The Akko Marina Archaeological Project 9781407315027, 9781407344935

This book analyses archaeological finds retrieved from the Akko marina and its surroundings. Analysis of structures and

240 86 456MB

English Pages [353] Year 2017

Report DMCA / Copyright

DOWNLOAD PDF FILE

Table of contents :
Front Cover
Title Page
Copyright
Contents
Foreword
Chapter 1. The Akko Marina Archaeological Project – Introduction
Chapter 2. The Akko Harbor in Numismatic, Epigraphic, Cartographicand Photographic Evidence
Chapter 3. Hellenistic, Roman and Byzantine Ceramics from the Akko Marina
Chapter 4. Amphora Stamps from the Akko Marina
Chapter 5. Maritime Depictions on Ceramics from the Akko Marina
Chapter 6. Ship Timbers from the Akko Marina: Early Use of Unpegged Mortise-and-tenon Joints and Mixed Shell-and-frame Construction
Chapter 7. The Fatimid, Crusader, Mamluk and Ottoman Ceramics from the Akko Marina
Chapter 8. Coin Finds from the Akko Marina: Isolated Coins and a Hoard of Thirteenth-Century Gold Florins
Chapter 9. A Fifteenth-Century Harbor Installation with Wooden Pillars in Akko
Chapter 10. Anchors from the Akko Marina
Chapter 11. Ship Fittings and Rigging Devices from the Akko Marina
Chapter 12. Fishing Gear from the Akko Marina
Chapter 13. Jewelry and Adornments from the Akko Marina
Chapter 14. Metal Objects from the Akko Marina
Chapter 15. Copper Bun-shaped Ingot from the Akko Marina
Chapter 16. Stone Finds from the Akko Marina
Chapter 17. Weapons and Ammunition from the Akko Marina
Chapter 18. Pig Tusks and Animal Bones from the Akko Marina
Chapter 19. Glass Vessels and Raw Glass Chunks from the Akko Marina
Chapter 20. Structures and Installations In and Around the Akko Marina
Chapter 21. Shipwrecks and Cargoes in and around the Akko Marina
Chapter 22. The Akko Marina Archaeological Project – Summary
Recommend Papers

The Akko Marina Archaeological Project
 9781407315027, 9781407344935

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

________ Ehud Galili (PhD) is a research fellow at the Zinman Institute of Archaeology, University of Haifa, a member of the National Committee for the Protection of the Coastal Environment, director of the Atlit-Yam project and excavator of the submerged Neolithic settlements off the Carmel coast, as well as an emeritus marine archaeologist and researcher (Israel Antiquities Authority).

BAR  S2862  2017   GALILI   THE AKKO MARINA ARCHAEOLOGICAL PROJECT

This book analyses archaeological finds retrieved from the Akko marina and its surroundings. Analysis of structures and installations casts light on the harbour’s building and destruction cycles; for example, a 15th century wooden mole has been discovered, indicating previously unknown activity in that period. Hellenistic to late Ottoman period ceramics reveal the city’s international connections and commerce. Glass artifacts and raw glass finds shed light on the famous local glass industry. Shipwrecks, anchors, rigging devices and cargoes starting from the Late Bronze Age tell us about shipbuilding and commercial ties. A unique 13th century hoard of gold florins reveals the last days of Crusader Akko as described in historical documents. Fishing gear indicates fishing activity and weapons and ammunitions provide a glimpse of the conflicts and battles in Akko and its role in local and world history. Numismatic, epigraphic, cartographic and photographic evidence of activity from the Hellenistic period onward depicts the harbour and associated facilities, including ancient and modern lighthouses, breakwaters, and other structures.

The Akko Marina Archaeological Project Ehud Galili with contributions by Gerald Finkielsztejn, Zaraza Friedman, Liora Kolska Horwitz, Yaacov Kahanov, Robert Kool, Baruch Rosen, Jacob Sharvit, Na’ama Silberstein, Dov Zviely and a foreword by David Jacoby

BAR International Series 2862 9 781407 315027

B A R

2017

The Akko Marina Archaeological Project Ehud Galili with contributions by Gerald Finkielsztejn, Zaraza Friedman, Liora Kolska Horwitz, Yaacov Kahanov, Robert Kool, Baruch Rosen, Jacob Sharvit, Na’ama Silberstein, Dov Zviely and a foreword by David Jacoby

BAR International Series 2862 2017

by Published in BAR Publishing, Oxford BAR International Series The Akko Marina Archaeological Project © Ehud Galili and the contributors severally Medieval grapnel iron anchor from Akko. Photo: E Galili. The Authors’ moral rights under the UK Copyright, Designs and Patents Act are hereby expressly asserted. All rights reser ved. No par t of this work may be copied, reproduced, stored, sold, distributed, scanned, saved in any for m of digital for mat or transmitted in any for m digitally, without the written per mission of the Publisher.

ISBN 9781407315027 paperback ISBN 9781407344935 e-format DOI https://doi.org/10.30861/9781407315027 A catalogue record for this book is available from the British Library

BAR titles are available from: BAR Publishing Banbury Rd, Oxford, [email protected] + ( ) + ( ) www.barpublishing.com

,

This book is dedicated to Carmela Hertzberg-Galili and Rachel Melzer Galili

Contents

Foreword David Jacoby.................................................................................................................................. vii Chapter 1: The Akko Marina Archaeological Project – Introduction Ehud Galili ....................................................................................................................................... 1 Chapter 2: The Akko Harbor in Numismatic, Epigraphic, Cartographic and Photographic Evidence Ehud Galili, Dov Zviely and Baruch Rosen .................................................................................. 12 Chapter 3: Hellenistic, Roman and Byzantine Ceramics from the Akko Marina Na‘ama Silberstein, Ehud Galili and Jacob Sharvit ....................................................................... 35 Chapter 4: Amphora Stamps from the Akko Marina Gerald Finkielsztejn ..................................................................................................................... 163 Chapter 5: Maritime Depictions on Ceramics from the Akko Marina Zaraza Friedman, Ehud Galili and Baruch Rosen ....................................................................... 181 Chapter 6: Ship Timbers from the Akko Marina: Early Use of Unpegged Mortise-and-tenon Joints and Mixed Shell-and-frame Construction Yaacov Kahanov .......................................................................................................................... 187 Chapter 7: The Fatimid, Crusader, Mamluk and Ottoman Ceramics from the Akko Marina Ehud Galili, Baruch Rosen and Jacob Sharvit ............................................................................. 204 Chapter 8: Coin Finds from the Akko Marina: Isolated Coins and a Hoard of ThirteenthCentury Gold Florins Robert Kool.................................................................................................................................. 216 Chapter 9: A Fifteenth-Century Harbor Installation with Wooden Pillars in Akko Ehud Galili and Baruch Rosen ..................................................................................................... 243 Chapter 10: Anchors from the Akko Marina Ehud Galili and Baruch Rosen ..................................................................................................... 246 Chapter 11: Ship Fittings and Rigging Devices from the Akko Marina Ehud Galili and Jacob Sharvit...................................................................................................... 252 Chapter 12: Fishing Gear from the Akko Marina Ehud Galili and Baruch Rosen ..................................................................................................... 259 Chapter 13: Jewelry and Adornments from the Akko Marina Ehud Galili ................................................................................................................................... 266 Chapter 14: Metal Objects from the Akko Marina Ehud Galili ................................................................................................................................... 269 Chapter 15: Copper Bun-shaped Ingot from the Akko Marina Ehud Galili and Jacob Sharvit...................................................................................................... 277 Chapter 16: Stone Finds from the Akko Marina Ehud Galili ................................................................................................................................... 279 Chapter 17: Weapons and Ammunition from the Akko Marina Ehud Galili and Baruch Rosen ..................................................................................................... 282

v

The Akko Marina Archaeological Project

Chapter 18: Pig Tusks and Animal Bones from the Akko Marina Liora Kolska Horwitz, Ehud Galili and Baruch Rosen................................................................ 286 Chapter 19: Glass Vessels and Raw Glass Chunks from the Akko Marina Ehud Galili, Baruch Rosen and Jacob Sharvit ............................................................................. 290 Chapter 20: Structures and Installations in and around the Akko Marina Ehud Galili ................................................................................................................................... 298 Chapter 21: Shipwrecks and Cargoes in and around the Akko Marina Ehud Galili ................................................................................................................................... 313 Chapter 22: The Akko Marina Archaeological Project – Summary Ehud Galili and Baruch Rosen ..................................................................................................... 320

vi

Foreword D. Jacoby Akko (Acre) has a long and tumultuous history. It is first mentioned in an inscription of 1479 BCE in the great temple of Amon at Karnak in Egypt commemorating the victorious campaign of Pharaoh Thutmose III against Megiddo in Canaan. Since then it has been successively occupied and governed by numerous political entities and rulers. It is often assumed that Akko fulfilled at all times a pivotal function in Levantine trade and shipping, thanks to its geographic location. This sweeping statement requires serious qualifications. Geography was undoubtedly an important component in Akko’s history. The promontory on which Akko is located at the north of a sandy beach in the bay of Akko created a relatively safe anchorage and enabled the beaching of ships. However, political and economic circumstances were the decisive factors determining the city’s evolving role in Mediterranean trade and shipping. Moreover, one should also remember that Akko’s economy was not exclusively geared toward the sea, but rather simultaneously integrated within the trading networks of its continental hinterland. Akko’s role as port and market evolved over time. Phoenician merchants traded in the city from the tenth century BCE. A Greek colony resided there by the middle of the fourth century BCE. Alexander the Great established a mint at Akko after defeating the Persians at Issos in 333 BCE, which also reflects the city’s active participation in regional land and maritime trade. Akko was called Ptolemais from 208 BCE until the Arab conquest of 636 CE. The city was favored by Julius Caesar. Akko served as the regular landing place of Roman and allied troops in the following two centuries. A breakwater was built in the first half of the second century BCE, clearly improving mooring conditions. A depiction of Akko with a circular basin enclosing a galley on a Roman coin confirms the existence of the breakwater. A colony of Roman army veterans was settled at Akko between 52 and 54 BCE. However, the function of the city and its harbor was limited to some extent by the building of an artificial harbor in Caesarea under Herod the Great in 25–13 BCE and, beginning shortly afterward, by that city’s role as the successive provincial capital of Roman Judea, Roman Syria Palaestina and Byzantine Palaestina Prima. Still, Akko remained the maritime outlet of Galilee, at the height of its prosperity in the Byzantine period until the Arab conquest of 636 CE, and retained its commercial relations with Damascus. A shipyard is attested at Akko in the late Byzantine period, shortly before 636 CE. It is unclear whether it served for the construction of ships for the navy or for private use. In any event, before he founded the Umayyad dynasty in Damascus, Mu’awiya made Akko one of the bases of his naval forces, which fought the Byzantine navy and furthered his conquests overseas. In 649 he established a shipyard at Akko or, more likely, reactivated the old one. The Umayyad Caliph Hisham I, who ruled from 724 to 743, turned Akko into his main naval base. In 861 the

Abbasid Caliph Mutawakkil equipped a fleet at Akko, as he did in other cities. In the late ninth century, Ahmad ibn Tulun, a semi-independent governor of Egypt who annexed Palestine and Syria, built a second jetty that improved the safety of Akko’s harbor. This was the last addition ever made to Akko’s harbor until modern times. Akko reached the peak of its demographic and urban development while under Crusader rule from 1104 to 1187 and from 1191 to 1291. This is also the period in which the city’s harbor enjoyed the most intense traffic throughout its history. It was the destination of merchants, crusaders, and thousands of Christian pilgrims on their way to Jerusalem. In that period the harbor was also involved in trans-Mediterranean trade with the West on a scale never reached earlier or later. The forces of Mamluk Egypt totally destroyed the city in 1291, and according to some sources filled its harbor with debris to prevent the return of Christian forces. Still, Akko remained stopover for ships sailing along the Levantine coast, as well as for Western ships bringing pilgrims and exporting cotton from the city’s hinterland. Following the Ottoman conquest of 1516 a colony of French merchants was established at Akko. The semi-independent governor of the province of Sidon, the Druse Emir Fakhr ed-Din, began to revive the city between 1595 and 1634, yet only to a limited extent. The modern development of Akko began under Daher el-Omar, who ruled the city under nominal Ottoman suzerainty around the mideighteenth century and turned it into the capital of province extending from Beirut to the north to Caesarea to the south. He established himself at Akko, restored and fortified it, yet enclosed a smaller area than the Crusader city. He also ensured its water supply by means of an aqueduct. The Albanian adventurer Ahmed el-Jazzar, governor of Akko from 1775 to 1804, also engaged in construction in Akko. Napoleon Bonaparte’s attempt to conquer the city in 1799 was thwarted by a British naval squadron which intercepted the French siege artillery sent from Egypt, as well as by improvements in the city’s defenses. Successive rulers of Akko during the nineteenth century also engaged in construction in the city and improved its fortifications. Akko was occupied by British forces in 1918 and remained under British rule until the establishment of the State of Israel in 1948. During the entire period extending from the Ottoman conquest of 1516 Akko’s port enjoyed modest traffic associated with export of cotton grown in its hinterland. It was eventually supplanted by the port of Haifa, opened in 1933. Since then it has served only as anchorage for fishing boats. The long evolution of Akko’s harbor and its changing fortunes over some three and a half millennia are partly reflected by the archaeological material recovered from the Akko Marina and its surroundings. The material is studied in 20 independent chapters, preceded by an introduction and followed by a discussion of the finds.

vii

Chapter 1

The Akko Marina Archaeological Project – Introduction Ehud Galili Israel Antiquities Authority and Zinman Institute of Archaeology, University of Haifa Key words: underwater archaeology, shipwrecks, ancient pottery, cultural heritage, Acre, dredging

Introduction

of its history. Geological, geomorphological and sedimentological processes naturally change the coast, causing erosion and destruction on the one hand, and accretion on the other. These processes – and human intervention in the coastal environment in the form of development and construction of maritime structures, quarrying of sand, reclaiming land from the sea and the quarrying of kurkar – quickly destroy nearby ancient harbor structures and installations. Marine archaeological remnants in and around the harbor and iconographic descriptions are key testimony that can help us understand the history of the harbor and the human activity it encompassed. Historic Akko, which exists to this day, is situated atop ancient remains. Extensive development and infrastructure work have been undertaken to maintain a level of modern services that meet the needs of a population in the late twentieth and early twenty-first centuries, and to improve the city’s appearance and make it attractive to local and foreign tourists. Infrastructure work for road construction, sewage, water supply, electricity, communications, and the preparation of structures for tourist services, commerce, residential housing and hotel accommodations, makes it necessary to deal daily with the ancient cultural remains found everywhere in the city. Construction and development, which involved work in the sea itself, required preliminary archaeological surveys, underwater excavations and ongoing supervision. Such work was needed to uncover ancient remains, study them and assess their importance, document and map them and identify the sites and finds that should be preserved in situ for the generations to come. This report presents findings from an underwater survey conducted by the IAA (Israel Antiquities Authority) during deepening of the Akko Marina from 1992 to 1993 and finds from subsequent surveys at the sites where material dredged from the marina was deposited in the open sea. In addition, remains of recovered water crafts are briefly described, along with cargo and findings during underwater and coastal surveys conducted by the IAA in and around the Akko Marina from 1988 to 2004.

