pebble with perforation at one end.

D. Miscellaneous (Type 28) Type 28. Miscellaneous One artifact from the Beam-Press Complex assemblage is included in the miscellaneous category, a flat limestone pebble (III.159; GS 1149) from Room 2 with a small hole ground into the center of both smoothed faces. If the holes were intended as perforations, they are incomplete (nor do they align). Although the pebble was found in a surface layer, it has been included in the publication for its rarity. The function of the piece is unknown, though one might suggest either a balance-pan weight or a counter weight. It is not possible to offer any date with certainty since no other examples are known from Bronze Age or historical deposits at Mochlos. III.159 (GS 1149; Room 2; 6701.2; Fig. 36; Pl. 12). Counter/weight? Intact. Dims. 0.0447 x 0.0399 x 0.0071.

Weight 22.6 g. Flat circular limestone pebble with both faces abraded smooth and a small hole ground into the center of each face.

Discussion Of the 31 ground stone implements from the Beam-Press Complex, only a few (III.138, III.143–III.146) are thought to be genuine products of the Hellenistic period; the remainder are of Bronze Age date. While most of the material is residual, this is not to say that none of it was used in the Hellenistic period. As noted above, elucidating the ground-stone tool kit of the inhabitants of the Beam-Press Complex is not without problems. But through consideration of an object’s state and

58

TRISTAN CARTER

context, one can make some progress. Simply stated, the best candidates are those that are whole (or near whole) and in floor deposits. Thus, where only a fragment of an implement is recovered, such as

III.148 (the corner of a balance-pan weight), one can make a good case that the piece relates neither to the Hellenistic building nor to the activities that occurred there.

Raw Materials The majority of the Minoan ground stone tools from the Beam-Press Complex were made from local raw materials, namely limestone and sandstone. Of the small group of “exotic” stones, the granodiorite pestle (III.136) and the balance-pan weight of “nonlocal” schist/shale (III.148) were both probably introduced from other parts of Crete during the Neopalatial period—the former likely procured from the Gournia region. Full descriptions of the Minoan raw materials and their sources are presented in Mochlos IC (Carter 2004b). Of greater significance for this publication, however, are the small number of Hellenistic tools made from imported raw materials. These include three implements made of a blue-gray andesite: the upper and lower stones of a hopper-rubber mill (III.143, III.144) plus a grinder (III.138). This non-Cretan raw material likely comes from either the Saronic Gulf or the Dodecanese. There is also a beam press made of nonlocal limestone (III.145) and a whetstone made of a different nonlocal sandstone (III.146), albeit of uncertain date. Following the scheme laid out in Mochlos IC, the raw materials from the site are categorized as follows: 1. Resources from the immediate vicinity (limestone, sandstone, phyllites/schists) 2. Resources from the coastal plain (amphibolite, serpentinite); not present in this study 3. Resources from East Crete (green quartzite, brown quartzite, granodiorite) 4. Resources from Crete (arenite sandstone, “nonlocal” schist, calc-schist, “nonlocal” sandstone) 5. Off-island resources (andesite, dacite, emery, pumice) 6. Resources of currently unknown derivation (hematite, dunite)

1. Resources from the Immediate Vicinity 1A. LIMESTONE

Limestone beach cobbles from the Mochlos waterfront were the main resource employed in the Beam-Press Complex ground stone assemblage with 28 pieces, 68% of the total. Limestone is represented in seven tool types, being the exclusive raw material in each case. Ground stone tool classes represented in this raw material include: Type 1a (1/1), Type 2 (5/5), Type 4 (1/1), Type 10 (1/1), Type 22 (4/4), Type 25 (5/5), and Type 28 (1/1). 1B. SANDSTONE

Despite the fact that sandstone is a locally occurring beach resource, it was poorly represented in the Beam-Press Complex ground stone assemblage, with only the perforated (“ring-shaped”) weight (III.153) made of this material. Ground stone tool classes represented in this raw material include Type 19 (1/1).

3. Resources from East Crete 3C. GRANODIORITE

This is a relatively fine-grained igneous rock of gray-brown color with black and white inclusions, a raw material known to be indigenous to the southern edge of the Mirabello Bay, specifically the area around Gournia and the northwest isthmus (Hayden, Moody, and Rackham 1992, 308–311; Haggis and Mook 1993, 290). Four implements from the Hellenistic structure were made of granodiorite: a pestle (III.136, Type 3) and three saddle querns (III.140–III.142, Type 14a). Ground stone tool classes represented in this raw material: Type 3 (1/2), Type 14a (3/3).

STONE IMPLEMENTS

59

4. Resources from Crete

5. Off-Island Resources

4B. “NONLOCAL” SCHIST

5A. ANDESITE

One tool from the Beam-Press Complex, a balance-pan weight (III.148), was made from a purplish-brown fine-grained schist/shale. The material bears no resemblance to the locally available schists and phyllites from the coast near the late LM IB Artisans’ Quarter. The specific origin of the stone is unknown, but schist is a relatively common resource in Crete. Ground stone tool classes represented in this raw material include Type 21 (1/1).

The Beam-Press Complex assemblage produced three implements of a blue-gray porous andesite, a grinder (III.138, Type 5) and a pair of hopperrubber mills (III.143, III.144, Type 14d). The Aegean volcanic arcs contain a number of andesite sources—in the Saronic Gulf, Cyclades, and Dodecanese—a number of which are known to have been exploited for the manufacture of querns and millstones from the later Neolithic to modern periods (Runnels 1981; Wagstaff 1982, 236– 241). The two most likely sources of the Mochlos millstones (and by extension the other two implements) are the Saronic Gulf (Aegina in particular) and Nisyros in the Dodecanese, both centers of andesite hopper-rubber mill production during the late Classical and Hellenistic periods (Kardulias and Runnels 1995, 121–123). Without petrographic analysis, the provenance of our examples cannot be certain (cf. Williams-Thorpe and Thorpe 1993); they do appear, however, to fit Runnels’s description of the raw material from Nisyros in that “they often have a texture that resembles a sponge or metallic slag” with “prominent vugs, or cavities” (Kardulias and Runnels 1995, 111). Ground stone tool classes represented in this raw material include Type 5 (1/2) and Type 14d (2/2).

4D. NONLOCAL SANDSTONE AND LIMESTONE

Two other implements from the Beam-Press Complex were made of nonlocal stone: the press bed itself (III.145) and a whetstone (III.146). The raw materials are quite different but are classed here together as the exact location of these stones is currently unknown, though a Cretan origin seems quite possible given the relatively commonplace occurrence of these stones on the island (Fassoulas 2000). The press bed was made of a white, fine-grained limestone with the occasional fossil shell inclusion, while the whetstone was made from an orange, fine-grained indurated sandstone with small pebble inclusions/surface adherents. Ground stone tool classes represented in these materials include Type 14e (1/1) and Type 16 (1/1).

The Chipped Stone The deposits excavated within the Beam-Press Complex contained 78 pieces of chipped stone, all obsidian. Although the manufacture and use of chipped stone tools in the historical period is documented from elsewhere in the Aegean (Runnels 1982; Kardulias and Yerkes 1996), the specific use of obsidian after the end of the Bronze Age is poorly attested. The latest secure evidence for the regular consumption of obsidian comes from the Late Cycladic IIIC cult center at Phylakopi on Melos (Torrence 1985), while in Crete the material was

being knapped and used during LM IIIB at Quartier Nu, Malia (pers. obs.), and LM IIIC in Chania (Karantzali 2000). At what date obsidian ceased to be procured from Melos is currently uncertain, though from a Cretan perspective one would be tempted to suggest the LM IIIC/Sub-Minoan period. Small quantities of obsidian appear to be in circulation in the latest Bronze Age and earliest Iron Age (it is recorded from both Kavousi Vronda and Kavousi Kastro; L.A. Turner, pers. comm.); however, this might in part represent the long-term

60

TRISTAN CARTER

curation of older implements and/or pick-up and reuse of earlier material. A similar interpretation might be accorded to the obsidian blades from a Geometric tomb (associated with toilet equipment) at Praisos (Bosanquet 1901–1902, 251; 1904, 232). Recent publications from Chania and Knossos have recorded obsidian from a number of deposits of Hellenistic, Roman, and later date (Sackett and Cocking 1992, 398; Karantzali 1997); in each

instance it seems safest to assume that the material is in secondary context, having been redeposited from Bronze Age levels below. The same can be argued for the Beam-Press Complex material given that the structure overlies/disturbs Building B2, a large Neopalatial structure that was particularly rich in obsidian (Carter 2004a, 298–305), not least in those upper story levels that suffered the most from the construction of the Hellenistic building.

4

Ceramic, Glass, Metal, and Shell Objects Natalia Vogeikoff-Brogan with contributions by Amanda Kelly and David S. Reese

Roof Tiles The excavation of the Beam-Press Complex yielded a large number of terracotta roof tiles (Figs. 37, 38; Pl. 12), which are mentioned in the excavation notebooks. The notebooks record the discovery of at least one intact pan tile on the floor level of Room 4 with the neck of a transport amphora (III.79; Fig. 25) resting on top of it (see Ch. 1). With few exceptions, however, the roof tiles were not collected systematically during the excavations of 1991 and 1992. In 2001, and in anticipation of the current publication, Amanda Kelly collected and drew a considerable number of tiles that had been piled in the vicinity of the Beam-Press Complex as well as in other areas with Hellenistic remains. As a result, the following report is limited by the method used to collect and record the tiles. A more thorough discussion of the roofing system of the Late Hellenistic structures at Mochlos will follow after the publication of the results of the more recent excavations in 2004 and 2005, which carefully recorded the contextual information of the roof tiles from all Hellenistic structures dug in those two seasons.

The pan tiles, which have a ridged upper end, a beveled lower end to overlap the ridge of the tile below, and raised borders along the entire length of the long sides (Fig. 37; Pl. 12), conform to Wikander’s Corinthian pan-tile type C2a (Wikander 1988, 208). This type of tile was popular in Greece, Asia Minor, and Cyprus from the sixth century b.c. through the Hellenistic period (Wikander 1988, 208 n. 59). The tiles that were collected in 2001 do not preserve any complete example of a pan tile. The only complete pan tile comes from the floor deposit of another Late Hellenistic structure that lies east of the church of St. Nicholas (III.160; Fig. 37; Pl. 12), about 40 m to the east of the Beam-Press Complex. It is a large tile (0.68 x 0.57 m) with a thickness of 0.045 m at the beveled lower end, and it weighs 17 kg (including plastered corner). Of special interest is the fact that the tile slightly tapers (0.01–0.02 m) at the beveled short side to fit tightly upon the raised edge of the next tile. In addition, the underside of the tile is not completely flat, but slightly curved. Like the majority of

62

NATALIA VOGEIKOFF-BROGAN

tiles from Mochlos, it is made of coarse, light red clay (10R 5/6 or 6/8) with phyllite inclusions (Fabric 1). This fabric is macroscopically similar to the Fine Phyllite Fabric of Neopalatial Mochlos (Barnard and Brogan 2003, 5), and it confirms that the tiles, unlike the table and cooking wares found in the Beam-Press Complex, were made locally since phyllite formations surround the entire coastal plain region. The majority of the fragments from the Beam-Press Complex are also made of Fabric 1, except for a small number of tiles (III.164, III.167, III.168 + uncataloged RT 11 and RT 25) made of a reddish-yellow fabric (5YR 6/6) that also contains phyllite inclusions (Fabric 2). As a rule, their measurements seem to be fairly consistent with those of the pan tile (III.160) from Area E4, although slight variations in size were not uncommon in ancient roofs, “but to which extent this is a result of repairs rather than a reflection of the original state of the roofs is difficult to ascertain” (Wikander 1993, 128, 132). For example, the thickness of the raised border of the long sides varies between 0.02 and 0.035 m (III.162 + uncataloged RT 22), and height of the raised border on the long sides could average between 0.045 and 0.054 m (III.162 + uncataloged RT 19). The cover tiles found in the excavation are gableshaped in section (Fig. 37; Pl. 12). Their length is equivalent to that of the pan tiles, and they conform to Wikander’s C1 type (Wikander 1988, 210). They appear to be significantly fewer than the pan tiles (about 1:4), although this may be accidental. In a random assemblage like ours, which was put together 10 years after the excavation of the building, it is possible that some cover tile fragments were not collected because they were not recognized as such. Since cover tiles are smaller and significantly thinner than pan tiles, their fragments sometimes do not differ from coarse pottery sherds. Nevertheless, Wikander has made a similar observation about the small number of cover tiles at Acquarossa, suggesting that “some houses had roofs totally or mainly covered with pan-tiles” (Wikander 1993, 127). As with the pan tiles, both fabrics were used for the manufacture of the cover tiles covering the Beam-Press Complex. Cover tiles made of Fabric 1 may have carried a pair of U-shaped indentations at one narrow end, as is the case with the cover tiles found in the floor deposit of a Late Hellenistic

structure in Trench E3 9600, to the north west of the Beam-Press Complex (III.166; Pl. 12). There are also examples of cover tiles made of Fabric 2, which carry three small parallel indentations resembling the Greek letter Ξ (III.167; Fig. 38). Their weight varies from 3 to 4 kg. Two of the cover tiles (III.167, III.168) associated with the BeamPress Complex display a small variation of about 0.01 m in the height of the gable. Variations in the dimensions of the cover and pan tiles may indicate repairs. Some pan and cover tiles are covered with a light brown slip (7.5YR 7/4) The lack of any ridge tiles or antefixes argues against a gabled roof at the Beam-Press Complex, which is not surprising for the time period. According to Wikander, “it is only from Hellenistic times onwards that we find indications that people tried consciously to avoid using ridge-tiles by building houses without a saddle roof, by replacing them with ordinary cover-tiles . . .” (Wikander 1993, 127). In the case of the Beam-Press Complex, because of the different elevations of its rooms, we have tentatively reconstructed four flat sloping roofs (Fig. 7). III.160 (RT 1; E4 7730.5; Fig. 37, Pl. 12). Pan tile, almost complete. Missing corner of beveled end with long side. Weight: 17 kg (including plastered corner). TF: coarse, light red 10R 6/8 (Fabric 1). Found in another Hellenistic building to the east of the Beam-Press Complex. III.161 (RT 15; Fig. 37). Pan tile. Small fragment preserving corner of beveled short side with long side. Weight: 1.20 kg. TF: Coarse with large phyllites (>0.01m), red 10R 6/6 (Fabric 1). Light brown slip (10YR 7/6). III.162 (RT 20; Fig. 37). Pan tile. Small fragment preserving long side with raised border. Remains of three fingerprints. TF: coarse with phyllites, weak red 10R 5/4 (Fabric 1). III.163 (RT 23; Fig. 37). Pan tile. Small fragment preserving corner of long side and short ridged side. Weight: 1.65 kg. TF: coarse with phyllites, red 10R 5/6 (Fabric 1). Light brown slip (10YR 7/6). III.164 (RT 24; Fig. 37). Pan tile. Fragment preserving short ridged side. Weight: 2.20 kg. TF: medium coarse with phyllites, reddish yellow 5YR 6/6 (Fabric 2). Light brown slip (10YR 7/6). III.165 (RT 29). Pan tile. Small fragment preserving corner of long side and short ridged side. Weight: 1.40 kg. TF: coarse with large phyllites, red 10R 5/6 (Fabric 1). Light brown slip (10YR 7/6).

CERAMIC, GLASS, METAL, AND SHELL OBJECTS

III.166 (RT 6; 9600; Pl. 12). Cover tile. One-half extant. U-shape lines at short end. Weight: 1.7 kg. TF: coarse with large phyllites, red 10R 6/8 (Fabric 1). III.167 (RT 7; Fig. 38). Cover tile. One-half extant. Groove running parallel to one end, and three small parallel lines on one side of gable. Weight: 2 kg. TF:

63

medium coarse, reddish yellow 5YR 6/6. Light brown slip (10YR 7/6). III.168 (RT 8; Fig. 38). Cover tile. One-third extant, preserving one end. Weight: 1.12 kg. TF: medium coarse with phyllite inclusions, reddish yellow 5YR 6/6. Light brown slip (10YR 7/6).

Water Pipe The removal of the North Balk in Room 2 recovered part of a narrow-bore water pipe (III.169; Fig. 38), which must have been part of a larger continuous pipe system that was not found during the excavation of the Beam-Press Complex. The exact location and depth of the pipe along the northern side of the room is uncertain, but if it did indeed belong to the room, it must have served some secondary purpose. Similar pipes, both “male” and “female,” have been found in Hellenistic/Roman contexts at Knossos (Sackett and Cocking 1992, 410–411, pl. 223:51, 52). A fragment of a male pipe was found together with sima fragments in Room

A1 at Kommos, which dates to the Late Hellenistic period. There it was restored on the roof of Room A1 as a drain pipe used to dispose of water accumulating on the roof (Shaw and Shaw 2000, 65, 77, pls. 1.186, 1.187). Similar pipes have been reported under the modern village of Mochlos, and another possible explanation is that pipes like III.169 tapped a nearby spring at Vaya (Βαγιά), behind the modern village, and provided water to the Hellenistic settlement (J.M. Flood, pers. comm.). III.169 (C 1000; E3 6600 North Balk; Fig. 38). Water pipe, male. Broken at one end. P.L. 0.36, d. (mouth) 0.06. TF: coarse, red 2.5YR 5/8.