Akko (Acre) is a Mediterranean harbor city that has been functioning for thousands of years (Fig. 1). It is a unique site and was recently recognized by UNESCO as a World Heritage Site. Akko has been the scene of diverse human activities over the millennia. The bay just south of the city has served as a natural anchorage and harbor. The unique geo-hydrographic characteristics of Akko’s southern shoreline, protected from western and northern winds by the Akko headland, gave the place natural advantages as a safe haven for anchoring sailing vessels and as a base for the construction of a harbor. The Akko Harbor is one of the three ancient built harbors on the coast of Israel. This harbor served as a gateway for thousands of years, with vessels arriving regularly from all across the Mediterranean. The harbor served foreign armies, pilgrims, local and international commerce, and is mentioned in numerous ancient sources. The importance of the Akko Harbor is attested by continuous descriptions, with documentation of the harbor and its means of access, beginning on coins of the Roman period, and, from the thirteenth century, drawings, sketches, portolans, charts and written descriptions (Galili et al. this volume a). No similar abundance of navigational aids can be found for other ancient anchorages or harbors along Israel’s coast, such as Atlit (Athlit), Caesarea, Jaffa or Ashkelon (Ascalon), which also hosted maritime activity beginning in antiquity. Numerous archaeological studies of the city of Akko and its harbor have been undertaken over recent decades. Nevertheless, several archaeological and historical issues remain unresolved. For example, the date of construction and end of use of the harbor’s facilities; its boundaries in various periods; its array of piers and arrangements for mooring and anchoring vessels of various sizes; the location of the harbor’s entrance; problems of siltation of the harbor in various periods and the harbor’s destruction by human and natural agents. The harbor exerted a major influence on the social and economic life of the city and the answers to these issues can greatly assist in the understanding

1

Ehud Galili

Fig. 1: Location map and aerial photograph of the Akko Peninsula, view from the southeast (aerial photo: A. Ohayon) and Akko. The bay is bounded on the north by the Ahihud Fault, through which the Na‘aman River flows. On the south it is bounded by the northern edges of the Carmel range, through which the Kishon River flows (Galili and Rosen this volume e: figs. 1, 4). The Haifa Bay faces the open sea to the west, and during winter storms cannot provide proper safe haven for ships. The Haifa Bay is a sedimentary sink and the edge of the Nile littoral cell. The shoreline of the bay is sandy except for the southern shore of the Akko headland and Haifa’s Bat Galim Beach. During winter storms the Akko Harbor is exposed to the waves and currents from the southwest; anchoring there is dangerous in winter. During spring summer, and fall,

Finds uncovered until 2014 during archaeological activities by other entities in the Akko Marina are also discussed. Some of the findings can be attributed to discrete assemblages from shipwrecks or harbor installations; others were scattered without clear archaeological context. The Physical Conditions of the Akko Harbor and its Surroundings The Akko Harbor is located at the northern end of the Haifa Bay, which is the largest coastal embayment in Israel and the continuation of the valleys of Zevulun 2

The Akko Marina Archaeological Project – Introduction

commerce, and so it may be assumed that the city had a harbor or anchorage. Research cannot clearly indicate the presence or location of such early harbor; however, it was suggested that the Na‘aman River and its estuary served as an inland safe haven for ships (Raban 1993). Beginning in the Persian period, in the sixth–fifth centuries BCE, the city began to spread beyond the mound, and Phoenician Akko was the staging area for the Persian forces on their way to Egypt. According to Raban (1993: 29–31), the great Persian fleet and the logistical forces supporting it, could not make do with the shelter provided by the Na‘aman. It was assumed that the harbor was first built in its current location in the sixth century BCE for the benefit of the fleet of the Persian King Cambyses. According to Raban (1983: 146), the Phoenicians built another, open anchorage near the existing harbor, now known as the Island of Flies. At the end of the fourth century BCE the city became a Greek polis. Its location and its convenient harbor significantly increased its importance, and it took the name Ptolemais, after King Ptolemy (the Ptolemaic dynasty ruled Egypt during the Hellenistic period, 305 BCE to 30 BCE). It fairly quickly became one of the most important cities in the region. At its height, its jurisdiction extended southward beyond Mount Carmel and northward to the mountain ridge known as the Ladder of Tyre. Historical and archaeological evidence from the Hellenistic period reveal the importance and prosperity of the city. Greek inscriptions, an abundance of wine jars imported from Greece, consumed by non-Jews and soldiers, and the bones of elephants belonging to the Hellenistic army, were discovered in excavations. The recently uncovered Hellenistic-era residential quarter, with its wealth of colored plaster fragments, attests to a high standard of living. As an important and wealthy capital city, it was surrounded by fortifications. A city wall built of ashlars, 3 m thick, has been discovered near the present-day Tambour factory. On the slopes of the Western Galilee Mountains five fortresses have been discovered, which apparently protected Akko’s land flank. Near Moshav Regba an aqueduct was discovered, hewn in the kurkar ridge, which led water from the Kabri Springs. Only a major city with ample resources could have funded the construction of such an aqueduct in the Hellenistic period. Beginning at that time, the city began to mint its own coins, a right reserved only for important cities. Because of its importance, Alexander Jannaeus (Hasmonean king, died 76 BCE) imposed a lengthy siege on the city, but was unable to conquer it. During the Roman period, until the construction of the harbor of Caesarea (building started in 10 BCE), the Akko Harbor was the most important gateway to Europe and to Rome and its European territories. Akko’s strategic location and the importance of its harbor led to its renovation, including the raising and

however, the Akko Harbor provides good protection from northern winds. Southwestern summer storms are not particularly powerful and it may be assumed that the Akko Harbor could be used during the summer. The bottom of the Akko Bay is covered with pebbles, coarse carbonatic sand, and clays carried by the flooding of the Na‘aman River and fine quartz sand brought by the sea currents.West of the foundation of the ancient breakwater,the seabed is rocky and the Crusader walls are built on kurkar abrasion platforms – remains of the kurkar ridge that was mostly quarried for building stone. The basin of the modern fishing harbor is in same place as the ancient harbor once was, and the line of the modern southern breakwater was built mainly over the ancient breakwater (Figs. 1, 2). Because the basin is closed, there was low energy in it. This resulted in the deposition of clays and fine sediments over thousands of years. In these clayey deposits, which are several meters thick, are remains of ships, cargos, and numerous archaeological objects originating from ships that were wrecked in the harbor or were abandoned. These layers also contain refuse that was discarded or inadvertently fell overboard. East of the marina is a large sandy area (the easern basin) measuring 300 × 300 m, bounded on the east by a rampart that connected the Island of Flies with the shore. According to several scholars, these are the remains of the eastern breakwater from the Early Muslim period, (see summary Galili and Rosen this volume e). This area, which may have served as part of the Crusader and Mamluk harbor basin, was blocked by sand as much as several meters thick. East of the rampart is a sandy area with a beach facing southwest and extending toward the eastern beaches of the Haifa Bay (Galili and Rosen this volume e: figs. 1, 4, 9).

Historical Background The coast of Israel, which is ca. 196 km long, is mostly straight, with few natural safe havens for ships. Over some 5,000 years, this coastline served as an active shipping lane, constituting a bridge between civilizations. This coast witnessed intensive shipping, trade, fishing, wars, pilgrimage and other seaborne activities. The Haifa Bay, the largest in Israel, is exposed to the wind and therefore could not serve as a safe haven for sailing vessels during storms. The Akko Bay, located in the northern part of the Haifa Bay, could provide relatively good shelter, particularly during northern storms. Thus the sea opposite Akko, between the Na‘aman estuary and the southern part of the city, served as a major center of maritime activities for thousands of years. The archaeological findings in the excavations of Tel Akko (Tell el-Fukhar) revealed extensive trade relations beginning in the early second millennium BCE (Middle Bronze Age IIA). At that time, the existence of the city was already tied to maritime

3

Ehud Galili

extension of the southern breakwater. Roman Akko continued to be a key non-Jewish urban center, which existed alongside Jewish Galilee. The importance of the city can be gleaned from the works of Josephus Flavius (a Jewish military leader who was captured by the Romans and later became a historian; died in 100 CE). In 39 BCE, Herod (A Jewish Roman client king of Judea, 74–4 BCE) disembarked at Akko when he returned from Rome after being crowned by the Roman Senate (Ant. 376; Wars 290). After his return from Rome, it was in Akko that Herod raised an army of mercenaries of various ethnic origins to fight Antigonus (the last Hasmonean king of Judea, died 37 BCE; Ant. 394; Wars 290). In 30 BCE, Herod met Octavian (Augustus) in Akko and welcomed him grandly to a luxurious palace, when Octavian was on his way to Egypt to fight Anthony and Cleopatra after the battle of Actium (Ant. 385). Herod knew the strategic value of Akko as the gateway to Rome and to the entire Mediterranean, and it may be assumed that the fact that this important harbor city was outside his realm influenced his decision to build the harbor of Caesarea. It was at the Akko Harbor that the Roman legions landed to quell the Jewish Great Revolt in Galilee (66–67 CE); the banners of these four legions appear on a bronze coin minted in the city during the time of Emperor Nero. On a city coin the harbor is depicted as a semicircle surrounded by towers with ships entering it. That unique depiction of a harbor on a city coin is a clear evidence of Akko’s importance. Another coin shows a mound east of Akko, topped by a fortress to which a winding road ascends. At the bottom of the coin a ship is shown with another detail that appears to be a lighthouse (Galili et al. this volume a). Beginning in the reign of Nero (54–68 CE) the title Colonia was bestowed on the city, and after the Great Revolt, the Jewish population returned there and once again became involved in the city’s commercial and economic life. There were many public buildings in the city from the Roman period, among them a temple to Jupiter and the Capitoline Triad. During the Byzantine period, Akko was a regional capital with its own bishop, as well as a shipbuilding yard. Although not a single church from this period has yet been discovered, written sources reveal the names of a number of bishops who served there. With the Arab conquest (644 CE), the city went back to its original name, Akka-Akko, after about 1,000 years during which the foreign name Ptolemais was used. During the Early Muslim period, at the time of the Umayyad Caliph Mu‘awiya (reign, 661–680 CE), the Akko Harbor was very active. The shipyard that had operated during the Byzantine period was reopened and the harbor became the second largest Arab navy base in the Eastern Mediterranean after Alexandria. During the 880s CE, Ahmed Ibn Tulun, the Abbasid governor of Palestine and Egypt, expanded and enlarged the harbor in a manner resembling the harbor of Tyre to the north.

The Jerusalem historian Al-Muqaddasi described the work, which was planned and implemented by his own grandfather, the Jerusalem architect Abu Bakr. The Crusaders made Akko and its harbor their main naval base linking them to Europe. Between 1191–1291 CE, Akko was the capital of the realm and the nexus of Christian military and the center of economic and commercial activities. The city was the headquarters of the Crusader military orders and of the administration established by the Italian maritime powers in the Levant. Activity in Akko and its harbor during the Crusader era left behind an abundance of maps, documents and archaeological remains. According to historical descriptions, the city’s harbor during Crusader times was identical to the one from the era of Ibn Tulun, its main drawback being its small size (ca. 60,000sq m), which was insufficient, especially in the spring and summer. Due to siltation of the harbor large ships had to anchor outside the protected area near the Na‘aman estuary, and south of the southern breakwater. During the Mamluk and the early Ottoman periods (from the destruction of the city at the end of the Crusader period in 1291 CE and until the early seventeenth century), the harbor lay in ruins. However, there is evidence that it was used at this time by Venetian merchants for the export of cotton and the transport of pilgrims (Ashtor 1983: 379; Shor 1990: 161–163; Galili and Rosen this volume a) and for trade with Venice. Fakhr al-Din II (1572–1635, leader of the Emirate of Shouf) restored the harbor and renewed international trade, but in 1613, shortly after the renovation, he ordered the harbor to be demolished to prevent Turkish access to the city. During the 1740s the city was taken over by Dahar al-Omar (virtually autonomous ruler of northern Palestine, lived 1688–1775), who encouraged merchants, especially the French, to visit the city. According to eighteenthcentury documents, Dahar al-Omar built warehouses at the harbor and renovated the southern breakwater and the Island of Flies. The renovation was carried out using fieldstones bonded with cement into which ancient columns were inserted, to which ships could moor. In the second half of the eighteenth century the Akko Harbor flourished as an important Eastern Mediterranean trade center, frequented by Ottoman, Russian and French ships. In April 1799, the French expeditionary force under Napoleon Bonaparte reached the gates of Akko after conquering Jaffa and Ramla. The French army tried for a month to breach the walls of the city and conquer it, but to no avail. During the siege, the defenders of Akko were assisted by the British fleet under the command of ViceAdmiral Sidney Smith. This fleet effectively stopped any French attempt to bring in reinforcements by sea; it even captured a number of French vessels carrying cannons from the siege on Jaffa. In May, the French forces withdrew southward by land via the coast. Contemporary French sources provide an abundance

4

The Akko Marina Archaeological Project – Introduction

Fig. 2: Map of the various underwater surveys conducted by the IAA in and around the Akko Harbor, 1990–2012, and the main archaeological remains (IAA).

History of Research – Underwater Excavations and Surveys at Akko, 1960s–2004 Several underwater archaeological activities were carried out in Akko since 1960 by various researchers and institutions. Underwater archaeological activities can be divided into a number of categories: predevelopment rescue excavations and/or surveys (aimed at studying, documenting and, if possible, rescuing archaeological finds in areas intended for destructive, modern development); rescue excavations and surveys associated with the deepening of the modern marina by heavy dredging machinery; research excavations; and didactic excavations (aimed at training students in underwater archaeology). Below is a review of the main underwater surveys and excavations carried out in Akko from the 1960s to 2004. Underwater Pre-development Rescue Surveys and Excavations at the Southern Breakwater and Island of Flies

Fig. 3: Map of work areas of the Akko Marina project, 1992–1993 (IAA)

From 1964 to 1966, the Underwater Exploration Society of Israel conducted underwater excavation and survey prior to the construction of Akko’s modern marina. The expedition, headed by E. Linder, examined the foundations of the Island of Flies, the

of maps and descriptions of the array of British and French forces. In the mid-nineteenth century the harbor began to fill with silt and declined in importance.

5

Ehud Galili

underwater survey was undertaken during 1991-1992 by the IAA in the areas intended for development (Fig 2: I). A sonar survey, a metal detection survey and water-jet drillings were conducted as well. Remains of thirteen shipwrecks were detected and examined during the survey (Galili et al. 1991, 2002; Galili and Sharvit 2002; Galili this volume f: fig. 1) (Fig. 2).

southern breakwater and the remains of shipwrecks at the entrance to the harbor. Test excavations were conducted in 1966 at the eastern rampart (Fig. 2: 15) and southern breakwater (Fig. 2: 13) and at the Island of Flies (Fig. 2: 14). A sub-bottom profiler sonar survey and proton magnetometer survey were also conducted around Akko and the various phases of construction of the harbor were reconstructed (Raban 1982a, 1982b; Linder and Raban 1964; Flinder et al. 1993; Raban 1983, 1993). Two wooden hulls of shipwrecks were discovered and checked (Fig. 2: 10, 11).

Underwater Rescue Survey in the Akko Area 1990–2004 The IAA conducted year-round underwater rescue surveys whose purposes were to locate, document and rescue findings uncovered on the sea bed that were at risk of destruction or theft. The surveys were carried out along the entire

Training Excavation of a Shipwreck In 1975 E. Linder of the University of Haifa and J.R. Steffy of Texas A&M University conducted a training excavation (in the course of a field school) of a shipwreck on the north side of the Island of Flies (Fig. 2: 11). The wooden hull of the ship was uncovered, containing copper alloy nails, along with a large quantity of schist stones used as ballast. In 2006, the University of Haifa renewed the excavation of this ship. Underwater Research Survey of the Southern Breakwater In the 1980s, R. Gertwagen of the University of Haifa conducted an underwater survey of the eastern end of the southern breakwater. Another survey and a test excavation were undertaken in the area of the Pisan Quarter of the Crusader period, in the so-called “Pisan Harbor” (Fig. 2: III) (Gertwagen 1989, 1996). Underwater Rescue Surveys – the 1983 Deepening of the Akko Marina, 1991-1992 From 1983 to1988, during and after the deepening of the marina, an underwater survey was carried out by E. Galili of the University of Haifa and S. Wachsmann of the IAA. The survey, which examined the material removed by the dredger that deepened the marina, revealed a few dozens pottery and metal objects. Also documented were finds retrieved by fishermen-divers in the areas where the dredged materials were deposited. Underwater Surveys Prior to Planned Expansion of the Akko Marina In 1988, a plan to expand the Akko Marina was submitted for approval by the IAA. The plan included significant expansion of the marina to the south and east, including reclaiming underwater areas and the construction of a residential neighborhood. An

Fig. 4: Map of the underwater and coastal surveys by the IAA on the west coast of Akko (IAA)

6

The Akko Marina Archaeological Project – Introduction

Mediterranean coast of Israel and inland waters. Priority was given to areas where there was a high likelihood of exposure, destruction or looting of finds. During the ongoing surveys in the Akko area, remains of ships (mainly anchors, nails and lead sheathing), weapons, ammunition and cargos were found. In addition valuable information was gathered from fishermen-divers and others with regard to items found at sea during diving and fishing (Galili 1987; Galili et al. 2002, 2004; Galili and Rosen 2008a, 2008b).

mapping of quarries and ancient installations by means of differential GPS and manual measurement of the abrasion platforms. The surveys revealed two millstone quarries (Fig. 4: VII, 4: VIII), a quarry for building stones (Fig. 4: V) and small rock-cut pools (Fig. 4: VI). (Galili and Sharvit 2002; Sharvit and Galili 2002). Underwater Pre-development Survey – the Pisan Quarter Seafront, 2004 In order to protect the sea walls and ancient structures on the shoreline of the Pisan Quarter, ancient submerged structures were surveyed and mapped. The survey included bathymetric mapping of the seabed with the assistance of the Israel Electric Corporation, water-jet drillings, a surface survey and mapping of the remains of the ancient breakwater (Fig. 2: III, Galili this volume e). The surveys were by the IAA (Galili and Sharvit 2002; Galili et al. 2010: figs. 17.2, 18).

Underwater Rescue Survey prior to the Extension of the Southern Breakwater, 1991 Following a plan to extend the southern breakwater by 200 m, an underwater survey was carried out along the zone of the planned extension. The survey, which included scans of the seabed by divers and water-jet drillings, was carried out by the IAA (Galili and Sharvit 2002, Fig. 2: IV).

The Akko Marina Archaeological Project, 1992–1993

Underwater Pre-development Survey – the Pisan Quarter Seafront, 1993

The findings from this project form most of the material presented in this volume. The archaeological project accompanied the deepening of the Akko Marina from November 1992 to June 1993 and also included salvage surveys conducted by the IAA

Following a plan to build a bridge connecting the fishing harbor with the western sea wall promenade, underwater survey was carried out south of the Pisan Quarter (Fig. 2: III). The survey was carried out by the IAA (Galili et al. 2002). Underwater and Coastal Pre-development Rescue Survey, Southwestern Akko, 1994–1995 In the framework of planning tourist accommodations in Akko, the possibility of building a hotel on the abrasion platform and the ancient quarry north of the lighthouse was assessed (Figs. 2: V, 4: V). An underwater and coastal survey was carried out by the IAA (Sharvit and Galili 2002). Finds included ancient quarries, the foundations of city walls and fragments of walls that were built on the coastal abrasion platforms. In that area, according to historical documents and drawings, the castle of the Templar military Order was located during the crusader period (Galili and Sharvit 2002; Sharvit and Galili 2002; Galili this volume e). Underwater Western Akko Pre-development Rescue Survey, 2003/2004 A coastal and underwater survey was conducted on the kurkar abrasion platforms at the western and northern beaches ahead of construction of the beach promenade and the laying of drainage pipes. These surveys included

Fig. 5: The dredger that deepened the marina (J. Galili) 7

Ehud Galili

marked by numbered buoys. The area of the marina was divided into six main work areas, each of which included a number of squares (Fig. 3). A – east of the marina; B – northeast of the entrance to the marina; C – entrance to the marina; and D, E and F – inside the marina. The deepening of the marina was carried out by a bucket-dredger, based on a barge, which raised sediment from the marina bed and loaded it into a discharge barge (Figs. 5, 6). Preliminary examination of the material was done aboard the barge, and finds identified were collected. Each barge received an identifying number connecting it to the various excavation areas in the marina. The sediment, including the archaeological material, was deposited into pre-determined, rocky flat-bottomed locations, at a depth of 15–20m in the open sea. The dumping was planned to avoid ecological damage to the marine environment and according to a special permit issued by the Environmental Protection Ministry. The deposit sites were surveyed by divers and examined by a metal detector. Indicative findings were collected and taken ashore. A data base was prepared coordinating the work areas in the marina and the numbers of the barges depositing the material into the sea. Findings retrieved from the sea were soaked in fresh water for several days. They were then dried and marked with the number of the barge that had retrieved them, so the area of the marina were they were found could be identified. All told, the contents of some 80 barges were deposited in the open sea; approximately 100 dives were made at the deposit sites and some 30 dives in the marina. Divers mapped the submerged installations in the marina by means of measuring tape and compass, and the data was recorded on site. Later the data was transferred to the graphic journal. The mapping of the marina and the identification of the work squares and the excavation areas were made by compass binocular and a laser rangefinder, using known reference points on the shoreline. Buoys were placed at points where remains to be preserved had been identified, and dredging in those areas was halted.

Fig. 6: Preliminary examination of the material removed from the marina by a barge that ferried it to the deposit sites at sea (J. Galili) in the Akko region from 1990 to 2004 (see above) (Figs. 2: II, 3). The Akko Marina was deepened so that it could serve water craft with a draft of several meters (Galili and Sharvit 2002; Galili and Rosen 2008a; Galili et al. 2010). Some 15,000 cubic m of material were dredged from the marina on behalf of the Shipping and Ports Department and the Tourism Ministry. The IAA archaeological work, directed by E. Galili, included ongoing supervision of the dredging and survey dives during which ancient remains on the marina bed were located, mapped and retrieved. The material dredged from the marina was examined by divers at the dumping areas (see below). Rescue Excavations at the Foot of the Southeastern City Wall, 2008–2012 In order to expose and repair the foundation of the Ottoman southeastern sea walls of Akko, which were damaged by marine erosion, a 250m section was excavated at the foot of the walls (Fig. 2: VI). The excavations, carried out by the IAA, the Old Akko Development Company and Rhodes University, revealed installations associated with the Early Roman-Hellenistic harbor and shed light on the foundation methods of the Ottoman city walls (Sharvit et al. 2013).

Underwater Rescue Surveys These surveys, whose purpose was to document and salvage ancient remains exposed on the seabed and at risk of destruction or looting, were carried out mainly after winter storms. Priority was given to areas where the risk of exposure and looting was greatest. Pairs of divers working from a boat surveyed the seabed using a compressed air system, measuring equipment, sketching and photography, and, frequently, a handheld metal detector. Sites were initially located and mapped manually (see below) and, beginning in the 1990s, by GPS.

Methodology To map the location of the archaeological finds, the marina area was divided into 132 squares (10 × 10m each). Each group of squares marked a defined excavation, which was allotted its own separate deposit site in the open sea. The deposit sites were 8

The Akko Marina Archaeological Project – Introduction

Sonar Scanning of the Seabed Mapping

To locate the remains of ships, cargos and structures underwater, the project was assisted by the Israel Oceanographic and Limnological Research Institute using ground-penetrating sonar to scan various parts of the marina and its surroundings. The survey was carried out by a boat towing a submerged sonar transmitter that sent acoustic waves to the seabed. The reflected acoustic waves were received by a hydrophone and transferred to a recorder that recorded on paper profiles and anomalies that might be ancient structures or shipwrecks. Targets located were marked by buoys and the seabed at these points was then checked by divers and water-jet drillings (see below). Navigation of the boat and mapping of the profiles and the targets were carried out by means of a Mini Ranger with the assistance of two points on the beach where transmitters had been installed to broadcast signals to the receiver on the boat and record its location at any given moment.

Bathymetric underwater mapping was done by the Israel Electric Corporation’s Marine Biology Department, utilizing a differential satellite navigation system and sonar equipment synchronized by computer. The equipment was installed on a boat moving back and forth as it mapped the area south of the Pisan Quarter. The mapping served to plan the protection of the sea walls and document the submerged remains of the ancient breakwater in this area, which were in situ there and had not been covered by the stones of the modern breakwater. The coastal mapping of the sea walls and the ancient installations on the shoreline in the western part of the marina and in the Pisan Quarter were implemented using the differential satellite navigation system. Remains of shipwrecks and harbor installations in shallow water were identified by aerial photographs. Test Excavations

Underwater Electromagnetic Scanning

Small probes (1 × 1m) using water-jet drilling were made in and around the marina, particularly in places were remains were located under the seabed.

To locate large metal objects, such as cannons or heaps of sherds containing large ferrous concentrations, a proton magnometer was used operating together with ground-penetrating sonar. The equipment measured and mapped the magnitude of the magnetic field of the areas where the boat passed. A map of magnetic field values was produced showing magnetic anomalies, where the likelihood was high of finding large metal objects or cargos of pottery. The experimental stage revealed numerous anomalies in and around the marina, originating in scrap metal. These scraps made the work difficult; thus magnetic mapping was not fully implemented and it was decided to rely mainly on the groundpenetrating sonar and diving surveys.

Registration of Finds The finds retrieved during the deepening of the marina and during the surveys in the deposit areas off shore were marked with the number of the dumping barge, the number of the working area in the marina and the item number. The finds were also given categories according to groups demonstrating their function. A few items retrieved during surveys before or after the dredging project were marked by the usual method for underwater rescue surveys of the IAA; these were given the number of the dive report, which includes: permit number (top), report number (bottom left in a triangle) and item number (bottom right).

Water-jet Drilling This method utilized a metal pipe inserted into the seabed through which a jet of water was ejected. A pump on the diving boat pumped in sea water and conveyed it by means of a flexible tube to the metal pipe inserted by divers into the seabed. By the nature of the sediments emerging from the drilling and the meter mark on the metal pipe, the divers recorded the depth of penetration and the nature of the subsurface of the seabed at each drill site. Water-jet drilling was utilized at points where the ground-penetrating sonar identified targets that seemed to be ancient remains. This method was also used in areas where information was required about the seabed subsurface for planning of a system of structures intended to protect the sea walls and the ancient remains in the Pisan Quarter seafront.

Acknowledgments The author wishes to thank the Israel Antiquities Authority for the institutional support, M. Feinberg Vamosh for the English editing, D. Jacoby for reading and reviewing the manuscript and B. Rosen and R. Galili for their notes on the text. Thanks are also given to E. Stern and R. Tweig, and to Y. Zomer; T. Shabi and A. Mukhiri (supervision of the barge that removed the material from the marina). Beside the authors of the various chapters in this volume, the following individuals assisted in processing the material and preparing it for publication: D. Segal I. Carmi (C14 dating); N. Liphschitz (identification

9

Ehud Galili

Galili E. and Rosen B. 2008a. The Harbor (Acco). In E. Stern, H. Geva and A. Paris (eds.). The New

of wood). Assisting in the logistical aspects of the diving and in technical processing: A. Greenberg; D. Moskowitz; A. Yakobovich (divers): G. Neulander (recording of finds); S. Ben-Yehuda (drafting and drawings); N. Ze’evi and H. Tahan (drawing); Z. Neuman, M. Saltzberger, J. Galili, T. Sagiv and C. Amit (photography of finds); M. Heiman and L. Barda (GPS mapping); K. Abramson from the Israel Electric Corporation (GPS mapping and sonar survey); G. Amit from the Israel Oceanographic and Limnological Research Institute (side scan sonar, ground penetrating radar and magnetic survey). J. Moshe from the IAA; the divers A. Benizri, A. Kotzer and R. Schomus assisted in the locating of finds and remains of vessels uncovered in and around the Akko Marina. Hetz Hatzafon, Avi Ohayon and Google Earth provided the aerial photos.

Encyclopedia of Archaeological Excavations in the Holy Land 5. Pp. 1558–1561. Jerusalem.

Galili E. and Rosen B. 2008b. Marine Archaeology in Israel–Recent Discoveries. In E. Stern (ed.). The

New Encyclopedia of Archaeological Excavations in the Holy Land 5. Pp. 1925–1934. Jerusalem.

Galili E., Rosen B., Zviely D., Silberstein N. and Finkielsztejn J. 2010. The Evolution of the Akko Harbour and its Trade Links Revealed by Recent Underwater and Coastal Archaeological Research. Journal of Island and Coastal Archaeology 5: 191– 211. Galili E. This volume e. Structures and Installations in and around the Akko Marina.

References Ashtor E. 1983. Levant Trade in the Later Middle Ages. Princeton.

Galili E., Rosen B. and Zviely D. This volume a. The Akko Harbor in Numismatic, Epigraphic, Cartographic and Photographic Evidence.

Flinder A., Linder E. and Hall E.T. 1993. Survey of the Ancient Harbour of Akko, 1964–1966. In M. Heltzer, A. Segal and D .Kaufman (eds.). Studies in

Gertwagen R. 1989. The Southern Shore of the Old City of Akko – ‘The Pisan Harbor.’ C.M.S. News, Rep. No. 16. Haifa.

Galili E. 1987. Akko Port, Underwater Surveys. Hadashot Arkheologiyot 89: 59–60. (Hebrew).

Gertwagen R. 1996. The Crusader Harbor of Acre: Layout and Problems of Maintenance. In M. Balard (ed.). Autour De la Première Croisade, Actes du Colloque de la Society for the Study of the Crusades and the Latin East, Clermont-Ferrand, 22–25 juin 1995. Pp. 553–582. Paris.

Archaeology and History of Ancient Israel, in honor of Moshe Dotan. Pp. 199–225. Haifa.

Galili E., Dahari U., Sharvit J., Aviam M. and Stern E. 1991. Survey of the Akko Marina. Final Report. Submitted to the Akko Development Corporation by the Israel Antiquities Authority. Pp. 1–54. (Hebrew).