Loomweights The excavation of the Beam-Press Complex yielded a small number of loomweights (Fig. 39; Pl. 13), more than twenty, some of which were found in outside areas. Only those associated with floor assemblages are cataloged here. There was a small concentration of weights in Room 1: six made of clay and one of lead (III.170–III.172, III.180– III.182, III.194; Figs. 39, 41; Pls. 13, 14). Most were found in the western half of the restricted paved area in the southwestern part of the room (Fig. 11). The weight of the clay loomweights varies from 35 g to 110 g. The wide range in the weight of the weights suggests that they were not used in weaving and most likely served other purposes. The loomweights recovered from the Beam-Press Complex fall into five categories based on shape. The majority of the weights are Type A: round, discoid in profile, and with two holes and a flattened top (III.170–III.174; Fig. 39, Pl. 13; for typology, see Soles et al. 2004, 28–33, 42 n. 33; Cheval

2008). The flattened edge frequently carries a shallow groove (Cheval 2008, 19, figs. 5, 8, 9). The weight of the intact ones varies from roughly 30 g to 38 g (III.170, III.172). Another type, Type B, is also round but with a single hole (III.175–III.177; Fig. 39; Pl. 13). These weights are heavier, ranging between 130 g to 280 g. Both types were produced locally and made of light red clay with various inclusions, including phyllite (Soles et al. 2004, 28). Although the discoid loomweights date to the Late Bronze Age, the Hellenistic inhabitants of Mochlos appear to have re-employed the intact ones (as they did with the stone tools) after retrieving them from the ruins of the Minoan houses (for LM IB parallels, see Soles et al. 2004, 28–33, 42 n. 33). There is no doubt about their reuse since several of them were found on the floor deposit of Room 1 (III.170–III.172; Fig. 39; Pl. 13), and this was a common practice in many periods. The same kind of reuse occurred at Mycenae, where Mycenaean

64

NATALIA VOGEIKOFF-BROGAN

clay loomweights and stone tools were reused in the Hellenistic period (Bowkett 1995, 36, 39), and earlier at Mochlos, where the Mycenaean occupants collected and reused Minoan loomweights (Soles et al. 2011, 54–55). Employing circular weights with one or two holes was not uncommon in the looms of the Hellenistic period (Tébar Megías and Wilson 2008, 53–54, and esp. 55 for reference to other Hellenistic sites). In Crete, an example of a round loomweight with two small holes has been found at Hellenistic House A in Eleutherna (Kalpaxis, Furtwängler, and Schnapp 1994, pl. 35). A third type of loomweight, Type C, which is lentoid in form (III.178; Fig. 39), was found in the floor deposit of Room 4. Lentoid weights are found in the Classical and Hellenistic contexts of the Unexplored Mansion at Knossos (Sackett and Cocking 1992, 401–402). The Beam-Press Complex also yielded two more types of clay loomweights: pyramidal (Type D) and bi-conical (Type E). Two of the pyramidal Type D (III.180, III.181; Fig. 39; Pl. 13) and one of the biconical Type E (III.182; Fig. 39; Pl. 13) loomweights were found in the floor deposit of Room 1. Both types were used in the Hellenistic period (Kalpaxis, Furtwängler, and Schnapp 1994, fig. 37, pl. 35; Tzachili 2006, figs. 5, 6, 9). Pyramidal weights III.180 and III.181 were made of brown clay with small inclusions, unlike weight III.179 (Fig. 39; a surface find in Room 1, but included here to show the variety of clays used for pyramidal weights), which is made of light red clay. Finally, the two biconical weights (III.182, III.183; Fig. 39; Pl. 13) are both light (32 and 35 g, respectively), but they differ in shape and fabric. Loomweight III.182 is elongated, resembling biconical weights from Axos, while III.183 is squatter (for parallels from Axos, Eleutherna, and Knossos, see Sackett and Cocking 1992, 401, pl. 333:22, 23; Kalpaxis, Furtwängler, and Schnapp 1994, fig. 38:K 309; Tzachili 2006, fig. 6). Additionally, III.182 is made of fine, light red clay, unlike III.183, which is made of coarse red clay resembling the Mirabello fabric of the cooking wares. The spindle whorl (III.184; Fig. 39; Pl. 13), which was recovered from Room 3, finds parallels from the Bronze Age to the Hellenistic times (for Bronze Age parallels, see Warren 1972, 246, no. 157; also Alberti 2008, 28; for Hellenistic parallels from Euesperides, see Tébar Megías and Wilson 2008, 53, fig. 6).

Type A. Round with Flattened Top, Two Holes III.170 (C 76; Room 1; 7607.2; Fig. 39; Pl. 13). Loomweight. Roughly round, flattened and grooved top, and two holes near upper edge. Intact. Weight 35 g. TF: coarse, red 2.5YR 5/6. III.171 (C 80; Room 1; 7607.2; Fig. 39; Pl. 13). Loomweight. Roughly round, flattened and grooved top, and two holes near upper edge. Almost intact, except for small chip. Pres. weight 30 g. TF: coarse, reddish brown 2.5YR 4/4. III.172 (C 81; Room 1; 7607.2; Fig. 39; Pl. 13). Loomweight. Roughly round, flattened and grooved top, and two holes near upper edge. Intact. Weight 38 g. TF: coarse, red 2.5YR 5/6. III.173 (C 87; Room 2; 6606.2; Fig. 39). Loomweight. Roughly round, flattened on top, and two holes near upper edge. Missing lower half, chipped on edges. Pres. weight 60 g. TF: coarse, light red 2.5YR 6/6 with gray core. III.174 (C 200; Room 8; 63/7314.1; Fig. 39). Loomweight. Flattened on top, with two holes. About one-third extant. Pres. weight 66 g. TF: coarse, light red 2.5YR 6/6.

Type B. Round with One Hole III.175 (C 84; Room 5; 7510.3; Fig. 39; Pl. 13). Loomweight. Roughly round, grooved top, with one large hole. Chipped on edge. Pres. weight 123.5 g. TF: coarse, light red 2.5YR 6/8. III.176 (C 117; Room 4; 6504.1; Fig. 39). Loomweight. Roughly round, flattened and grooved on top. Small fragment with hole. Pres. weight 49 g. TF: coarse, 5YR 6/6. III.177 (C 202; Room 8; 63/7314.1; Fig. 39). Loomweight. Small fragment, one hole. Pres. weight 27.5 g. TF: coarse, red 2.5YR 5/8.

Type C. Discoid III.178 (C 431; Room 4; 6504.2; Fig. 39). Loomweight. Thickened lentoid disk with one hole near upper edge. Almost intact, except for large chip. Pres. weight 98 g. TF: coarse, reddish yellow 5YR 6/6.

Type D. Pyramidal III.179 (C 48; Room 2; 6703.1; Fig. 39). Loomweight. Small, square in section, flat-topped. Missing lower part. Pres. weight 35 g. TF: medium, reddish yellow 5YR 7/6. III.180 (C 78; Room 1; 7607.2; Fig. 39; Pl. 13). Loomweight. Tall, flat-topped. Almost intact, missing small part of one corner at base. Incised T on one side. Pres. weight 110 g. TF: medium, brown 10YR 7/4–6/4.

CERAMIC, GLASS, METAL, AND SHELL OBJECTS

III.181 (C 79; Room 1; 7607.2; Fig. 39; Pl. 13). Loomweight. Medium sized, square in section, flattopped, but worn. Almost intact, chipped at one corner. Pres. weight 49 g. TF: medium, brown 10YR 6/3.

65

III.183 (C 93; Room 4; 6504.1; Fig. 39). Loomweight. Intact, except for a few light chips. Weight 35 g. TF: medium, red 2.5YR 5/6.

Spindle Whorl

Type E. Biconical III.182 (C 82; Room 1; 7607.2; Fig. 39; Pl. 13). Loomweight. Intact. Weight 32 g. TF: fine, light red 2.5YR 7/6.

III.184 (C 127; Room 3; 86/8716.4; Fig. 39; Pl. 13). Spindle whorl. Biconical with vertical perforation. Intact. Weight: 18 g. TF: medium, yellowish red, 5YR 5/6.

Figurines Outside Rooms 1 and 2, the excavation unearthed two figurines, one almost complete (III.185; Fig. 40; Pl. 13), the other in fragments (III.186; Fig. 40; Pl. 14). Figurine III.185 depicts a short and pudgy naked figure, probably a child (Soles and Davaras 1994, 434, fig. 22). Its fabric is very close to that of ECCW. Of special interest is the second figurine (III.186), which is part of an amphora in the shape of a seated satyr. Its fragments were found in the floor assemblage of Room 1 and also in the outside area, which could be the result of several depositional processes (e.g., primary residual refuse or post-depositional disturbances). Similar plastic vases either in the form of Bes, an Egyptian diety, or Bes-Silenus have been found at Delos in contexts of the late second/early first century b.c. (Barrett 2011, 278, 332, figs. D71, D77; see also Hatzidakis 2004a, 374, pl. 169:ε, στ). In the Delian examples, the figures are depicted seated with a protruding belly, legs open, and hands resting on their knees. Another plastic vase in the form of Bes-Silenus is displayed at

the Ierapetra Museum, but it is without provenance (Papadakis 1986, 83). The Mochlos satyr has been restored somewhat differently, with his arms raised toward his head (see mold from Olynthus in Robinson 1952, pl. 108:364B), unlike the examples of Bes-Silenus from Delos and Hierapytna. Coincidentally, Seager reported the discovery of “. . . a curious rhyton, which may be late Greek, showing the bust of a Satyr holding both hands to his head” (Seager 1909, 276). III.185 (C 132; Outside Room 1; 7708.5; Fig. 40; Pl. 13). Figurine, male. Complete, restored from five fragments. H. 0.10. TF: fine, very pale brown 10YR 8/3. Short, stout, naked male, perhaps a child. Large head with indication of hair on the front, arms bent at elbows. III.186 (C 150; Outside Room 1; 6601.1, 6604.1, 6605.1, 6606.1, 6621.1, 6600 North Balk, 7708.4; Fig. 40; Pl. 14). Plastic vase. Many fragments, several joining, preserving part of ring base with one leg bent at knee, part of a second leg, and one arm bent at elbow. Restored h. 0.165. TF: fine, very pale brown. 10YR 8/4. Black glaze, brown at places.