Linder E. and Raban A. 1964. Underwater Survey in the Akko Harbor. The Western Part of the Galilee and the Galilee Coast, Nineteenth Annual Conference of the Israel Exploration Society. Jerusalem. Pp. 180–193. (Hebrew).

Galili E., Sharvit J., Dahari U., Bahat-Zilberstein N., Finkielsztejn G., Stern E.J., Kool R., Kahanov Y. and Rosen B. 2002. Akko Harbor, Underwater Surveys. Hadashot Arkheologiyot 114: 13–16, 12*–15*. (Hebrew).

Raban A. 1982a. The Ports of Akko. In M. Yeda‘aya (ed.). Antiquities of the Western Galilee. Pp. 180–207. Haifa. (Hebrew).

Galili E. and Sharvit J. 2002. Millstone Quarry Along the Coast of Acre. ‘Atiqot 42: 73–78. (Hebrew).

Raban A. 1982b. Fragment of a Ship from Napoleon’s Siege on Akko 1799. In M. Yeda‘aya (ed.). Antiquities of the Western Galilee. Pp. 207–195. Haifa. (Hebrew).

Galili E., Raban A. and Sharvit J. 2002. Forty Years of Marine Archaeology in Israel. In H. Tzalas (ed.). Tropis VII, Proceedings of the 7th International Symposium on Ship Construction in Antiquity, Pylos, Greece, 1999. Pp. 927–961. Athens.

Raban A. 1983, Akko Harbor. Kardom 24–25: 141–146. (Hebrew). Raban A. 1993. Maritime Acco. In E. Stern (ed.). The

New Encyclopedia of Archaeological Excavations in the Holy Land 1. Pp. 29–31. Jerusalem.

Galili E., Zviely D. and Rosen B. 2004. The Akko Harbor and its Environs on Coins and Graphic Representations. Horizons in Geography 62: 130– 113. (Hebrew).

10

The Akko Marina Archaeological Project – Introduction

Sharvit J. and Galili E. 2002. Akko West – Coastal and Underwater Surveys. Hadashot Arkheologiyot 114: 11–13, 10*–12*. (Hebrew). Sharvit J., Planer D. and Bauxton B. 2013. Preliminary Finds from Archaeological Excavations along the Foot of the Southern Seawall at Akko, 2008–2012. Michmanim 24: 39–52. Shor N. 1990. The History of Akko. Pp. 161–163. Tel Aviv. (Hebrew).

11

Chapter 2

The Akko Harbor in Numismatic, Epigraphic, Cartographic and Photographic Evidence Ehud Galili (1), Dov Zviely (2) and Baruch Rosen (3) 1. Israel Antiquities Authority and Zinman Institute of Archaeology, University of Haifa, Israel. 2. School for Marine Sciences, Ruppin Academic Center, Michmoret and the Leon Recanati Institute for Maritime Studies, University of Haifa, Israel. 3. Israel Antiquities Authority. Key words: coins, lighthouse, nautical charts, navigation, marine archaeology, hydrography, Akko, navigation

Introduction

as local and international trade and it is mentioned by numerous historical documents (Ashtor 1983; Shor 1990). Akko’s importance as a harbor city is attested by a long and continuous literary tradition; the earliest sources mentioning the city harbor date from the Bronze Age. The earliest known pictorial-graphic depiction of the city harbor and its approaches appear on coins from the Roman period. Proto-charts of the harbor appear in the thirteenth century, continuing with illustrations, charts, maps and verbal descriptions up to the present. No similar abundance of written navigational aids describes any other ancient harbor along the coast of Israel, such

The Akko waterfront is protected on the west and north by the Akko headland. As such the Akko Bay has seen widespread human activity for thousands of years, with the bay south and east of the city serving as a natural anchorage and later, a harbor. Akko is now an urban complex consisting of a walled Mediterranean city and an adjacent harbor (Fig. 1). The Akko Harbor served as the gateway for millennia, visited by ships from all across the Mediterranean. The harbor served foreign armies and pilgrims as well

Fig. 1: A-I: Sites mentioned in the text (aerial photo: modified after A. Ohayon) 12

The Akko Harbor in Numismatic, Epigraphic, Cartographic and Photographic Evidence

as Atlit, Caesarea, Jaffa or Ashkelon. Recent decades have seen numerous archaeological studies of the city and its harbor. However many archeological and historical questions concerning them remain unanswered. Among these are: When were the harbor’s facilities built and when did they fall out of use? What were the layout and function of the harbor in various periods? Among the unknowns are the location and the array of piers and the anchoring arrangements of ships of all sizes, the location of the harbor’s entrance and the silting processes in various periods. Also not completely known is how the city’s inhabitants dealt with the deterioration of the harbor by human or natural agents in various periods. Resolving the above questions can help in understanding Akko’s history and the impact of the harbor on its society and economy. The graphic depictions presented in coins, anchoring charts, coastal panoramas, maps and aerial photographs of the harbor and its vicinity often reveal details of ancient structures and facilities that have disappeared. As such they complement the archaeological research and may help resolve some of the key issues associated with the maritime activity in Akko. To navigate safely to an anchorage, a ship’s captain needs reliable information, obtained today through electronic, satellite-guided navigation equipment, nautical charts and pilot’s books. In the past, as a means of disseminating information and as an aid for navigation, the port area was depicted on coins, written documents, illustrations, plans, maps and nautical guides. The production of these aids met the needs of sailing and anchoring vessels using the harbor, carrying warriors, merchants, tourists and pilgrims. The graphic depictions of the Akko Harbor may be divided into five main groups according to the order in which they appeared as far as we know. 1. The earliest graphic depictions of the Akko port appear on Roman coins (Galili et al. 2004; Rosen et al. 2011; Galili et al. 2010). 2. Illustrations of the city and the harbor are known from the late thirteenth century onward (Dichter 1973; Zviely et al. 2003; Galili et al. 2004). These are not on a standard scale, and have no cartographic grid or indications of water depth. 3. Harbor plans are known from the early seventeenth century, but seem to have been in use before this. They were apparently sketched by various mariners (Rosen 1993; Zviely et al. 2003; 2010). The sketches often note the ship’s anchoring point and the depth of the surrounding water, nearby ports and prominent coastal landmarks to help the ships navigate. However, these plans do not have a cartographic grid and they cannot be linked to a broader map system. 4. Coastal views of the city and the area of the harbor, as seen from the sea, which continued to appear in modern maps and pilot books (Zviely et al. 2003; 2009b; 2010). 5. Nautical charts, which started appearing in the

eighteenth century and provide essential updated hydrographic information important for the safety of ships and also contain information on coastal characteristics and ancient structures (Rosen 1993; Zviely 2006; Zviely et al. 2009a; 2009b). 6. The first aerial photographs of the Akko area, taken in the early twentieth century, and subsequent photos provide a source of information about the harbor and its facilities at the time (Zviely 2006; Zviely et al. 2009a). This article surveys the development of the graphic depictions of the Akko Harbor and its approaches until the end of the twentieth century. It also presents and analyzes archaeological and historical information that can be gleaned from these depictions regarding access to the harbor, its functions and facilities.

The Akko Harbor on Coins from the Roman Period Two-dimensional graphic depictions of the geographical structure of the current Israel appeared as early as the Roman period. Mapping of this region from the Byzantine period, including nautical information, has survived in the Madaba Map (Avi-Yonah 1954). Probably in the Roman period, as in later times, iconographic information was accompanied by written sources, and both served seafarers. As early as the Greek period, sailing instructions appeared in the form of the periplus for coastal sailing and the diaplus, used on the open seas (Kadmon 1982: 28). Map-like information also appears on Roman coins. Although coins do not equal maps in terms of their descriptive value, the details depicted shed light on how in those times three-dimensional graphic information was conveyed in a two-dimensional medium. The information was frequently factual and useful, and was disseminated to all interested parties in thousands of copies. For example, on a city coin of Shekhem, Mount Gerizim, the acropolis of the city, appears along with the temple at the top of the mountain and the road that led there. The information conveyed on the coin has been confirmed by archaeological excavation (Hill 1914: Pl. 5). Nautical information appears on a coin of Antoninus Pius (reign 138–161 CE), depicting the lighthouse of Alexandria in Egypt (Shaw 1972). A coin of Elgabalus (reign 218–222 CE) from the early second century CE (Fig. 2), shows the Akko Harbor as a semicircle surrounded by boathouses or warehouses similar to those found in the excavations of the inner harbor of Caesarea (Raban 1993). The coin shows a ship anchoring in the harbor center and a small boat at the harbor entrance (Meshorer 1984: 14, 15). An inscription surrounds the boat, reading COLONI PTOLM, referring to Akko’s Roman name, Colonia Ptolomais. That harbor depiction is the only one known on coins of Roman Palestine and it attests to Akko’s importance during that period. This coin, in

13

Ehud Galili, Dov Zviely and Baruch Rosen

Fig. 2: A Roman coin of Elgabalus, showing the Akko Harbor, within which a sailing vessel is shown with furled sails (after Meshorer 1984: 14)

Fig. 3: A Roman coin of Elgabalus, showing the acropolis of the city of Akko: on the top of the coin, to the right, is a tall tower with images on its top, presumably a lighthouse, and at the foot of the coin is a ship and a small lighthouse (after Rosen et al. 2011)

circulation during the period in which it was minted, conveyed to a wide variety of people who handled it that the city of Akko had a harbor and that sailing vessels could anchor there for safe haven, repairs and unloading and loading merchandise and passengers. Another coin of the city of Akko, minted at the beginning of the second century CE by Elgabalus (Fig. 3), clearly shows the acropolis and the road leading to it. Moreover, a ship appears at the bottom of the coin (Fig. 3: B) and an object that Meshorer believed was an altar (1984: 12, 14) (Fig. 3: A). It was recently proposed that the object on the bottom of the Elgabalus coin represented a small lighthouse that marked the entrance to the harbor and that a tall tower on the top right of the city acropolis represented another lighthouse (Galili et al. 2010; Rosen et al. 2011). The identification of this structure as a lighthouse rather than an altar is based on a number of assertions: 1. Useful information of a maritime nature is presented on the above-mentioned Roman coin from Alexandria and thus it is reasonable to assume that similar information is depicted on the coin of maritime Akko. 2. Lighthouses in Roman harbors (as described above) were often depicted on contemporaneous coins and mosaic floors (Figs. 4, 5). 3. The depiction of this object on the city’s seafront together with a ship underscores the claim that this was a nautical apparatus and not an altar. 4. A lighthouse, built after the Roman period, indeed stood on the Island of Flies at the entrance to the Akko Harbor. This lighthouse (al-Manara in Arabic) was renewed during the Ottoman period, illustrating the need for this nautical aid at that point in the harbor

Fig. 4: Roman coin of Alexandria depicting a lighthouse (after Rosen et al. 2011)

entrance (Fig. 6). From these coins we learn that Akko had a closed built harbor with boathouses and/or storehouses and a lighthouse. Written information from the Byzantine period (Dan 1972) confirms some of the information about the maritime structures in the Akko Harbor conveyed by these two Roman coins. 14

The Akko Harbor in Numismatic, Epigraphic, Cartographic and Photographic Evidence

improved to meet the needs of navigation and anchoring in the Akko area. The illustration attributed to Matthew Paris, executed before 1259 (Fig. 7) (Dichter 1973: 14; Jacoby 1979: Pl. 7), depicts the city of Akko as a rectangle surrounded by a thick wall, two of whose sides (west and south) border the sea. On the left side of the illustration the city’s northern suburb appears, outside the thick wall (Fig. 7: A). This suburb was surrounded on its northern and eastern sides with a smaller wall and was not protected on the west by a sea wall. Near the northwestern corner of the city wall, a vessel is depicted without mast or sails, having a two-armed admiralty anchor and a narrow rudder (Fig. 4: B). Another vessel appears near the northwestern corner of the city, without mast, sails, rudder or anchor, seemingly moored to the wall (Fig. 4: C). Anchoring outside the harbor, without the ability to maneuver, would have been dangerous for ships, and it may be surmised that the ship’s riggings had been dismantled for maintenance. Assuming that the drawing is complete, it appears that the ships depicted in the Paris illustration had not anchored inside the harbor either for lack of space, too-shallow water or to save anchoring fees. The illustration shows three additional ships in the area of the harbor, sailing northward with their sails unfurled. The drawing is general, not to a standard scale, with the city of Akko more prominently shown than other areas of the coast, and depicting a number of important Crusader public buildings by name. At the end of the early fourteenth century, just prior to the fall of Crusader Akko, various illustrations of the city and its harbor began to appear. Marino Sanudo of Venice was involved in preparations for a crusade to the Holy Land and in that capacity he wrote a treatise advocating a crusade to the Holy Land (Liber secretorum fidelium crucis). During an early visit to Akko in 1285 or 1286, he collected sketches and verbal descriptions of Akko and its harbor, which he gave to Pietro Vesconte, an illustrator, who executed the drawings known as the Sanudo Maps. Most scholars agree that these illustrations, which have been described and discussed by several scholars, among them Dichter (1973), Jacoby (1979), Prawer (1991) and Gertwagen (1996), were copied from a previous source or a small number of sources. There are differences among the Sanudo illustrations; however, the basic lines reiterate themselves in various versions. Some of these differences probably stem from copyist errors, while others are the result of additions and of the illustration based on information conveyed to the artists by visitors to Akko. None of the pictures show harbor facilities and no harbor is seen on the city’s west coast. However, various maritime facilities appear from the southern corner of the Akko headland eastward. A number of illustrations show a circular or elliptical symbol in the southern part of the city of Akko beside

Fig. 5: Mosaic floor from Ostia depicting various types of lighthouses, including a small one similar to the one appearing on the Akko coin (After Rosen et al. 2011: fig. 3)

Fig. 6: The modern and ancient lighthouse on the Island of Flies at the entrance to the Akko Harbor (D. Zviely)

Illustrations of the City and Harbor of Akko from the Crusader Period and Thereafter The earliest illustrations of Akko in the Crusader period are from the thirteenth century (Galili et al. 2004). These include useful nautical information concerning the coastline contour, the settled area near the harbor, possible anchorages, and natural and manmade navigational obstacles. Mariners who navigated from Europe directly to Akko lacked sufficient prior knowledge of the area, leading to a real need for navigational aids. A nautical guide compiled in the second half of the thirteenth century provides detailed instructions for ships entering the Akko Harbor (Motzo 1947). A study of nautical illustrations found in various archives and collections in Israel and abroad, shows that from that time on, the quality of these maritime drawings gradually

15

Ehud Galili, Dov Zviely and Baruch Rosen

Sea wal l

N

Akko

Sea w

all

Akko Bay

Fig. 7: An illustration of Akko made before 1259, attributed to Matthew Paris (modified after Dichter 1973: 13) 16

The Akko Harbor in Numismatic, Epigraphic, Cartographic and Photographic Evidence

the sea. Some drawings show a narrow connection between this symbol and the open sea (Dichter 1973: 20) (Fig. 8: A), while in others, no such connection to the sea appears (Dichter 1973: 23) (Fig. 9: A). According to the illustrations, this symbol is located at the Pisan Quarter seafront, near the remnants of the foundation of the ancient breakwater (Figs. 1: 18). According to several scholars, the Pisans had a small external harbor here with an opening to the south that functioned in the summer (Linder and Raban 1965: 192; Raban 1983). Others have suggested that this area was a colonnaded plaza, close to the southern wall of the Pisan Quarter, and the place could not have served as an anchorage or an outer port (Kesten 1993). No historical evidence exists for such an outer harbor in this place, and there are no archaeological remains of such a facility. There is a segment of Crusader wall at this point, built atop remains of the early breakwater (Fig. 1: 18). This wall is part of a complex of structures and vaults whose remains have been found in the Pisan Quarter seafront west of the harbor (Galili et al. 2002; Galili and Rosen 2008). The seabed in this area is shallow and rocky and the area is exposed to south-southwestern winds. In surveys carried out in 2004 during a construction of a protective sea wall in the city southern part, remains of the early breakwater were documented on the seabed south of the Pisan Quarter (Galili this volume e: fig. 11). These surveys revealed that access from the Pisan Quarter to the sea was blocked by the massive remains of the breakwater of the ancient harbor. In light of the obstacles to access from the sea created by the ancient breakwater remains and the sea wall, there is no possibility that an external harbor or anchorage open to the south and serving the Pisan Quarter existed there. It is illogical that where a high, thick sea wall stood whose purpose was to protect the city, the Pisans would have created a breach for an external harbor. Such a breach could have endangered the city and exposed it to a naval attack from the south. Some Sanudo illustrations and subsequent depictions show a small bay and a kind of inner harbor north of the shore side of the southern breakwater (Figs 8: B, 9: B, 10: B, 11: B). Some believe that the place, whose remains are now located under Khan al-Umdan, served as an inner harbor, which was approached from the main harbor (Fig. 1: 25) (Flinder et al. 1993: 223; Raban 1983). Others, based on Linder and Raban’s research (1965: 198) contend that there was no inner harbor there (Jacoby 1979: 13). Based on Sivan’s research (1981) Gertwagen (1996: 561), claims that there could not have been an inner harbor at that location because Khan al-Umdan is constructed on solid kurkar rock. However, Sivan (1981) stated that no drillings are available from the area of the old city. To ascertain whether Khan al-Umdan is built on a site that served as an inner harbor during the Crusader period, drillings and archaeological excavations must be undertaken there. Recent archaeological excavation adjacent to Khan al-Umdan revealed no

Fig. 8: Illustration by Marino Sanudo showing the southern part of the city of Akko and a body of water connected to the sea (modified after Dichter 1973:29, from the British Museum, Catalogue No. 27376)

Fig. 9: An illustration by Marino Sanudo showing the southern part of the city of Akko and a body of water that is not connected to the sea (modified after Dichter 1973: 22, Riccardian Library, Florence, Italy)

17

Ehud Galili, Dov Zviely and Baruch Rosen

bedrock, indicating the possibility of a connection to the sea. A tunnel leading from the area close to Khan al-Umdan to the western shoreline in the Templar Quarter, where the fortress was located, may provide an indirect evidence for a connection to the harbor. The tunnel, flooded by city drainage and sea water, was used to move rafts bearing goods from such an inner harbor to the seaside fortress on the northwest shore. Grooves cut on the sides of the tunnel allowed portions of it to be dammed to allow quarrying below sea level (Galili et al. 2010; Galili and Rosen this volume e). Akko’s southern breakwater appears in the Sanudo illustrations in several versions, most of which also show a tower in the sea. In some of the drawings the tower is not connected to the shore (Dichter 1973: 23) (Fig. 9) and in others a number of dark segments are seen between the tower and the beach. These could N have been reefs, remains of buildings or a breakwater (Dichter 1973: 20–22) (Fig. 8: C). In one illustration, executed in 1323 by Paulino Veneto, a contemporary of Sanudo, the tower is connected to the beach by a number of hatched lines (Jacoby 1979: fig. 2) (Fig. 10: A). According to Jacoby (1979: 14), this crosshatching was intended to represent a chain preventing the entry of vessels to the harbor. Jacoby also notes that a chain at this point does not make sense. It is suggested that the crosshatching between the tower and the southeastern shore of Akko is not a chain, but a line of reefs, the remains of the ancient breakwater. This would mean that during the Crusader period the southern breakwater of Akko was breached, parts were sunken and parts appeared on the surface of the water as reefs. Support for this supposition appears in a written description of the entrance to the harbor in the thirteenth century: “The entrance to the harbor was between Said tower [the Tower/Island of Flies] and the rocks” (Fig. 1: between 13 and 14) (Motzo 1947; Taylor 1971:106–109). These rocks were apparently the remains of the early southern breakwater (Fig. 1: 13). Some illustrations show the tower at the eastern end of the line of reefs (the southern breakwater) in a prone position with its top facing north (Dichter 1973: 18, 20, 22, 23; Figs. 8, 9). In later illustration the tower is standing (Dichter 1973: 24, 40, 41; Fig. 11: A). Based on various sources preceding the construction of the modern southern breakwater in 1966 (Fried, Glaswein, Werner Engineers 1957; Linder and Raban 1965; Zviely et al. 2003: fig. 15), it seems that the eastern end of the early breakwater had an additional, north-facing extension (Fig. 1: 26). This extension also appears in aerial photographs and in bathymetric maps; it is a rampart consisting of small stones and few fragments of columns in secondary use (Linder and Raban 1965). The highest point of the rampart is at a depth of 1.0–1.5m. It is reasonable to assume that at Sanudo’s time illustrations were copied and

N

Fig. 10: Illustration of Akko by Paulino Veneto from 1323, showing, according to Jacoby (modified after Jacoby 1973: 2), a tower connected to the coastline by means of a chain

Fig. 11: Drawing of Akko by Augustine Antoine Calmet from 1720, in which a tower stands on the southern breakwater of the Akko Harbor (modified after Dichter 1973: 40)

18

The Akko Harbor in Numismatic, Epigraphic, Cartographic and Photographic Evidence

does not mention construction of a breakwater, but rather a structure that was connected to the southern part of the city wall (Shor 1990) built of segments standing on pillars with spaces in between them. Trial excavation at the eastern rampart showed no construction remains. Thus it was proposed that the structure described by Al-Muqadassi was the northfacing extension from the eastern end of the southern breakwater (Linder and Raban 1965). Flinder et al. (1993) noted that there is some doubt as to the purpose of the eastern rampart, claiming that there is no certainty that this was an artificial marine structure. It is proposed that there was no need for a breakwater on the eastern side of the harbor because the distance between the harbor and the eastern coastline of the bay is small and does not allow high waves to develop when eastern winds are blowing. Based on the archaeological finds discovered so far it seems that this rampart was used as a causeway/temporary road during construction works on the Island of Flies, like the one that leads to the sea castle in Sidon (Fig. 12). Additionally the rampart could have served to block entry of ships from the east to the anchoring basin (Galili and Rosen volume e).

the artists, who never saw Akko, assumed that the addition facing north was a tower. Therefore they drew it mistakenly as such, first, as a horizontally drawn tower (Figs. 8, 9) and later, as a standing tower (Fig. 11: A). Thus the existence of a tower at the end of the southern breakwater is dubious. In 1291 Akko was destroyed by the Mamluks. Many scholars believe that the city was razed and abandoned for centuries thereafter. Some also contend that the harbor facilities surviving after the Crusader period were probably demolished to prevent the Christians from returning (Ashtor 1983). However pilgrims landed there during the fourteenth century and merchants from Venice and Damascus used it to export cotton to Europe (Ashtor 1983; Shor 1990: 161–163). Archaeological findings recently unearthed at Tel Yassaf north of Akko attest to commercial ties with Venice soon after the Crusader period, apparently connected to the cotton trade (Stern 1999). In underwater surveys carried out by the Israel Antiquities Authority from 1992 to 1993, wooden pilings were discovered on the seabed northeast of the eastern end of the southern breakwater. These could have been used as the foundation for some sort of marine installation (Fig. 1: 19). The wooden pilings were dated by C14 to the years 1425–1461 (calibrated). These pilings may have formed part of a pier or a dock built near the eastern end of the southern breakwater, which enabled larger vessels to dock in relatively deep water (3–5m) and transfer merchandise and passengers via lighters (Galili et al. 2002; Galili and Rosen this volume a). The prominent object north of the breakwater, seen in some later editions of the Sanudo illustrations, may have been a depiction, based on a later source, of such awooden pier or similar installation that existed on the southern breakwater tip. Based on Sanudo drawings, the historical sources and the archaeological findings at Tel Yassaf, it is proposed that contrary to Atlit, Caesarea, Jaffa and Ashkelon, activity at the Akko Port persisted after the Crusader period. In some of the Sanudo drawings the Tower of Flies appears as a structure or tower located at the southern end of a rampart/wall built in the sea as a continuation of the eastern city wall (Figs. 8, 9, 10, 11). According to Gertwagen, the submerged structure that can be seen in aerial photographs (Fig. 1: 15) extending from the southeastern city wall southward into the sea was built by Ibn Tulun during the Early Muslim period as a breakwater that protected the harbor from strong eastern winds (Gertwagen 1996: 568– 570). According to Linder and Raban (1965), the purpose of this structure was to keep sand out of the harbor, or, it may have been a part of the eastern harbor that was never completed (like the eastern harbor in Tyre, in Lebanon). In 985 CE, the Muslim historian Al-Muqadassi describes the construction of the harbor (Galili this volume e). The description

N

Fig. 12: The harbor of Sidon in southern Lebanon. At top right: the sea castle and the bridge connecting it to the coastline (the general outlines of the Akko Harbor can be seen in mirror image) (modified after Google Earth)

Harbor Plans Nautical literature from the fifteenth to the nineteenth centuries includes plans of the Haifa/Akko Bay, which constituted part of compendia of plans that were in use along the Mediterranean coast (Rosen 1993; Zviely et al. 2003; 2010). Generally no other anchorages along the Israeli coast appear in these compendia. The plans provided information to mariners about access to the coast and anchoring zones in the area of these

19

Ehud Galili, Dov Zviely and Baruch Rosen

harbors, as was the usual practice in that period. Only a few of these compilations, based on observations by visiting mariners and early schematic drawings, have survived. The plans do not show global coordinates and their cartographic characteristics are very basic and schematic. Most of them started appearing prior to the nineteenth century in atlases by J.N. Bellin and J. Roux (Rosen 1993; Zviely et al. 2003; 2010). Below we discuss six of these harbor plans depicting the Haifa Bay, often called the Akko Bay.

East of the city a large crescent-shaped form can be seen (perhaps the eastern rampart), connected to the coast and facing southwest. Between the Akko headland and the crescent is a protected, almost enclosed area of water, connected to the open sea by a narrow passage. From the southwestern end of the peninsula, an elliptical shape appears very prominently, extending south for hundreds of meters (Fig. 13: A). At the center of this element there appears to be a depth marking. This protrusion seems to depict a known submerged reef, which constituted a navigation hazard and also appears on modern maps (Zviely et al. 2003). Remains of lead sheathing and nails from Roman-era vessels discovered on this reef show that vessels with drafts of 2.5m and hit the rock there (Galili et al. 2002). A few hundred meters south of the entrance to the harbor an anchor symbol appears, apparently marking a point outside the harbor where large vessels could anchor. In the harbor basin a single anchor appears, and in the open sea, west and southwest of Akko are three anchor symbols, at a point where the sea is 25–35m deep. The city of Haifa (Caiffe) is depicted as a group of circles near the beach, in the center of which is an inlet south of which is yet another anchor symbol – a recommended anchorage. In the southern part of the Haifa Bay near the coastline an inner, enclosed body of water can be seen, separated from the coast by a narrow strip of land. This body of water may have been a coastal lagoon or a large swamp that existed at the time in the area of the Kishon estuary.

French Harbor Plan of Akko from 1707, Attributed to Jean Olivier This plan, called “Plan de la Rade de S: Jean d’Acre il de Caiffe il l‘Andtoit du bon Mouillage,” numbered 64 (Fig. 13), is earlier than many other Akko Harbor plans. The area covered by the plan extends from Mount Carmel (“N.D. de mont Carmel”) northward to Rosh Haniqra (Cap Blanc). The map shows a compass rose with the four cardinal directions and a fleur-de-lis design pointing north. At the edges of the map is a scale in Toize units (Zviely et al. 2003; 2010). Akko (S. Jean d’Acre) is depicted as a group of circles covering the area of the Akko headland.

The Akko Harbor Plan from Malta, ca. 1760, Apparently Drawn by Antonio Borg The British National Maritime Museum, Greenwich holds a collection of Mediterranean harbor plans apparently drawn by Antonio Borg from Malta. A copy of one of the plans in the collection (No. 100) is held by the Laor Library of the Hebrew University of Jerusalem. It is called “Chaifa E S.Gio: DAchiri” (Fig. 14). The plan covers an area from the Carmel coast to north of Akko and the coast of western Galilee. Haifa Bay is shown surrounded by mountains, with the cities of Haifa and Akko on the coast. The plan shows a compass rose divided into 32 directions and a fleur-de-lis pointing north. The scale at the edge of the plan is shown in Italian miles (“Scala d’un miglio Itagliano”). The text accompanying the plan mentions three places – Akko (S.Gio: Dachiri or S: gio: D’acri), Haifa (Caifa or Chaifa) and Atlit (Castell Pellegrino). Two settlements are marked on the plan and their spelling in the titles and in the body of the texts differ. The name Atlit, mentioned in the text, is absent from the plan. In the area of the Akko headland three elliptical forms are shown in the sea. Two are connected to the shoreline, perhaps depicting the southern breakwater and the eastern rampart of the

Fig. 13: A French harbor plan of Akko from 1707, attributed to Jean Olivier (modified after Kedar 1997: 166)

20

The Akko Harbor in Numismatic, Epigraphic, Cartographic and Photographic Evidence

N

N

Fig. 15: French harbor plan of Akko, presented in an atlas by Josef Roux, 1764 (modified after Archive of Historic Sites, Hebrew University of Jerusalem) At some distance southwest of the end of the rampart a small island is shown (apparently the Island of Flies). Between the southern part of the city and the rampart, two small, adjacent inlets can be seen, marked with four small crosses, signifying a hazardous area for sailing vessels. The eastern inlet may mark the Akko Harbor, and the western inlet the place known as the Pisan Harbor (see above). South of the southwestern end of the Akko headland a circle appears with closely packed points (Fig. 15: A), apparently marking the location of the submerged reef. West and southwest of the city of Akko, next to the shoreline, an area of the sea is shown, demarcated with a dotted line. This is apparently the broad abrasion platform located north of the Akko lighthouse. South of Akko three anchors, shown in the water, mark recommended anchorages outside the harbor. Many depths markings appear in the area of those points and near the harbor, in units of brasses (1 brasse = 1.62m, the length of two outstretched arms). The city is depicted as a group of buildings, spires and minarets, and east of the city, a fortress can be seen, apparently on the site of Tel Akko (Fig. 15: B).

Fig. 14: A plan from Malta depicting the Akko Harbor, apparently drawn by Antonio Borg, 1764 (modified after Archive of Historic Sites, Hebrew University of Jerusalem) Old City, or the remains of these marine structures. Between these two shapes is a relatively protected area of water, connected to the open sea by a narrow passage, delineating the Akko Harbor. The third and southernmost form (Fig. 14: A) is not connected to the shoreline and seems to mark the rocky submerged reef, marked on modern maps as well (Zviely et al. 2003). A single anchor symbol seen a few hundred meters south of the harbor marks a recommended anchorage. Also seen in the vicinity of the harbor are a number of depth markers, apparently in fathoms (1 fathom =1.83m). Farther back from the Akko coastline are two large structures with tall mosque towers alongside fortified towers, which apparently served as navigation and orientation landmarks for mariners.