Glass The Beam-Press Complex yielded a very small number of fragments from glass vessels (Fig. 40). All have been included in the publication, although only III.187 and III.190 come from floor deposits. Fragment III.187 found in Room 5 preserves the rim of a heavy conical or hemispherical bowl in emerald green glass. The shape is common in the Late Hellenistic/Early Roman deposits of the Unexplored Mansion at Knossos (Price

1992, 419–420, pl. 337), and the bowl was probably imported from the Syria-Palestine region, to judge from its popularity in that area. Fragment III.190, found in Room 4, preserves the ring base of a small footed bowl or a skyphos; it is made of green glass that has been painted white. Finally, fragment III.188, which is not from the floor assemblage of Room 4, has a wide groove between two narrow ones on the interior surface of the rim

66

NATALIA VOGEIKOFF-BROGAN

and is comparable to a glass vessel in the National Archaeological Museum in Athens (inv. no. NM 2692; see Weinberg 1992, 100, no. 53). III.187 (G 1; Room 5; 7510.3; Fig. 40). Bowl, emerald green, cast. Part of rim. Pair of wheel cut grooves inside below rim. Pres. h. 0.08. III.188 (G 2; Room 4; 6501.5; Fig. 40). Bowl, brown, cast. Tiny bubbles. Part of rim and upper side wall. Three grooves—a wide one between two narrow ones—inside below rim. Pres. h. 0.05.

III.189 (G 3; Room 3; 8610.2). Bowl, light green(?), cast. Small fragment preserving part of upper side wall and one groove inside. III.190. (G 12; Room 4; 6511.1; Fig. 40). Bowl or skyphos, green with white paint on interior and exterior. Part of base and side wall. Base d. 0.065. III.191 (G 41; Room 1; 7601.4; Fig. 40). Bowl, brown, cast. Small fragment preserving part of rim. Pair of shallow grooves, inside, below rim. Pres. h. 0.016.

Shell The floor deposit of Room 5 yielded one triton shell (III.192; Pl. 14), nearly complete aside from one recent hole. The shell was collected alive, to be eaten foremost. Afterward, it may have been used as a scoop for everyday tasks (Soles et al. 2004, 41). Fragments from another 46 triton shells were found in the Beam-Press Complex. Of interest is a body fragment that may have two cut

parallel sides, which came from the floor deposit of Room 6 (Locus 64/7412.1–2). Worked triton shells used as vessels were popular in the Bronze Age (see App. E). III.192 (Sh 7; Room 5; 7510.3; Pl. 14) Charonia sequenzae. Complete, some vermetids on exterior, recent holes on body, slightly worn apex, lip a bit encrusted. L. 0.218; w. 0.107.

Metal Objects Very few of the metal objects found in the Beam-Press Complex (Figs. 41, 42; Pl. 14) can contribute further to the understanding of the activities taking place in its space. Of interest is a lead pyramidal weight (III.194; Fig. 41; Pl. 14), which was found together with six clay and four stone loomweights in Room 1. A similar weight comes from an Augustan deposit at Knossos (Branigan 1992, 371, no. M 326, pl. 313). In addition to the lead weights, which are usually associated with fishing, the Beam-Press Complex yielded substantial evidence for the consumption of sea shells (see Ch. 1, Room 6) and contained additional fishing gear, like perforated stone weights (III.151, III.152, III.154, III.157) and a ring-shaped stone weight (III.153). Other metal objects of interest are the lid of a small lead pyxis from Room 2 (III.193; Fig. 41) and a lead sheet with radiating incisions (III.195; Fig. 41), which could be a game board. Sheet III.195 from Room 1 was not found in the floor

assemblage, but it is included here because of its incised pattern. Lead binding III.203 (Fig. 41), found together with a folded sheet, must belong to a binding similar to the one found in Deposit H 32 at the Unexplored Mansion (Branigan 1992, 366, no. M 72, pl. 304:13). Some of the lead sheets found in Room 6 (III.199–III.202) might have been used in the construction of the press. Finally, the excavations recovered several nails made of bronze and iron, and a small ring-shaped piece made of bronze (III.211; Fig. 42). All bronze nails are square in section throughout their shank. With the exception of large nail III.206 (Fig. 42), which is 0.13 m in length, the rest of the bronze nails are of medium size. Nails III.206 and III.208 had their ends bent after being driven into the wood (Shaw and Shaw 2000, 384). A flat iron object (III.215; Pl. 14) of unknown use was recovered from the floor assemblage of Room 1.

CERAMIC, GLASS, METAL, AND SHELL OBJECTS

Lead III.193 (Pb 3; Room 2; 6621.1; Fig. 41). Lead lid with knob. Two fragments preserving full rim and part of dome with knob. III.194 (Pb 7; Room 1; 7610.1; Fig. 41; Pl. 14). Lead weight. Intact. Pyramidal, perforated at top. Weight: 0.166 kg. III.195 (Pb 8; Room 1; 7701.2; Fig. 41). Lead plaque, folded. Incised, radiating lines on circumscribed(?) surface. H. 0.115; thickness: 0.004. Est. length, if unfolded 0.10. Weight: 0.336 kg. III.196 (Pb 11; Room 4; 6511.2; Fig. 41). Lead sheet, folded at places. Hole (probably recent, from hand-pick). Thickness: 0.001. Weight: 0.113 kg. III.197 (Pb 12; Room 4; 6505.1, found in unexcavated spur of dirt to the east of threshold; Fig. 41). Lead sheet, folded at places. Incised lines? Two small holes. Thickness: 0.001. Weight: 0.0783 kg. III.198 (Pb 13; Room 4; 6505.1; Fig. 41). Lead sheet, folded at places. Thickness: 0.001. III.199 (Pb 14; Room 6; 64/7412.1; Fig. 41). Pair of lead sheets, folded. Narrow strips. Est. L.: (a) 0.11; (b) 0.18. W.: (a) 0.026; (b) 0.003. Thickness: 0.001. Weight: (a) 0.0389 kg; (b) 0.055 kg. III.200 (Pb 15; Room 6; 64/7412.1; Fig. 41). Small mass of folded lead sheets. Weight: 0.089 kg. III.201 (Pb 17; Room 6; 64/7412.1; Fig. 41). Lead sheet with small holes. Thickness: 0.002. III.202 (Pb 33; Room 6; 64/7412.2; Fig. 41). Lead sheet. Thickness: 0.002. Weight: 0.017 kg. III.203 (Pb 38; Room 2; 6621.1; Fig. 41). Lead sheet, folded at places, and part of binding. P.L. of sheet: 0.10; p.w. of sheet: 0.06. H. of binding: 0.03; p.L. of binding: 0.025. Weight of sheet: 0.057 kg. Weight of binding: 0.0245 kg.

67

III.204 (Pb 39; Room 4; 6504.2; Fig. 41). Lead object, folded and disfigured. Weight: 0.013 kg. III.205 (Pb 55; Room 2; 6600 North Balk; Fig. 41). Lead strip. Width: 0.05m. Weight: 0.024 kg.

Copper Alloy III.206 (CA 68; Room 8; 63/7314.1; Fig. 42). Bronze nail. Bent at end, square in section. Three incisions on top. Est. length (before bending): 0.13. III.207 (CA 208; Room 4; 6505.1; Fig. 42). Bronze nail. Bent near top, square in section. Est. length (before bending): 0.08. III.208 (CA 209; Room 4; 6504.1; Fig. 42). Bronze nail. Bent at end, square in section. Est. length (before bending): 0.07. III.209 (CA 215; Room 4; 6504.1; Fig. 42). Bronze nail. Slightly bent, square in section. L. 0.10. III.210 (CA 117; Room 2; 6604.1; Fig. 42). Rectangular bronze sheet. Minor chips at edges, recent hole. Dims. 0.048 x 0.057. Thickness: 0.001. III.211 (CA 85; Room 4; 6504.1; Fig. 42). Bronze ring. Intact. Triangular in section. Outer d. 0.018.

Iron III.212 (Fe 3; Room 4; 6511.1; Fig. 42). Iron nail. Corroded. Missing its head. Rectangular in section. P.L. 0.11. III.213 (Fe 12; Room 4; 6505.1; Fig. 42). Iron nail. Small fragment. Corroded. P.L. 0.084. III.214 (Fe 13; Room 4; 6504.1). Iron nail or slag. Two fragments. Corroded. III.215 (Fe 15; Room 1; 7607.1; Pl. 14). Five fragments, most joining, of a flat piece. Dim. 0.185 x 0.115. Thickness: 0.07. Corroded.

5

The Late Hellenistic Settlement at Mochlos and the Political and Economic Sovereignty of Hierapytna Natalia Vogeikoff-Brogan

The Beam-Press Complex forms part of the larger Late Hellenistic settlement at Mochlos. The settlement is located on the south slope of the island and is surrounded by a circuit wall that runs along the southern, eastern, and northern sides of the island (Fig. 3). The abrupt cliffs of the western side would have formed a natural fortification. Most of the buildings are oriented along an east–west axis, and they were built inside this circuit wall. The Beam-Press Complex is one of only two or three buildings found outside the walls of the Hellenistic settlement, a location which may have been determined by the non-residential activities conducted in these structures. Study of the pottery from the Beam-Press Complex, as well as from the other buildings, suggests that the southern side of the island, after its abandonment in the Late Minoan IIIB period, was not occupied again until the end of the second century b.c. Our evidence supports the results of three recent surveys in the area of the Mirabello Bay, which have recorded a dearth of coastal Hellenistic settlements to the east of Istron (Fig. 2) in the time before the Early Roman period (Hayden 2004, 171–172, fig.23; Haggis

2005, 85–88; Vogeikoff-Brogan 2012a). There is evidence, however, for a brief, limited occupation of the summit at Mochlos during the seventh and fourth centuries b.c., but the broad south slope of the island remained entirely unoccupied until the Late Hellenistic period (Vogeikoff-Brogan 2012b). The Mochlos Hellenistic settlement was thus short lived, beginning in the late second century b.c. and lasting less than 100 years. Study of the pottery has shown that most of the tableware and the storage and transport vessels used by the inhabitants of Mochlos were made of East Cretan Cream Ware fabric. Petrographic analysis has suggested that the clay resources employed for the fabric originated near the Myrtos Valley, at the western edge of the Ierapetra Plain, not far from the city of ancient Hierapytna. For a type of ware like the ECCW to have enjoyed such wide distribution, it must have been produced en masse by a large production center, one that should have been located at Hierapytna. The fact that the Mochlos inhabitants consumed large quantities of the ECCW most likely indicates that the site was under the political and economic influence of Hierapytna. That

70

NATALIA VOGEIKOFF-BROGAN

Mochlos did not fall under the sovereignty of any other neighboring city on the north coast is also evident from an isopolity treaty between Hierapytna and Lato dating to 111/110 b.c. The treaty defined the northwest boundaries of Hierapytna in detail, and it excludes any expansion of the city of Lato to the east, with the possible exception of a Lyttian exclave in the area of Pacheia Ammos or Gournia (Chaniotis 1996, 338–351, esp. 345–346, no. 59). Connecting pottery with political geography should be done with caution since, in some cases, shared pottery styles may not reflect political expansion but only common cultural tradition. The latter especially may be the case when one deals with pottery that imitated or was influenced by the production of a larger center or city. For example, it has been hypothesized that the Lasithi region fell within the territory of Malia in the MM I period on the basis of stylistic similarities between Maliote and Myrtos Pyrgos pottery. In an exhaustive study examining in detail the pottery of the

two sites, Carl Knappett concluded that, despite similarities, most of the Myrtos Pyrgos pottery was produced locally. He consequently argued that the Maliote influence was due to ideological rather than political control (Knappett 1999). On the other hand, in the 1990s Catherine Morgan and Todd Whitelaw argued that Argive pottery found in graves near Lerna in the Late Geometric period should mean that the settlement there was a daughter community/colony of Argos (Morgan and Whitelaw 1991, esp. 90–91). In the same vein, Cyprian Broodbank argued that the Cretan features, both in shape and decoration, in the pottery of the Middle Bronze Age on Kythera (Phase 2) constitute evidence for the Cretan origin not only of the potters but also of the consumers. In other words, there was a strong presence of Cretans on the island during this period (Broodbank 2004, esp. 75–76). The case of Mochlos is hardly any different from that of Late Geometric Lerna or Middle Bronze Age Kythera.

Mochlos as Hierapytna’s Harborage on the North Coast The systematic excavation of the Late Hellenistic settlement at Mochlos, combined with the results of the Vrokastro, Gournia, and Kavousi surveys, offers a unique opportunity to review the political and economic history of the eastern part of Crete in this period. According to both Strabo (10.479) and the wellknown treaty between Hierapytna and Itanos, the defining event of the second century b.c. in East Crete was the conquest of Praisos by Hierapytna in the 140s (Chaniotis 1996, 303–310, nos. 47–49, esp. no. 47), which allowed Hierapytna to expand its eastern borders to a point near Itanos and thus access the north coast. Previously, a number of isopolity treaties with Lato, Arkades, Priansos, and possibly Viannos show how Hierapytna had extended her territory to the lower elevations of the Lasithi Mountains on the west, securing access to the central part of the Mirabello Bay (Chaniotis 1996, 217–222, no. 14, 255–264, no. 28, 273–274, no. 35, 338–351, no. 59, with previous bibliography). In his thorough study of the Cretan isopolity treaties, Angelos Chaniotis has argued that Hierapytna’s large territorial expansion in the second

century b.c., through aggressive (conquest) and non-aggressive (isopolity treaties) behavior, should be explained by her need for more pasture in order to cope with her growing population and other social problems (e.g., concentration of land in the hands of few landowners; see Chaniotis 1995, 72– 77; 1999). The search for extra pasture was an important, but probably not the only, reason for Hierapytna’s expansion to the north. Strategically located at a harbor that benefited from trade between Rhodes and Alexandria, the city of Hierapytna, like other Cretan cities (e.g., Itanos), increased her revenues by collecting harborage fees and levies on nonCretan products in transit; this especially was the case after the conquest of Praisos, when the islands of Leuke and Chryse on the south coast, which originally belonged to Praisos, most likely fell under the economic control of Hierapytna, and Leuke became the object of a long dispute between Itanos and Hierapytna (Viviers 1999; Vogeikoff-Brogan 2004, 217–219; Chalikias 2013, 32–36, 50). The establishment of a port facility at a key location on the north coast would have increased

THE LH SETTLEMENT AT MOCHLOS AND THE SOVEREIGNTY OF HIERAPYTNA

Hierapytna’s income from the trade goods passing along the north coast of Crete. Whether or not the Cretan cities exported their own goods, it is certain that they were profiting from transit trade around the island, as is obvious from a number of bilateral treaties (Viviers 1999). The choice of Mochlos over harbors at Pacheia Ammos and Tholos (Fig. 2) is puzzling at first because Mochlos is difficult to access by land. This isolation is, moreover, reflected in the finds from both the Gournia and the Kavousi surveys, which revealed scant evidence for any occupation of the north isthmus and the Kavousi plains, the surrounding lowlands, and the hill slopes before the first century a.d. (Vogeikoff-Brogan 2012a). But transshipment of goods across the isthmus may not have been Hierapytna’s goal in occupying Mochlos during the last decades of her supremacy in East Crete before the arrival of the Romans. Hierapytna’s expansion to Mochlos is better understood if we view the latter as a coastal station for collecting harbor fees, in the way the city of Itanos on the north coast had controlled the island of Leuke on the south coast for part of the Hellenistic period. Furthermore, Mochlos may not have been Hierapytna’s first choice; she may have been “forced” in this direction after unsuccessful attempts to gain control of other harbors on the north coast of the isthmus (Pacheia Ammos and Tholos) during the second century b.c. (Vogeikoff-Brogan 2012a, esp. 81–83). A funerary inscription of the late second/early first century b.c. found near Vasiliki may have marked a polyandrion connected to Hierapytna’s

71

attempts at a northern expansion (van Effenterre and van Effenterre 1989). The lack of any significant settlements in the eastern part of the Mirabello Bay for much of the second and first centuries b.c. could also indicate that this area represented a bone of contention between Lato, Hierapytna, and possibly Lyttos, and was, therefore, not suitable for the establishment of dependable coastal stations. Furthermore, it has been argued, based on reference to Lyttos in an isopolity treaty between Hierapytna and Lato at 111/10 b.c., as well as on a passage in Strabo (10.4.3 [C 475]), who attributes Minoa to Lyttos, that Lyttos had an exclave in the area of Gournia or Pacheia Ammos (Chaniotis 1996, no. 59, esp. 346; see also Vogeikoff-Brogan 2012a, esp. 82). The study of the pottery from the Beam-Press Complex, as well as from the other buildings in the settlement, suggests that the settlement was abandoned in the course of the first century b.c., beginning in the second quarter. The final abandonment of the settlement at Mochlos at the end of the first century b.c. is better understood when two factors are viewed together: (1) the development of Tholos as Hierapytna’s prime harbor on the north coast after the first century a.d.; and (2) the network of settlements in the isthmus and Kavousi plains, which was created in order to facilitate the city’s transshipment trade between the north and south coasts. This expansion was made possible only after the establishment of Roman rule throughout Crete, which brought about both the unification of the Cretan cities and an end to earlier impermeable boundaries.