Copy of Harbor Plan, 1770–1774 The Dichter Collection of maps at the University of Haifa holds an Italian harbor plan dated to 1770. It seems to be an imprecise copy of a harbor plan of the Haifa Bay appearing in the Roux Atlas (Fig. 15). This plan, and others like it, e.g., a French plan dated to 1773 (Dichter 1973: 153) and a German plan from 1774 (Carmel 1969: fig. 3), depict the coastline from south of the Carmel promontory to Rosh Haniqra, including the Haifa and Akko harbors. However, this plan is less detailed than the Roux plan (Zviely et al. 2003; 2010).

French Harbor Plan in the Josef Roux Atlas, 1764 The Laor Library holds a copy of the Roux Atlas showing a harbor plan (Pl. 119) of the Haifa Bay, entitled “Golfe de Caiffe Sirie” (Fig. 15). The area depicted on the plan extends from the promontory of Mount Carmel to Rosh Haniqra (C: Blanc). The map’s compass rose is divided into eight directions (the orthodromic lines) based on the north, which is marked by a fleur-de-lis. At the edges of the map the scale is shown in units of lieues (leagues). At the eastern end of Akko a long, broad rampart is depicted jutting southward into the sea (possibly the eastern rampart).

21

Ehud Galili, Dov Zviely and Baruch Rosen

by a wall, and east of it, extending into the sea is formation shaped like an upside-down Y (the eastern rampart?). Between this formation and the end of the wall of Acre, a small bay is seen with a broad opening to the south. From the southern part of the Akko headland, an elongated formation juts into the sea to the southwest – indicating the rocky reef depicted in the plans described above. South of the city of Akko and east of the rocky reef, a small, dark spot is shown in the sea, probably the Island of Flies. No symbol is shown for the southern breakwater. South of Akko is a large concentration of depth markers shown in fathoms. Three anchors are also seen, marking recommended anchorage spots. The depths in the area of the anchorages are delimited on the south by the words “The Road” (short for “roadstead” marking an area of the sea where ships can anchor), at a depth of 3–7 fathoms (5.5–12.8m). The Kishon and the Na‘aman streams can be seen along the eastern shoreline of the bay. Although the Luffman harbor plan provides new information, it was probably copied from earlier sources and numerous imprecise elements show that the compiler never visited the area (Zviely et al. 2003; 2010).

Fig. 16: British harbor plan of Akko from 1799, published by Luffman (from the private collection of Baruch Rosen)

Coastal Views Coastal views from the sea (panoramas) are important navigational aids based on the coastal formations and are also used in modern navigational charts. Some of these sketches look like artistic depictions of the coast. In the series of views from 1686 (Kedar 1997: 165), the coastline of the Akko promontory is shown, from the southwestern end through the area of the harbor and to the southern beach located east of the ancient city wall. In the western part of the series of views (see the upper series, Kedar 1997: 165, fig. 7), the houses of the city are seen along with monumental structures and the harbor area (Fig. 17). From the southeastern end of the Akko headland a line of low rocks is shown, apparently the remains of the southern breakwater (Fig. 17: A). These views indicate that in this period there were no structures on those rocks. North of the line of low rocks, a protected area can be seen in which two mediumsized ships, 15–25m long, are anchored (Fig. 17: B). These vessels have one mast and an array of furled

British Harbor Plan, Luffman, 1799 The Luffman Harbor Plan (No. 17) entitled the “Bay of Acre,” is part of a collection of plans of important cities, ports and fortresses worldwide. It was published by John Luffman in London in 1799 (Fig. 16). The plan shows the area from south of the Carmel promontory to north of Rosh Haniqra. Its compass rose is divided into 16 directions, with a fleur-de-lis pointing north. At the edge of the map, a scale appears in miles (apparently nautical miles, each of which equals 1.853 m). The text at the bottom of the plan includes, for the first time, longitude and latitude: 39.25 degrees longitude east of London; latitude 32.40 degrees north. The above coordinates are more or less on the same latitude as the Haifa Bay, albeit four degrees longitude east of it. The Luffman plan shows the city of Akko surrounded

W

E

Fig. 17: Coastal view of Akko from 1686 (modified after Kedar 1997: 165)

22

The Akko Harbor in Numismatic, Epigraphic, Cartographic and Photographic Evidence

lateen sails. East of the row of low rocks was an area without obstacles, apparently the harbor entrance (Fig. 17: C). From the southeastern shoreline of the city, seen on the right side of the view (Fig. 17: D) a rampart extends southwest (probably the abovementioned eastern rampart). The southwestern end of the rampart is a few meters above the water level. The protruding shape at the end of the rampart may represent the Island of Flies. Northwest of the end of the rampart (Fig. 17: E) a number of large vessels, more than 25 m long, with two masts and oars, are anchored. These vessels may have been anchored northwest of the eastern rampart in the protected area (the eastern basin northwest of the Island of Flies) where the water depth is 4 to 5 m. However, because the structure on the Island of Flies cannot be seen clearly, it is possible that the large ships seen in this plan are anchored outside the eastern basin, and are blocking the view of the Island of Flies. This coastal view clearly shows that at this time, the remains of the southern breakwater and the eastern rampart protruded at least half a meter above the water. No structure or tower (or remains of such) are seen at the eastern end of the southern breakwater. It also shows that medium-sized vessels could anchor in the western basin of the harbor in the seventeenth century. It seems that at this time the western basin was partly protected by the remains of the ancient southern breakwater. In later coastal views, for example, those of Bedford and Mansell from 1864 (Zviely et al. 2003: fig. 6), a number of relatively tall structures can be seen in the center and at the end of the southern breakwater, at a height similar to the structure on the Island of Flies. Remains of these structures also appear on nautical charts from the nineteenth and the twentieth centuries (Zviely et al. 2003), see below. Large structures seem to have been built on the southern breakwater at the end of the seventeenth century, which survived until the early twentieth century.

limited information about the harbor and the coast (Zviely et al. 2010). However, these could not serve as precise navigational aids in measuring distances and directions. The British Admiralty (founded in 1795) was the first to extensively publish nautical charts, called “Admiralty Charts” (Taylor 1971; Trostler 1982; Rosen 1983). From the nineteenth century to World War I, surveys were done manually. Soundings and bottom sediments were observed by a lead line (Galili et al. 2000; 2009; Rosen and Galili 2002; Oleson 2000). Location on the world grid was estimated by a chronometer and angles to the celestial bodies. Underwater obstacles were located by means of a towed wire drag (Grinker 2001). In 1913, the echo sounder was invented. A side-scan sonar, producing a continuous morphological map of the seabed, was developed in 1958. During 1940 and 1943 the nautical navigation system appeared, based on lowfrequency radio signals (LORAN – Long Range Navigation) and later, on satellites (GPS – Global Positioning System). A description of nautical charts of the area of Akko and its approaches is presented below. Additionally the archaeological significance of ancient marine remains and their function as revealed by analysis of nautical charts is discussed. The Development of Modern Maps of the Akko Harbor and its Approaches The first map of the land of Israel based on a modern system of surveying was produced by Jacotin in 1799, during the French occupation of Egypt and Palestine by Napoleon’s Army. Akko’s southern breakwater, on which the new breakwater now stands, is seen on this map as a line of reefs connected to the southeastern end of the Akko peninsula and extending eastward (Fig. 18). Just east of the end of this line of reefs a marine formation can be seen (the Island of Flies)

Nautical Charts Nautical charts are an important source of information in geographical-historical research of harbor cities and coastal areas. Old nautical charts feature objects that, in many cases, no longer exist. Today nautical charts provide essential hydrographic information useful for safety navigation. Nautical charts are classified according to scale, content or subject, for example, coastal charts, approach charts and harbor charts (Trostler 1982; Rosen 1983; May 1973). The late eighteenth century saw the start of evolution leading to the modern nautical chart, which constituted a conceptual revolution in maritime navigation (Zviely et al. 2009b). Until that time, in addition to portolanos, mariners had at their disposal harbor plans and illustrations, which provided only

N

Fig. 18: Enlargement of the Akko area, from the Jacotin Map (modified after Jacotin 1799)

23

Ehud Galili, Dov Zviely and Baruch Rosen

next to which the word phare (French for lighthouse) appears. Within the area north of the Island of Flies is an anchor symbol, indicating an anchorage. The anchor is located in the northern and central part of the eastern basin (between the eastern rampart north of the Island of Flies and the modern marina) and not in the western basin (within the Akko Marina). The first modern nautical navigational chart is the British Admiralty Chart 1242 (Acre) published in 1840 (Fig. 19). The map shows depths in fathoms, depth contours around the Akko headland, the coastline and a sketch of old Akko. A scale bar also appears in units of cables and nautical miles. The contours of the landscape, landmarks on the coast and known coastal bearings are drawn to assist in directing vessels sailing to the harbor. An unbuilt area can be seen along the western and southwestern

wall, a short pier can be seen protruding into the sea directed to the south (Fig. 19: A). Southwest of it are four dark, elongated spots near which the words “ancient mole” appear. The three western spots (Fig. 19, marked with the letter B) represent remnants of the southern breakwater and the easternmost spot (Fig. 19, marked with the letter C) depicts the Island of Flies. The area where the old harbor is marked is located in the eastern basin an continues hundreds of meters east of the boundaries of the actual location of that harbor. This deviation is an error of the cartographer who drew the map, and most probably was not familiar with the harbor area. In 1844, the British Admiralty published Chart 1585 (Bay of Acre), including the Haifa Bay, from the Carmel promontory to the area north of the Akko

N N Fig. 19: Enlargement of the Akko area, from Admiralty Chart 1242 (modified after Biddlecombe 1840)

Fig. 20: Enlargement of the Akko area, from Admiralty Chart 1585 (modified after Biddlecombe and Dillen 1844)

portion of the city wall, near the coastline (Zviely et

promontory (Biddlecombe and Dillon 1844) (Fig. 20). This map shows Na‘aman River and new navigational landmarks such as a bearing to the pier in map 1242 from 1840 (Fig. 19) or a wooden pier that was mentioned at an earlier version as a “marine structure jutting south into the sea” (Biddlecombe 1840). The depths of the harbor, from the eastern rampart and westward to the wall of Akko, range from 1 1/4 fathoms to 1 3/4 fathoms (0.5 to 3.2m). The western and northern parts of the harbor up to about 100m from the coastline are marked as up to 0.5m deep. The depth in the center of the harbor is about 2.75m, northwest of the Island of Flies it is about 3.2m, along the southern breakwater it is 2m to 2.75m and in the area of the eastern rampart the depth is 0.5–1.5m. South of the harbor the depth is over 6m, which is

al. 2003; 2009b).

The northern part of this area, at the foot of the wall, is a narrow abrasion platform and the southwestern part is a broad, mostly quarried, rocky surface. Historical sources, both written and graphic, show that in this area the fortress of the order of the Templars stood in the Crusader period. Archaeological surveys at this location have documented the rock-cut features, but no evidence of a Templar fortress was discovered (Galili this volume e). Along the southeastern coast of Akko the words “old harbor” appear. The submerged rampart extending northward from the Island of Flies (Linder and Raban 1965; Gertwagen 1996), known as the “eastern rampart,” does not appear on this map. Near the southeastern coastline, close to the

24

The Akko Harbor in Numismatic, Epigraphic, Cartographic and Photographic Evidence

W

E

Fig. 21: Panorama view of Akko from the sea to the east-northeast, from Admiralty Chart 1585 (modified after Biddlecombe and Dillen, 1844) suitable for the anchoring of large vessels. About 150m to 300m south of the southern end of the Akko peninsula, a reef is shown whose minimum depth is c. 4m (Fig. 20). On the edge of the chart is a panorama view of Akko, shown from a point at sea marked on the nautical chart with the letter A, looking east-northeast. On the right side of the panorama view (Fig. 21) four bodies can be seen protruding from the sea, with Tel Akko (“Tel Napoleon”) rising behind them in the background. The three left bodies (Fig. 21: 1–3) are remains of the southern breakwater, and the structure on the right (Fig. 21: 4) is the Island of Flies. The Island and the first structure to its left (Nos. 4, 3) are shown on the chart in different shapes and colors than the two structures which are closer to Akko’s city wall (Nos. 1, 2): The first two appear as lightcolored, symmetrical trapezoids, while the other two appear as dark, asymmetrical spots. It seems that the two right-hand (light) bodies are built marine installations, while the left-hand (dark) bodies depict reefs or ruins (see also Galili this volume e: figs. 7, 14). The space between the two marine structures (Nos. 3, 4) apparently marks the opening of the main entrance to the Akko Harbor. This opening was located between the Island of Flies and the eastern end of the southern breakwater. The panoramic view and the data from the chart show, at the eastern end of the southern breakwater, in 1840, a structure slightly lower than the Island of Flies. The presence of such a structure is supported by Roberts’ sketch dated to 1842 (Hartal 1997), which shows the Island of Flies a few meters above the water’s surface (Fig. 22: 1) and to its right, a slightly lower marine structure (Fig. 22: 2). Another sketch from 1855 (Fig. 23) shows similar patterns. Beginning with Napoleon’s war, the coast of Palestine saw extensive marine activity, which included partial surveys and nautical charting, the results of which were published mainly by the British Admiralty. In 1863 the British Admiralty published a new version of Chart 1262 (Fig. 24) which included Akko map (called Acre or Akka) on a scale of 1: 18.000 and a map of Haifa (also called Khaifa), on a scale of 1: 22,800 (Zviely et al. 2003; 2009b). Depths are shown on the chart in feet. In the area of Akko’s ancient harbor, from the eastern rampart to the Old City wall,

Fig. 22: Panorama view of Akko from Tel Napoleon in the east. Panoramic drawing of Roberts from 1842 (modified after Hartal 1997) depths were shown as ranging from 2 to 13 feet (0.6 to 4m). The depths in the western and northern parts of the harbor, up to a distance of about 100m from the coastline, are 1m, the depth in the center of the harbor is about 2m, the depth northwest of the Island of Flies is about 4m, the depth along the inner part of the southern breakwater is 1.2m, and the depth in the inner area along the eastern rampart is 0.6 to 1.5m. Depth contours and depth points were also added, and the number of points marking the bottom sediments were increased, the names of the large underwater kurkar ridges were noted in the area north of the Haifa Bay (Talbot and Vernon Reefs), named after Royal Navy ships that charted the area before 1860 (Zviely et al. 2009b). An anchorage was marked by an anchor symbol at a distance of about 300 m south of the entrance to the Akko Harbor, at a depth of 11m. In the area of the ancient harbor, important repairs and updates were marked: At the eastern end of the harbor entrance the Island of Flies was marked in the form of an inverted C, called for the first time by the Arabic name Manara, meaning a lighthouse/torch. The coordinates of the Manara are noted on the lower part of the chart. From the Island of Flies to the north, toward the coast, the eastern rampart can be seen, marked for the first time by a continuous, narrow arched line of marks (rocks) which almost completely blocks access to the harbor from the east. From the southeastern end of the wall of Akko toward the eastern end of the entrance to the harbor, the chart showed seven short, dark spots (the southern breakwater) alongside of which is the name “ancient mole,” which appears in its correct