Olive Cultivation and Pressing at Mochlos The Beam-Press Complex is a large building measuring about 200 m2. The study of its architecture and finds produced evidence for several industrial activities, including the grinding of cereals and the production of olive oil. It is possible that the East Unit of the complex (Rooms 1–3) functioned as a set of “magazines” because of the easy access to the different rooms in this unit and the industrial features of Room 1. The processing installation in Room 1 is difficult to understand. It could have served as a multi-purpose installation

used for crushing olives or even for grinding cereals, depending on the time of year. The West Unit housed a stone-built beam press in Room 6; more importantly, over 50 samples of charred complete and crushed olives were collected and saved from the floor deposit of this room, representing only some part of the total that was once present (App. F). Rooms 7 and 8 were probably used either as storerooms for the produce waiting to be processed or as an animal shelter. It is likely that several parts of the complex must have

72

NATALIA VOGEIKOFF-BROGAN

been seriously depleted or disturbed after its abandonment, especially if life continued on other parts of the island. This holds especially true for Room 6, where everything “valuable” and unbroken— like stone weights or olive crushers—would have been taken away except for the press bed itself, which was built-in and not easily removable. With the exception of House A at Eleutherna, final publications of Cretan Hellenistic buildings of either domestic or industrial character are nonexistent. While pioneering in its holistic approach to the study of Hellenistic domestic structures, the publication of House A at Eleutherna lacks a contextual approach that considers the spatial distribution of the artifacts and the organic remains in light of the many types of depositional processes (Kalpaxis, Furtwängler, and Schnapp 1994). The so-called Andreion (Almond Tree House) at Praisos, with a beam press preserved in situ, is perhaps the closest published Hellenistic parallel to the Mochlos Beam-Press Complex. Excavated and published in the 1900s, however, the function of the Praisos Andreion building is not well understood, although the results of James Whitley’s recent excavation of the site suggest that the installation of the beam press belongs to a secondary phase following Hierapytna’s conquest of Praisos (Bosanquet 1901–1902, 263–269; Brun 2004, 114–115; HadziVallianou 2004, 79–80; Westgate 2007, 434–435; Whitley 2011, 42). Other examples of Hellenistic buildings in Crete with in-situ installations for oil production are located at Lato, at Sfakoryako, the port of Phaistos, and at Chalara of Phaistos (HadziVallianou 2004, 80–81). The best-preserved beam press on Crete, however, was discovered recently at Azoria near Kavousi, and it dates to the sixth century b.c. (Haggis et al. 2011, 46–61). While the evidence from the beam-press installation in Room 6 is clearly connected with pressing olives, Lin Foxhall’s most recent study on beam presses of the Classical and Hellenistic periods in Greece argues that the presses were multipurpose installations. Moreover, the small amount of pressing equipment recovered by the intensive surveys in the areas of Methana, the southern Argolid, Laconia, Kea, and southeast Attica suggests that oil and wine manufacture was not an incomegenerating enterprise before the advent of the Romans (Foxhall 2007, 202). One wonders, however, what the intensive survey of a wine-producing

area like Rhodes, Kos, or Thasos would have yielded in terms of pressing equipment. Of interest is the result of a survey on the island of Megisti (Kastellorizo) in the Dodecanese, which identified at least 47 open-air treaderies (πατητήρια, places where the treading of grapes took place), all located on flat platforms of bedrock, dating from the sixth century b.c. to the early centuries a.d. (Ashton 2002). In Foxhall’s opinion, the presence of crushing equipment within a pressing installation is the leading criterion for its identification as a dedicated olive press (Foxhall 2007, 146, 164, 194). But she does acknowledge that people could have crushed olives by other means and that crushing equipment could have been moved off-site in later periods (Foxhall 2007, 146, 176; see also Brun 2003, 10). For example, there is evidence in mainland Greece for the use of the trapetum, or olive mill, already from the fourth century b.c. (e.g., at Olynthus and Argilos). But the presence of the trapetum is not mandatory for the identification of a site as an olive press because the trapeta were probably one of the first items to be removed when the site was abandoned (Brun 2003, 10; Foxhall 2007, 146). Since no crushing equipment was found in the Beam-Press Complex, the issue of how the olives were crushed needs to be examined further, especially if there were alternative methods. To explain the lack of crushing equipment at the so-called Andreion (Almond Tree House) at Praisos, Hamish Forbes and Foxhall suggested that the long and low stone-built wall in the middle of the olive-press room was used as a crushing bed (Forbes and Foxhall 1978, 41). Sophocles Hadjisavvas has proposed a similar function for a low wall covered with fine slabs at the northern end of the Nicosia-PASYDY olive press (Hadjisavvas 1992, 28). From the early Hellenistic building at Sfakoryako, south of Phaistos, comes a shallow and wide basin with a small hole at the bottom, which combined crushing and pressing (HadziVallianou 2004, 80, fig. 15:b; see also HadziVallianou 1987, pls. 317, 318). There is evidence that a variety of less-expensive methods for crushing olives survived through the early 20th century in Greece (Oikonomou 1996, figs. 4–6). Harriet Blitzer reports that until the early 20th century, the Mesara farmers in Crete crushed their olives in the field “by beating and

THE LH SETTLEMENT AT MOCHLOS AND THE SOVEREIGNTY OF HIERAPYTNA

rolling them on a stone slab with a waterworn cobble or boulder” (Blitzer 1993, 170–171). In the case of Mochlos, however, there is another possibility worth exploring: that the olives were crushed in the processing installation of Room 1, with the aid of some sort of crusher (a large boulder?). Somebody would have run or rolled a heavy stone on top of the olives, which would have been spread on the paved surface, as was still done in several places in Greece until the beginning of the 20th century (Oikonomou 1996, 364, 367, figs. 2, 6). This is the way the olives were crushed in the Azoria olive press, where the excavators found a roller-crusher (Haggis et al. 2011, 52–54). The Mochlos beam press is an important discovery since there are few examples of in-situ beam presses of the historical periods in Crete (until Roman times) despite the many press beds and crushers that have been found on the island (HadziVallianou 2004). The recent discovery of the Azoria olive press, which dates to the sixth century b.c., has been a most valuable addition to the small corpus of preserved olive presses on Crete (Haggis et al. 2011, 46–60). Compared to other known olive presses of the Classical and Hellenistic periods in Greece, Cyprus, and other Mediterranean sites, the Mochlos press is unusual in two ways: (1) the large size of the platform that supported the press bed, and (2) the absence of a functional channel (canalis rotunda) on the press bed. The large platform is missing from the olive presses at Praisos, Nicosia-PASYDY (Cyprus), and the twin-press at Mari-Kopetra (Cyprus). The large stone-built platform, however, might be a Cretan feature, since similar platforms were found at Late Minoan Kommos, Late Classical–Hellenistic Lato, and Hellenistic Phaistos (Shaw and Shaw, eds., 1995, 486, pl. 8.63; Hadzi-Vallianou 2004, 80–81, figs. 14:b, 16:a). Hadjisavvas has noted that in early or lesssophisticated olive presses, crushing and pressing were probably combined into one procedure (contra Foxhall, who believes that olives need to be crushed separately while the press is working; pers. comm.); the large press bed from Maroni-Vournes (1.06 x 0. 87 m) was probably used both for crushing and pressing (Hadjisavvas 1996, 60–63, fig. 1). Likewise, in explaining the large size of the older press bed at Praisos, Robert Bosanquet suggested that the olives were crushed in a single bag, which would have spread out under pressure, hence the

73

need for a wider space (Bosanquet 1901–1902, 267, fig. 34). Assuming that the pressing of the olives took place on the wide platform of the press bed, the olive pulp, packed in square or round sacks, would have been placed directly on the press bed (Bosanquet 1901–1902, 265–266; Foxhall 2007, 135). A beam press anchored in a niche at the south wall of Room 6 would have applied, with the help of weights at its other end, the necessary pressure on the pulp required for the extraction of the olive oil. The only problem with this hypothetical restoration is that the anchoring of the beam on the south wall would have made access to Room 4 difficult during the pressing period, which could have lasted for several months. When the pressing period was over, one assumes that the beam would have been removed, thus freeing access to Room 4. Our artistic restoration of the beam press (Fig. 18) has opted for a different solution, anchoring the beam on a perforated boulder on top of the large back “support” of the platform. Although this solution may limit the length and the pressing strength of the beam, the length might have been sufficient (about 2.5 m) for pressing olives. One problem with this restoration is, however, that the operation of the press probably would have blocked the doorway to Room 5. In addition to the structural details of the Mochlos beam press, its possible function as an olive press in a short-lived, small port facility like Mochlos—a different setup from that of the affluent households of the Classical Attic or Delian countryside (Foxhall 2007, 55–83)—needs to be discussed. Blitzer’s recent ethnographic work in the Mesara Plain in South-Central Crete and Leland Allbaugh’s post–World War II documentation of olive culture and oil consumption in Crete may be more helpful in understanding the mechanisms behind olive cultivation in ancient Crete (Allbaugh 1953; Blitzer 2004). Although the modern Cretan landscape is synonymous with olive groves and vineyards, this may not have been the case in antiquity. Despite the references to olive culture and production in the Linear A and B tablets, the press installations at Late Minoan Kommos, and the various pressing equipment of Late Minoan and Hellenistic/ Roman date found throughout the island (HadziVallianou 2004), it has been argued that until the

74

NATALIA VOGEIKOFF-BROGAN

late 19th/early 20th century, the cultivation of the olive tree in Crete was an extremely laborious process with a rather limited yield. Unlike most studies about olive oil production, which concentrate on equipment and methods of extraction and production of olive oil from the fruit or on estimates of oil production based on the loading capacities of the different types of presses in antiquity (Mattingly 1993), Blitzer has focused on the cultivation methods of the olive tree and its yield capacity before the introduction of irrigation, fertilizers, and systematic pruning in Crete (Blitzer 2004). According to Blitzer, there were only two choices in the cultivation of olive trees in Crete before World War II: either (1) to graft wild olive plants in situ, a technique that produced a limited yield; or (2) to transplant wild olive plants to richer soils, a labor-intensive process that exceeded the immediate subsistence needs of a family, counteracted the production of grain, and required the care of many generations (Blitzer 1993, 170–171; 2004, 150–158; Kamilakis 2004, 220–221; for other methods of olive propagation, see Foxhall 2007, 97–112). Based on her work in the Mesara, Blitzer concluded that “the span of time needed for both the establishment of extensive olive groves and the production of surplus of olives and olive oil” has been seriously underestimated (Blitzer 1993, 171). On the other hand, the considerable number of crushers and bed presses that Hadzi-Vallianou has recorded throughout the island, especially in the Mesara Plain, constitutes strong evidence for a vigorous olive culture in Crete already from the Hellenistic period; how productive this was remains uncertain, since the annual oil yield per tree was most likely low. In a study of Crete, which took place in the immediate post–World War II years (1947–1949), the annual oil yield per tree for the pre-war years did not exceed five pounds (2.26 kg; Allbaugh 1953, 269), although Blitzer’s ethnographic study reports a higher amount of oil yield per tree (5–6 okades = 6.4–7.6 kg) when referring to trees that are 10–15 years old (Blitzer 2004, 154). Taking into consideration the above, it would be difficult to connect the construction of the Mochlos beam press with any large scale production of olive oil. First, if Mochlos was settled after a long period of abandonment, there would not have been enough time for newly grafted olive trees to

produce considerable quantities of fruit. To maintain a hypothetical annual consumption of about 100–150 kg of olive oil, a small-scale farming household would have needed to tend more than 50 olive trees (with an average bi-annual yield of 5 kg of oil per tree) based on the data from the Mesara study and the Allbaugh report (Allbaugh 1953; Blitzer 2004, 154; see also Foxhall 2007, 94, which estimated the annual consumption of wealthy households in classical Greece at 200–330 kg of olive oil). We might restore a landscape of subsistence farms in the coastal plain and the lower elevations of the ravines that lead to Mochlos; tending this many olive trees, however, especially if they were not all located in one place, would have been quite laborious for the individual farmer and his family in addition to their other agricultural activities, including the cultivation of grain and pulses (see Forbes’s comments in Zarinebaf, Bennet, and Davis 2005, 185 n. 127; on the laboriousness of the olive harvest, Kamilakis 2004, 240; on scattered land holdings, see Foxhall 2007, 56–58; also Allbaugh 1953, 245). It is unclear to what extent the practice of slavery, especially in the context of Cretan piracy following the descent of the Cretans from the mountains to the coast (Brulé 1978), might alter the picture during this period. A recent review of house plans of Classical and Hellenistic sites in East Crete, like Lato, Azoria, and Trypitos, reveals a preference for the linear type of house that left no room for segregation of the inhabitants, thus suggesting a very small number of slaves per household. Ruth Westgate has argued that the linear house plan with its modest character may “reflect a difference in the role or treatment of slaves in these households, or their absence” (Westgate 2007, 448; emphasis added). What then does the construction of the Mochlos beam press indicate? David Mattingly assumes that the construction of olive mills and presses is an indication of surplus oil production because subsistence production could be carried out using rudimentary equipment. Modern scholarship tends to equate the construction of oil presses with surplus and trade, although Foxhall has recently introduced the term “domestic production” to indicate a level of production that went beyond subsistence levels but was not enough for market-oriented olive oil production (Foxhall 2007, 38, 250).

THE LH SETTLEMENT AT MOCHLOS AND THE SOVEREIGNTY OF HIERAPYTNA

Keeping in mind the unusual nature of the settlement at Mochlos, as well as the results of the two ethnographic studies on the agricultural economy of Crete until the first decades of the 20th century, the construction of a permanent (noncollapsible, non-transferable) beam press suggests processing on some considerable scale, probably beyond small-scale household “subsistence” level, although the extent remains uncertain. More importantly, a permanent installation could also reflect the interest of the inhabitants, not so much in increasing the level of oil extraction, as in

75

decreasing the time taken to process the load (the best parallel for it is the fast potters’ wheel, the invention of which enabled potters to produce vessels more quickly, but not necessarily for purposes of trade). The Mochlos beam press may have acted simply as a convenient and time-saving facility that allowed the inhabitants of Mochlos to concentrate on other tasks that were not necessarily agricultural, especially if the primary purpose of the settlement was to collect harborage fees and facilitate sea faring on the north coast of Crete.

Wine Production at Hierapytna in the Late Hellenistic Period If one accepts Foxhall’s proposal that multipurpose pressing installations existed, the Mochlos beam-press installation in Room 6 could also have been used for crushing grapes. But again, even if it was used for pressing grapes, the Mochlos inhabitants should not be considered dedicated farmers; the purpose of the Mochlos occupation in the late second to early first century b.c. was connected with maritime profit. The discussion about wine production here is not required because of the beam-press installation, but because several of the transport amphorae found at Mochlos were made of ECCW and appear to have been produced by Hierapytnian workshops (see above, “Introduction”). Until recently, most scholars would have argued that Crete did not become involved in the production and export of wine before the Roman conquest of the island in 67 b.c. With the exception of a small number of stamped amphora handles deriving from Hierapytna, there was no other archaeological evidence to suggest an earlier date for Crete’s wine production. Moreover, Marangou, in her extensive survey of the island, had not located any amphora workshops of Hellenistic date (Marangou-Lerat 1995). Since then, and taking into consideration the evidence from Mochlos, Chaniotis has recently raised the chronological limit of his argument to ca. 110 b.c., when the wars between the Cretan cities stopped and Crete entered into a new period that allowed for an increase of outside contacts (Hadjisavvas and Chaniotis 2012, esp. 166–167­). Recent field work in East Crete has produced evidence to

support the hypothesis that Cretans—already from the first half of the second century b.c.—were producing wine for export, and that the case of the Hierapytnian amphorae was not the exception to the rule (a small number of stamped amphora handles from Hierapytna dating to the second century b.c. were found in Alexandria, Callatis [Black Sea], and Trypitos; see Marangou-Lerat 1995, 123–124; Vogeikoff-Brogan and Apostolakou 2004). At the sixth Scientific Meeting for Hellenistic Pottery held in 2000, the results of a petrographic analysis that included samples from transport amphorae dating to the late second to early first century b.c. were presented from three sites: Mochlos, Myrtos Pyrgos, and Knossos (Vogeikoff-Brogan et al. 2004; also Eiring, Boileau, and Whitbread 2002). The results show that a considerable number of the transport amphorae from Mochlos and Myrtos Pyrgos were made of the ECCW, the clay source of which was placed in the vicinity of Hierapytna (Ch. 2; see also Hadjisavvas and Chaniotis 2012). It was then argued that the ECCW transport amphorae were produced at Hierapytna in order to carry the city’s wine. Independently, Vasiliki Stefanaki, reconsidering the date of the Hierapytnian stamped amphora handles from Alexandria in connection with the city’s coinage production, suggested that the appearance of names of magistrates on both the coins and transport amphorae of Hierapytna indicates that the city was making a serious effort in the second century b.c. to assert administrative control over her economy, and she was no longer depending on subsistence alone (Stefanaki 2001, 138).

76

NATALIA VOGEIKOFF-BROGAN

Subsequently, additional evidence for local production of transport amphorae from another site in East Crete (Trypitos) was found, already from the early second century b.c. (Vogeikoff-Brogan and Apostolakou 2004; the results of the petrographic analysis of the amphorae from Mochlos and Trypitos were published in Vogeikoff-Brogan, Nodarou, and Boileau 2008). Chaniotis, on the evidence of one Rhodian stamped amphora dated to ca. 162– 155 b.c., has suggested a date later than the middle of the second century b.c. for the abandonment of Trypitos (Hadjisavvas and Chaniotis 2012, 168 n. 83); the study of the pottery, however, does not support a date lower than ca. 150 b.c. Chaniotis’s additional argument that the 32 Rhodian amphora stamps (ca. 245–155 b.c.) do not constitute enough evidence to support exhaustive trade with Rhodes when “a single excavation in the agora of Thessaloniki has yielded 60 Rhodian amphorae, almost as many as the whole of Crete,” is not methodologically sound (Hadjisavvas and Chaniotis 2012, 168 n. 84). Chaniotis compares the yields of a market place with those of residential contexts in the case of Trypitos, not to mention that the houses were small and had limited storage capacities. Furthermore, the AC6 type of transport amphora found at Hagios Nikolaos (ancient Lato pros Kamara), which has been considered to be Early Roman (Marangou-Lerat 1995, pl. XXII, esp. nos. A 147, A 148), may well belong to a Hellenistic period of production since it has been found together with Knidian coins of ca. 300 b.c. (Vogeikoff-Brogan and Apostolakou 2004; contra Hadjisavvas and Chaniotis 2012, 167 n. 77, where the concurrence of Roman amphorae and early Hellenistic coinage is not satisfactorily explained). The evidence from Trypitos and Mochlos also adds archaeological substance to Polybius’s mention of Cretan wine (Ath. Deipnosophistae 440e– f), as well as to an inscription of the third century b.c. referring to several hectares of vineyards at Kydonia (Marangou-Lerat 1995, 11–13; Marangou 1999, 270; Vogeikoff-Brogan and Apostolakou 2004; Vogeikoff-Brogan, Nodarou, and Boileau 2008). Today, after 15 years of research, there is enough archaeological evidence to argue that there was sufficient wine production in Hellenistic Crete to require the local manufacture of large pointed amphorae for sea transportation. The evidence is