25

Ehud Galili, Dov Zviely and Baruch Rosen

Fig. 23: Panorama view of the southern coast of Akko (after Bentley and Bartlett, 1855) geographical location. At the central spot of the mole a dark, hollow rectangle appears, apparently marking a structure on the southern breakwater. North of the eastern end of the southern breakwater, five small crosses appear for the first time; they mark an area of shallow water hazardous for sailing vessels. The

range of depths is marked in the ancient harbor on the 1863 chart, like the chart of 1844, and the words “Anc. Harbor” appear in the correct geographical location. The chart also shows a new panoramic view of the city of Akko from the sea (Fig. 25). In this view, the Island of Flies is seen, marked as “Manara” (Fig. 25: 4) and to the right a row of dark spots protruding slightly from the sea, apparently indicating the reefs of the eastern rampart. To the left of the Island of Flies is an area of sea (the entrance to the harbor) from which no objects protrude and to the left is a reef on top of which is a trapezoid-shaped object that is slightly lower than the Island of Flies (Fig. 25: 3). Farther to the left, two more, smaller, asymmetrical objects can be seen (Fig. 25: 1, 2). The first is located on a reef in the sea and the second is at the foot of the Akko wall. Southeast of the Island of Flies a large, three-mast ship is shown at anchor, facing west with its sails furled. Another nautical chart published by the British Admiralty in 1863 is Chart 1585 (Bay of Acre), which extends from the area of the Carmel coast in the south to the northern boundary of present-day Akko (Fig. 26). The chart is on a scale of 1: 45,400 and the depths are marked in fathoms (Zviely et al. 2009b). For the first time the angle of magnetic variation is shown. A coastal view from the sea of Akko and its approaches is also depicted, looking northeast at quite a distance from shore. In the Akko Harbor the minimal depth of the rocky reef in the older chart 1242 from 1840 is updated from 4 to 3 m, the Island of Flies is

N

Fig. 24: Enlargement of the Akko area, from Admiralty Chart 1262 (modified after Bedford and Mansell 1863)

26

The Akko Harbor in Numismatic, Epigraphic, Cartographic and Photographic Evidence

Akko sea walls

Harbor entrance

W

E

Fig. 25: Panorama view of Akko from the sea, looking north, from Admiralty Chart 1242 (modified after Bedford and Mansell 1863)

N

N

Fig. 26: Enlargement of the Akko area from Admiralty Chart 1585 (partial updating of details up to 1920) (modified after Hull et al.1863) Fig. 27: Enlargement of the Akko area from Admiralty Chart 1565 (partial updating of details up to 1920) (modified after Bedford and Mansell 1862)

marked as a hollow rectangle, the structure at the eastern end of the southern breakwater is marked as a hollow square and the structure in the center of the southern breakwater is marked as a hollow rectangle, partially open in its northern portion. In this chart, only five reefs appear instead of seven, possibly due to cartographic considerations. The charts and the coastal views drawn by Mansell and his subordinates were copied and corrected by various national and private agencies. Details that were added are: the Akko lighthouse and details about its illumination and a fishermen’s pier in the inner, western part of the Akko Harbor, as it is seen in the charts (Hull et al. 1863) (Fig. 26) and the chart by Bedford and Mansell (1863) (Fig. 27). From 1930 to 1932, the survey ship H.M.S. Endeavor operated along the shores of Palestine and in 1931, the British Admiralty published a new version of maritime Chart 1585 (Rosen 1993, Zviely et al. 2009b). The hydrographic data presented on this chart were based on an extensive hydrographic survey carried out in 1930 (Edgell et al. 1930, 1931). The map features three areas: Akko (Acre or Akka) on a scale of 1: 50,000; “Haifa Harbour,” on a scale

of 1:12,500 and the “Akka Anchorage.” A comparison of the depths measured in fathoms on Chart 1585 (Fig. 26), to depths measured in feet on chart 1585 and a fair sheet (Figs 27, 28), shows that the depths on Chart 1585 are dozens of centimeters shallower. The difference stems from an approximate correction to the mean sea-level, carried out according to depth data obtained in a hydrographic survey. The shallow water in the Akko Harbor in 1930 is apparently the result of siltation due to poor maintenance. The poor maintenance may indicate a decline in its regional importance following the construction of the Haifa Harbor between 1929 and 1932. Updates entered on the chart from 1931 in the area of the Akko Harbor include notation of the height of the structure on the Island of Flies as 22 feet (6.6m). A change in the marking and location of the eastern rampart on this chart shows that its high points are submerged and that it is connected to the shoreline and not to the Island of Flies. The marking of the southern breakwater was changed from the row 27

Ehud Galili, Dov Zviely and Baruch Rosen

of reefs on the previous chart to a contiguous marine structure protruding from the water (Fig. 29). The latest version of Chart 1585, which, for the first time, shows depths in meters (Clark 1998) (Fig. 30), was published in 1998 following collaboration between the British Hydrography and the Israeli Navy (Zviely 2006; Zviely et al. 2009b).

Aerial Photographs of the Akko Harbor and its Approaches Charts and illustrations present selective details that are important for their purposes, and therefore, much information and many details are missing. Among them are archaeological and historical details, some of which are important for understanding the function of the harbor and the city. Aerial photographs, on the other hand, show the situation as it is, and thus are a source of information for historical research on ports, anchorages and coastal areas. The first aerial photographs of the Akko Harbor were taken from 1917 to 1918 during World War I by the Bavarian Squadron No. 304, which was active in Palestine (Kedar 1992; Dichter 1993). These show the eastern rampart under water and the small rampart protruding from the eastern end of the southern breakwater northward. They also depict an isolated structure (the remains of a lighthouse) on the flooded platform north of the recent lighthouse (Fig. 36). From that time and until the establishment of the State of Israel, the area was photographed by the Royal Air Force. After the establishment of the State of Israel the area was frequently photographed by governmental and private entities (Fig. 31) (Zviely 2006; Zviely et al. 2009a). In 1966, a 230m-long breakwater was built, which covered most of the remains of the ancient southern

Fig. 28: Enlargement of the Akko area from the hydrographic survey (fair sheet) H7647 (after Edgell et al. 1930)

N

Fig. 29: Enlargement of the Akko area from the 1585 Admiralty Chart (modified after Edgell et al. 1931)

N

Fig. 31: Aerial photograph of the Akko Harbor, August 26, 1961, photo 4781, flight mm/60 (modified after Survey of Israel)

Fig. 30: Enlargement of the Akko area from Admiralty Chart 1585 (after Clarke 1998) 28

The Akko Harbor in Numismatic, Epigraphic, Cartographic and Photographic Evidence

this structure (The so-called northern lighthouse) as an active lighthouse with an added top story (Figs. 34a, 34b). The lighthouse stands on the flat, flooded rock platform, with a small foot-bridge connecting it to the coast. Fig. 33 show no lighthouse structure on the high, round platform on the sea wall (the present location of the lighthouse, Fig. 33b) at that time. Admiralty Chart 1242 from 1863 (Mansell et al. 1863) (Fig. 35) depicts the lighthouse standing on the center of the flat platform northwest of the location of the current lighthouse (marked with a black arrow). According to the chart its elevation is 46 ft. Aerial photos taken by the Bavarian Squadron during 1917-1918 depict this northern lighthouse in ruins and a new lighthouse built on the city wall ca. 150m to the south (Fig. 36, marked with a white arrow). In later corrections to Admiralty chart 1242 from 1922 (Bedford and Mansell 1863: Small corr. 1920) the elevation of the lighthouse was changed to 51 feet. That change may reflect the relocation of the lighthouse, as depicted in the photos. It is suggested that the first lighthouse, seen in Figs. 33, 34, was built on a low, flat area during the second half of the eighteenth century, probably in 1864. Most probably it is the one listed among the lighthouses maintained by the French company by 1864 (Zviely et al. 2013: 15 and Table 1). That northern lighthouse may have been damaged by sea storms or an earthquake and stopped functioning. Around 1915–1920, a new lighthouse, which functioned until recently, was built on the elevated platform on the southwest corner of the sea wall. Traces of the old, abandoned and destroyed, lighthouse (Figs. 37: a, 38: a) as well as the current lighthouse (Figs 37: b, 38: b) are seen in photos from 1927 and 1934. Scanty remains of the northern lighthouse are seen today on the flooded rock platform (Fig. 39 marked with a white arrow).

breakwater. In the 1970s, three docks were built along the inner part of the breakwater and in the 1980s, another dock was expanded (Fig. 32).

N

Fig. 32: Aerial photograph of the Akko Harbor, November 4, 1995, photo 3247, flight ml/501 (modified after Survey of Israel)

Late Eighteenth- to Early TwentiethCentury Photographs Photograph taken by the Italian photographer L. Fiorillo from Alexandria (who was active in the region during 1870-1889) depicts Akko`s southwest corner and the sea walls looking from the east. A onestory building with tiled roof, like a classic Frenchstyle lighthouse is depicted on the coastline (Fig. 33a) northwest of the location of the current lighthouse (Fig. 33b). An additional photo before 1914 depicts

a

b

Fig. 33. 1870-1889 looking west: (a) the northern lighthouse of Akko. (b) the round platform on the sea walls (Photo: L. Fiorillo, courtesy of the State Museum of the History of Religion, St. Petersburg, P 3421) 29

Ehud Galili, Dov Zviely and Baruch Rosen

Fig. 34a. Taken before 1914, looking west: The functioning northern lighthouse (courtesy of Jean Michell the Tarragon, The Saint Etienne Bibliotheque, Ecole Biblique, Jerusalem, and the Assomptionniste French Community of former Notre-Dame de France, Jerusalem).

Fig. 34b. Taken before 1914, looking southwest: The northern lighthouse two-story building, on the right, a two-mast vessel sailing north (photo: courtesy of the Saint Etienne Bibliotheque, Ecole Biblique, Jerusalem. (after Abel 1914, fig. 7).

Fig. 36. Akko 1918 looking north: a) the damaged northern lighthouse on the left (marked with a white arrow), b) the new lighthouse under construction or already functioning (marked with a white arrow) (photo: the Bavarian Squadron, modified after Kedar 1991: 220, fig. m6)

Fig. 37. 1927 Views at Akka from the barracks, looking south: (a) The ruined northern lighthouse, (b) the new southern lighthouse (Courtesy of the Copyright holder: the Baha’i World Centre, after Birgham, Star of the West 18: 9, December 1927, p. 281)

Fig. 35. The location of the first Akko Lighthouse depicted in admiralty chart 1242 from 1864 (marked with an arrow) (after Zviely et al. 2013 fig. 8) 30

The Akko Harbor in Numismatic, Epigraphic, Cartographic and Photographic Evidence

Fig. 38. Photo taken before 1934 by Cliches Barrois, looking south. Akko northern lighthouse in ruins and the new elevated southern lighthouse, with the Carmel range in the background (courtesy of Jean Michell the Tarragon, The Saint Etienne Bibliotheque, Ecole Biblique, Jerusalem, negative No. 6219), and see also Abel F. M. 1934 pl. XLIII, p71. the harbor’s history. Extensive development is now being planned for the Akko seafront, marina and its environs. Before the archaeological remains are destroyed and the outlines of the ancient harbor and its approaches totally change, it is crucial to exploit the remaining archaeological evidence and compare it to the iconographic and historical sources. Recently, in the framework of Israel Antiquities Authority excavations, numerous finds were retrieved from the Akko Marina. The finds are presented in this volume, and together with the iconographic depictions, they contribute to our understanding of the history and development of the Akko Harbor: 1. Cross-referencing current hydrographic information with illustrations on coins from the Roman period and descriptive written information gives rise to a general understanding of the Roman harbor of Akko. It seems that this was a built harbor that included docks, storehouses and/or shipyards, breakwaters and most probably two lighthouses. 2. Written and illustrated documents resulting from the intensive activity in Akko and its harbor during the Crusader period have added much information on the seafront of the city at that period. Later illustrations and anchorage plans, coastal views (panoramas), charts and aerial photographs, combined with marine archaeological findings, provide a great deal of information about the harbor’s evolution to this day. The finds indicate that the harbor was built during the Late Hellenistic-Early Roman period (Galili this volume g). 3. The area nicknamed “the Pisan Harbor” is shallow and exposed to winds and waves from the southwest.

Fig. 39. 2007 looking north: the flat rock-cut platform and the remains of the northern lighthouse (marked with an arrow) (E. Galili)

Summary and Conclusions Geomorphological and sedimentological processes in the Akko area change the coast naturally, leading to erosion and destruction on the one hand, and accretion and accumulation on the other. Human intervention in the coastal environment through maritime constructions, sand and stone quarrying, land reclamation and secondary use of building stones also destroy ancient structures and facilities in and around the harbor. Despite all these building and destruction processes, the archaeological remains left behind may be used as evidence in understanding

31

Ehud Galili, Dov Zviely and Baruch Rosen

anchorage and landing place. 12. In nineteenth-century charts, a wooden pier is depicted on the northeastern shore of Akko (east of the Old City). This pier no longer appears in maritime charts from 1863. 13. Late 18th/early 20th century photographs and maps suggest that the location of the Akko lighthouse was changed around 1913.

The approach to the coast from the sea is blocked by the remains of the ancient breakwater and the southern sea wall. It seems that there was no harbor, dock or anchorage in this area after the Roman period. 4. Based on some of the Sanudo illustrations, the structure on the Island of Flies must have been built before the fourteenth century. 5. The “tower” feature depicted in some of the Sanudo drawings on the eastern end of the southern breakwater probably did not exist. It is reasonable to assume that the appearance of this tower-like formation is a mistake stemming from the copying of earlier drawings by illustrators who had not actually visited the Akko Harbor. 6. The broken lines connecting the “tower” to the Akko coast in the 1323 illustration (Fig. 10: C), apparently mark a row of reefs or remnants of a breakwater rather than a chain. The existence of such reefs was confirmed in a description of “rocks” near the entrance to the Akko Harbor in a nautical guide from this period. 7. The eastern rampart that was built during or before the Early Muslim period apparently served as a physical obstacle that prevented enemy ships from entering the harbor from the east. It may also have acted as a causeway or a pier leading to the Island of Flies. This rampart apparently did not serve as a breakwater because waves from the east cannot reach a significant height over such a short fetch (the distance from the shore that enables the wind to raise the waves). 8. After the Mamluk conquest that marks the end of the Crusader period in the south Levant coast, a wooden pier was built at the end of the southern breakwater. This pier allowed relatively large vessels to dock, load and unload cargoes and passengers near the southern waterfront of Akko. 9. During the mid-fourteenth century the southern breakwater was in ruins, leaving behind only rocky reefs that protruded slightly from the water. At that time medium-sized vessels could anchor within the western basin (currently Akko’s marina), while large ships anchored near the Island of Flies (south and southwest of the island), where the water was deeper. 10. In the coastal view of Akko from 1686, a very low silhouette of the southern breakwater is seen, with no buildings on it. Beginning in 1840, nautical charts, photos (Galili and Rosen this volume f: fig. 8) and plans of Akko show structures on the southern breakwater. These structures, which take various forms, appear until 1920 in the center and at the eastern end of the southern breakwater. Beginning in the 1930s' these structures are no longer seen. Various structures that were on the southern breakwater during the end of the seventeenth century and the early eighteenth century survived until the early twentieth century. 11. On charts of Akko and its approaches from the late eighteenth to early nineteenth centuries, the eastern basin of the Akko Harbor is shown as an

Acknowledgments The authors wish to thank the Israel Antiquities Authority for the institutional support, M. Feinberg Vamosh for the English editing, D. Jacoby and A. Fichman for their remarks on the manuscript, M. Lyubov Musiyenko, director, the State museum of the history of religion, St. Petersburg and M. Anna Kudrik for providing Fig. 33, Jean-Michel de Tarragon, archivist of Ecole Biblique, Jerusalem, for providing Figs. 34, 38, the Survey of Israel, A. Ohayon and Google Earth for the aerial photos and S. Sachyan for his help in locating Figs. 34, 37.