no longer limited to the Late Hellenistic period; the study of the pottery from Trypitos has confirmed the local production of transport amphorae as early as the first half of the second century b.c., and material from Hagios Nikolaos, when it is fully published, may even push the date earlier to the third century b.c. What remains unsolved is the extent of the wine trade on Crete, and whether it reached off-island markets or traveled only locally by sea. Not much is known about wine installations and their form in Crete during the historical periods. The survey at Gavdos has produced a large number of rock-cut treaderies throughout the island, which are difficult to date, and types A and B could be dated anywhere from prehistoric to Roman times (Christodoulakos et al. 2000, esp. 572–573). Their distribution pattern shows that the treading of grapes took place near the vineyards. Very similar treaderies, cut in the natural soft marl, have also been discovered on the island of Megisti (Kastellorizo) in the Dodecanese, covering a chronological spectrum from the sixth century b.c. to the early centuries a.d. (Ashton 2002). Medieval treaderies cut in soft marl have been found in several locations of Central and central West Crete. The evidence from Gavdos and Kastellorizo supports the existence of a long-standing tradition of rockcut, open-air treaderies in the southern Aegean (Yapitzoglou and Moschovi 2002, 186). The rockcut treaderies have been connected with the production of a special type of sweet wine in antiquity, which required the sun-drying of the grapes before treading (Ashton 2002, 151–152; Yapitzoglou and Moschovi 2002, 186). The recent surveys of Vrokastro, Gournia, and Kavousi, however, did not identify any open-air treaderies, although northeast Crete is abundant in marlstone. Their absence is intriguing; either there was no viticulture in these areas or grapes were pressed by a different method. On the one hand, a lack of viticulture is quite possible since there was very little habitation in these areas during the Hellenistic period, but it does not explain the absence of treaderies from later periods when the isthmus and the Kavousi plains were extensively occupied. On the other hand, if Hierapytna, both Hellenistic and Roman, exercised state control on wine production, it is quite possible that the vineyards and their processing installations were concentrated on

THE LH SETTLEMENT AT MOCHLOS AND THE SOVEREIGNTY OF HIERAPYTNA

the south coast, and this would explain why they are not found on the north coast. While regional studies have been largely connected with surface surveys, this publication shows the breadth of information that can be extracted from studying the material culture of an excavated small settlement on the periphery of a city’s region, combining with it historical resources and the results of recent surveys in the area. Even though the purpose of an extensive or intensive survey is

77

to fan out from the city in ever-widening arcs to search for the location of settlements contemporary with the city (to paraphrase William McDonald [1984, 186]), the excavation and publication of a regional settlement like Mochlos allows us to look back toward the large city from its periphery. This approach becomes even more valuable since the material culture of the large city, Hierapytna, is itself largely irrecoverable because of modern development.

Appendix A

Petrographic Analysis of Local and Imported Transport Amphorae from Knossos, Mochlos, and Myrtos Pyrgos Marie-Claude Boileau and Ian Whitbread

The Project The aims of this project, which analyzed transport amphorae of the Late Hellenistic period from Knossos, Mochlos, and Myrtos Pyrgos, were to define: (1) local and nonlocal fabric types; (2) the relationship in fabrics between Central and East Crete; and (3) the origin of non-Cretan amphorae (Tables 2–4; Pls. 15–18A). The analysis took place at the Fitch Laboratory of the British School of Archaeology in 1999. Two more samples (NV 36-2004 and NV 37-2004) from the Beam-Press Complex were analyzed in 2004 at the petrographic laboratory of the INSTAP Study Center for East Crete by Eleni Nodarou. Petrographic analysis was conducted with the permission of the Greek Ministry of Culture (Διεύθυνση Συντήρησης Αρχαιοτὴτων).

Preliminary reports have appeared in: Eiring 2000; Eiring, Boileau, and Whitbread 2002; Vogeikoff-Brogan et al. 2004; and VogeikoffBrogan, Nodarou, and Boileau 2008. For a concordance of selected sherds and a summary of final results, see Tables 2 and 3. The abbreviations NV and JE stand for samples taken from Mochlos by Natalia Vogeikoff-Brogan (NV) and from Myrtos Pyrgos and Knossos by Jonas Eiring (JE), respectively. The bold number in parenthesis that follows the NV abbreviation refers to the Mochlos catalog number of the vessels discussed. For example, in referencing NV 15 (III.80), III.80 is the publication catalog number of amphora P 2154 (see Ch. 2).

The Geological Background The geology of Crete, although varied within each region, is repetitive, with similar rock formations in the eastern, center, and western parts of the island.

It is covered at 1:200,000 by Nikolaus Creutzberg and colleagues (1977), and certain areas of the island have been mapped at 1:50,000 by the Institute

80

MARIE-CLAUDE BOILEAU AND IAN WHITBREAD

of Geology and Mineral Expoloration (I.G.M.E; see, e.g., I.G.M.E. 1959 for the Ierapetra region).

Geology of the Knossos Area, Central Crete (after Creutzberg et al. 1977 and Wilson and Day 1994): Quaternary: Alluvium Pliocene: Predominantly white to yellowish, partly finely bedded marine marls and bioclastic limestones; locally, gray-bluish sandy clays. Upper Miocene-Lower Pliocene: Bioclastic and recifal limestones alternating with white and yellowish laminated or homogeneous marls; locally, evaporites. Jurassic to Eocene: Outcrops of Gavrovo-Tripolitza series: Thickbedded to massive, usually gray limestones and dolomites of tidal to shallow water origin; locally, brecciated or fine crystalline. The geology around Knossos is essentially characterized by Neogene limestones, marls, and clays. Metamorphic outcrops of the phyllite-quartzite series are found to the south, near Archanes (Mt. Juktas).

Geology of the Myrtos Pyrgos Area, Southeast Crete (Creutzberg et al. 1977): Quaternary: Alluvium Upper Miocene-Lower Pliocene: See description above for Knossos regional geology. Middle to Upper Miocene: Continental deposits including fluviatile and lacustrine sediments: conglomerates, sandstones, clays, lignite. Limestone breccias and cemented rubble of different composition. Brackish and marine sands, clays, and marls. The region northwest of Myrtos Pyrgos is characterized by the system of ophiolites composed

mainly of serpentinized periodotites, subordinate gabbros, and basalts. It is also associated with the amphibolites gneisses, micaschists, calcsilicate rocks, and marbles, locally intruded by granitic rocks. The Flysch melange with outcrops of mafic volcanites and ultramafic rocks (mainly ultramafic rocks: peridotites, mostly serpentinized and predominantly mafic volcanites: diabase, pillow lavas, spilite) and Gavrovro-Tripolitza are also present in this area.

Geology of the Mochlos Area, East Crete (Creutzberg et al. 1977) Quaternary: Alluvium Permian to Eocene: “Plattenkalk” formation: mostly dark, locally light colored, thin-bedded to platy, finely to coarsly crystalline limestones with bands and nodules of chert. Locally, thin intercalations of reddish or greenish phyllites. Mafic volcanites and ultramafic rocks (see decription in Myrtos Pyrgos regional geology). Gavrovro-Tripolitza series (see description in Knossos regional geology). The phyllite-quartzite rocks are the continuation of the geology of the mainland and are overlying the thin-bedded cherty limestones of the Plattenkalk series. The phyllite-quartzite series contains rocks of different age, origin, and grade of metamorphism, some of them looking very similar. The main part of the phyllite-quartzite series is formed by Permo-Triassic shales, phyllites, quartzphyllites, and quartzites. Local marbles and conglomerates are also found. Gypsum and graywacke are characteristic but minor constituents. Metamorphosed basalts and andesites are common.

Geology of the Islands The geology of the islands of Kos, Chios, Knidos (Datça Peninsula), Paros, and Rhodes has already been summarized by Whitbread (1995, 55–57 [Rhodes], 83–85 [Kos], 69–72 [Knidos/ Datça Peninsula], 138 [Chios], 225–227 [Paros]).

PETROGRAPHIC ANALYSIS OF LOCAL AND IMPORTED TRANSPORT AMPHORAE

81

The Pottery The material from all three sites was sorted according to fabric and shape (the Mochlos material by Vogeikoff-Brogan, and the Knossos and Myrtos Pyrgos material by Eiring). All samples are from amphorae except for sample number JE 17. The macroscopic sorting identified one fabric that

is common to both eastern sites and was named East Cretan Cream Ware (ECCW). The other fabric variations represent local and imported wares or local imitations of imported wares. For a concordance of samples, see Table 2.

The Methodology The methodology used in this report follows Whitbread’s proposal for the systematic description of ceramic thin sections (1995, 379–388). The conventions used are the following: Frequency labels (after Kemp 1985, 17): Predominant >70% Dominant 50%–70% Frequent 30%–50% Common 15%–30% Few 5%–15% Very few 2%–5% Rare 0.5%–2% Very rare 2 mm Macro 0.5–2 mm Meso 0.05–0.5 mm Micro