References Abel F. M. 1914. Le littoral palestinien et ses ports. Revue Biblique 4: 556-90. Abel F. M. 1934. Le couvent des Frères Prêcheurs a Saint-Jean d’Acre. Revue Biblique 43: 265-284. Ashtor E. Levant Trade in the Later Middle Ages. Princeton. Avi-Yonah M. 1954. The Madaba Mosaic Jerusalem.

Map.

Bedford F.D.G. and Mansell A.L. 1963. Acre or Akka – Chart 1242. 1:18,200 Scale. London. Bentley J.C. and Bartlett W.H. 1855. Acre. Holy Land. Original steel engraving drawn by W.H. Bartlett engraved by J.C. Bentley. Biddlecombe G. 1840. Acre – Chart 1242. London. Biddlecombe G. and Dillon C.H. 1844. Bay of Acre

– Chart 1585. London.

Brigham G.R. 1927. View at Akka. Star of the west 18/9. p. 281. Carmel A. 1969. The History of Haifa in the Turkish Times. Jerusalem, Figs. 3, 10. (Hebrew). Clarke J.P. 1998. Hefa (Haifa) – Chart 1585. 1:20,400 Scale. Taunton, U.K.

32

The Akko Harbor in Numismatic, Epigraphic, Cartographic and Photographic Evidence

Dan Y. 1972. Information about Akko from a Greek Composition from the Seventh Century. In B. Oded, B. Rappaport, A. Shohat and Y. Shmitzler (eds.).

Galili E. This volume e. Structures and Installations in and around the Akko Marina.

Studies in the History of the Jewish People and the Land of Israel. Pp. 53–62. Haifa. (Hebrew).

Galili E. and Rosen B. This volume e. The Akko Marina Archaeological Project – Summary.

Dichter B. 1973. The Maps of Acre – An Historical Cartography, Acre.

Gertwagen R. 1996. The Crusader Port of Acre: Layout and Problems of Maintenance. In M. Balard (ed.). Autour de la Premiere Croisade, Actes du Colloque de la Society for the Study of the Crusaders and the Latin East, Clermont-Ferrand, 22–25 juin 1995. Pp. 553–582. Paris.

Edgell J.A. and the Officers of H.M Surveying Ship Endeavour 1930. Acre Anchorage – Chart H7647 (fair sheet). 1:25,000 Scale. London.

Grinker B. 2001. Nautical Charting – Principles of the Hydrographic Survey. Etmoded 18: 9–11. Civil Maritime Engineering, Ltd. 1962. Akko Harbor – Proposal for a sport-boat harbor. Map 2-218T, 1: 1,000 Scale, Haifa. (Hebrew).

Edgell J.A. and the Officers of H.M Surveying Ship Endeavour 1931. Acre Anchorage – Chart 1585. 1: 25,000 Scale. London. Flinder A., Linder E. and Hall E.T. 1993. Survey of the Ancient Harbour of Akko, 1964, 1966. In M. Heltzer, A. Segal and D. Kaufman (eds.). Studies in the Archaeology and History of Ancient Israel. Pp. 201–225. Haifa.

Hartal M. 1997. Excavation of the Courthouse Site at Akko: Summary and Historical Discussion. ‘Atiqot 31: 113. Hill G.F. 1914. Catalogue of Greek Coins. London. Hull T.A., Christian F.B. and Mansell A.L. 1863. Bay of Acre – Chart 1585. 1:45,400 Scale, London.

Fried, Glaswein, Werner Engineers 1957. Municipality of Akko. The Harbor Plan. Map 191–4, Scale 1:1,000. Haifa. (Hebrew).

Jacoby D. 1979. Crusader Acre in the thirteenth century: Urban layout and topography. Studi Mediveali 3a Serie, XX: 1, 197: 1–45.

Galili E., Sharvit J., Dahari U., Bahat-Zilberstein N., Finkielsztejn G., Stern E.J., Kool R., Kahanov Y. and Rosen B. 2002. Akko Port, Underwater Surveys. Hadashot Arkheologiyot 114: 13–16, 12*–15*. (Hebrew/English*).

Jacotin P. 1799. Acre, Nazareth, Le Jourdain – chart Flle.46, 1:100,000 Scale. In Carte topographique de L`Egypte et de plusieurs parties des pays limitrophes levee pendant l`expedition de l`Armee francaise. Paris, 1810.

Galili E., Zviely D. and Rosen B. 2004. Akko Port and Surroundings on Coins and Graphic Representations. Horizons in Geography 62: 113–130. (Hebrew, English abstract).

Kadmon N. 1982. The History of Maps of Eretz Israel. In N. Kadmon and A. Shmueli, Maps and Mapping. Jerusalem. (Hebrew).

Galili E. and Rosen B. 2008. The Harbor (Acco). In E. Stern, H. Geva and A. Paris (eds.). The New

Kedar B. Z. 1991. Looking twice at the Land of Israel, Aerial Photographs of 1917-1918 and 1987-91.

Encyclopedia of Archaeological Excavations in the Holy Land 5. Pp. 1558–1561. Jerusalem.

Kedar B.Z. 1997. The Outer Walls of Frankish Acre.

Galili E., Rosen, B. and Zviely, D. 2009. Sounding Weights Recovered off the Mediterranean Coast of Israel. International Journal of Nautical Archaeology 38 (2): 343–368.

‘Atiqot 31: 157–180. Jerusalem.

Linder E. and Raban A. 1964. Underwater Survey of the Akko Harbor. In Western Galilee and the Galilee

Coast, 19th Israel Exploration Society Conference, 1964. Pp. 180–194. Jerusalem. (Hebrew).

Galili E., Rosen B. and Sharvit J. 2010. Artifact Assemblage Recovered From a Roman Shipwreck Off the Carmel Coast, Israel. ‘Atiqot 63: 61–110.

May W.E. 1973. A History of Marine Navigation. Oxfordshire, U.K.

Galili E. This volume a. The Akko Marina Archaeological Project – Introduction.

Meshorer Y. 1984. City Coins of Israel and TransJordan in the Roman Period. Jerusalem. (Hebrew).

Galili E., Rosen B., Zviely D., Silberstein N., Finkielsztejn J. 2010. The Evolution of the Akko Harbor and its Mediterranean Maritime Trade Links. Journal of Island and Coastal Archaeology 5: 191–211.

Motzo B.R. 1947. II Compasso da Navigare. Annali della Facoltà di Lettere e Filosofìa della Università di Cagliari Vili.

33

Ehud Galili, Dov Zviely and Baruch Rosen

Zviely D., Kit E., Rosen B., Galili E. and Klein M. 2009a. Shoreline migrations and beach-nearshore sand balance over the last 200 years in Haifa Bay (SE Mediterranean). Geo-Marine Letters 29 (2): 93–110.

Oleson J.P. 2000. Ancient Sounding-Weights: A Contribution to the History of Mediterranean Navigation. Journal of Roman Archaeology 13: 294–310. Prawer J. 1991. The Crusader Contribution to the Cartography of Eretz Israel. Cathedra 57: 20–30. (Hebrew).

Zviely D., Rosen B. and Galili E. 2009b. The Hydrographic surveys of Haifa Bay by the British Admiralty 1832–1931. Horizons in Geography 73: 142–166. (Hebrew, English abstract).

Raban A. 1983. The Akko Harbor, Kardom 24–25. Pp. 141–146. (Hebrew).

Zviely D., Rosen B. and Galili E. 2010. Anchoring plans along the Israeli Coast from the 16th Century to the 19th Century CE. Horizons in Geography 76: 92–111. (Hebrew, English abstract).

Raban A. 1986. The Ports of Akko. In M. Yed‘aya (ed.). The Western Galilee Antiquities. Tel Aviv. (Hebrew).

Zviely D., Rosen B. and Galili E. 2013. The history of lighthouses along the Israeli coast. Horizons in Geography 84: 5–23. (Hebrew).

Raban A. 1993. Maritime Caesarea. In In E. Stern (ed.). The New Encyclopedia of Archaeological Excavations in the Holy Land 2. Pp. 286–291. Jerusalem. Rosen B. 1993. Mappings of the Coast of Palestine by the Royal Navy. Cathedra 64: 59–78. (Hebrew). Rosen B., Galili E. and Zviely D. 2011. The Roman lighthouse in Akko, Israel. International Journal of Nautical Archaeology 41(1): 171–178. Shaw J.W. 1972. Greek and Roman Harbour Works. In G.F. Bass. A History of Seafaring. Pp. 101–106. London. Shor N. 1990. The History of Akko. Tel Aviv. (Hebrew). Sivan D. 1981. Paleogeography of the Akko Area in the Holocene Period (M.A. thesis, University of Haifa). P. 103. (Hebrew). Haifa. Stern E. 1999. The pottery of the thirteenth–fifteenth centuries from Giv‘at Yassaf (Tell er-Ras). ‘Atiqot 37: 125–136. (Hebrew). Taylor E.G.R. 1971. The Haven-Finding Art. Pp. 106–109. London. Trostler Y. 1982. Nautical Charts. In N. Kadmon and A. Shmueli. Cartography and Maps. Jerusalem and Tel Aviv. Pp. 80–93. (Hebrew). Zviely D., Galili E. and Rosen B. 2003. The Port of Acre and its Approaches in Modern Nautical Charts.

Horizons in Geography 56: 62–77. (Hebrew, English

abstract).

Zviely D. 2006. Sedimentological processes in Haifa Bay in context of the Nile Littoral Cell. (Ph.D. diss., University of Haifa. Haifa. (Hebrew, English summary).

34

Chapter 3

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina Na‘ama Silberstein (1), Ehud Galili (2) and Jacob Sharvit (1) 1. Israel Antiquities Authority 2. Israel Antiquities Authority and Zinman Institute of Archaeology, University of Haifa Key words: ancient harbors, marine archaeology, amphora, pottery, Akko, Acre

Introduction From November 1992 to June 1993, the Akko Marina was deepened so that it could serve ships with a draft of several meters (Galili et al. 2002, 2010). The marina was deepened using a barge-based dredger, which raised mud and sand from the harbor bed and transferred it to another barge with a bottom that could be opened (Galili this volume a). About 15,000 cubic m of material were dredged from the harbor. The project was accompanied by IAA archaeological work. The dredged material was dumped in selected loci in the open sea. The deposit sites were surveyed by divers using a metal detector. Findings that could be classified were collected in plastic baskets. Survey dives were also carried out in the marina where ancient remains were located and mapped. To map the origin of the archaeological finds, the marina was divided into six main work areas, each of which included a number of 10 ×10 squares (Fig. 1): Area A – east of the marina entrance; Area B – north of the marina entrance; Area C – entrance to the marina; and Areas D, E and F – inside the marina. This article will chronologically describe the pottery discovered from the Hellenistic, Early Roman (third century BCE–second century CE), Late Roman and the Byzantine periods (second–seventh centuries CE). The study of the pottery, which included identifying the origin of the various vessels, helps to understand the functioning and location of the port throughout its history and enables an understanding of trade and economic relations between Akko and its port and other coastal cities and production centers around the Mediterranean basin.

Fig. 1: Akko Marina and the areas of excavations (E. Galili) and engraved vessels, the vessels are presented by type rather than chronologically, noting the period of each type based on parallels. The vessels shown in the plates below represent almost all forms and types identified in the assemblage.1 All the fragments identified were classified into types and for each type, a typical selection of sherds was drawn and/or photographed. In some groups a minimum number of vessels of each type was counted.2 The distribution of the fragments of each type throughout the marina was also examined. The purpose of recording the location of the finds in the marina was to clarify which parts of the marina were in use in which periods. The vessels are described based on their forms and families. In each form or family, the division into types was done, where possible, according to the usual division, particularly the high-quality, widely

Methodology In each period the vessels have been categorized in the accepted manner, by type, from open to closed. In a number of categories, such as lids, oil lamps

1 Some types may not have been identified due to lack of indicative sherds.

2 The count was made possible because of the relatively small number of items of each type.

35

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

distributed vessels that have already been thoroughly studied. Vessels that have not yet been extensively researched were divided typologically, based on parallels at other sites. Usually, it was preferred to cite references to sites in which the types are well known and to publications where discussion is broad and based on numerous parallels. It should be noted that we did not present all published parallels. If a type was identified that had not been discussed extensively in a previous publication, as many parallels as possible were presented so as to show the geographical and chronological attribution as securely as possible. Some vessels, usually with no parallels, are depicted in the plates with no citation in the text. The time span of the vessels was determined to within half a century accuracy due to the constraints of the salvage operation and in the absence of clear stratigraphy; greater precision was not considered important in this repertoire, which extends over many centuries. The determination of geographical origin was made based on other publications and on typology. The fabric of the assemblage was not examined petrographically; in some cases the characteristics of the material from the marina was compared to published descriptions of materials. In other cases the source was identified by published distribution models. In describing the characteristics of the assemblage, a minimum number of vessels was counted for each of the types in the groups of bowls, cooking vessels and amphorae. The bowls and the cooking vessels were counted based on an examination of the diameters, base heights, rim lengths and other parameters of each type that assist in distinguishing between fragments of different vessels. A number of fragments whose parameters were similar or possibly similar were counted as one vessel. In counting the bowls, in cases where numerous bases or rims could not be associated with each other, the rims and the bases were counted separately and the larger number was determined to be the minimal count. In cooking vessels with handles the calculation was made based on the number of handles of each type found, which was divided by the number of handles usually found on a single vessel of the type under consideration. With regard to the amphorae and the jugs, the fragments of the necks and bases were counted. The purpose of counting the vessels was to determine the quantitative relationship between imported and local vessels in various periods and for various vessels. The families that were examined were bowls, jugs and juglets, representing table ware, cooking vessels of various types and amphorae. As for the rest of the groups, quantitative examinations were not undertaken because only one vessel of each type was present for each period, and therefore they do not impact the picture of the relatively common vessels in the assemblage. The typological division of the amphorae was

based mainly on the works of Dressel (1899) and the research of Peacock and Williams (1986) and Sciallano and Sibella (1994). A detailed description of the amphorae is presented in tables based on numbers of vessels, without a division into types. Similarly, amphorae bases are presented in tables without division into types. The bases are described as one assemblage without classifying them chronologically because the lack of whole vessels precluded connecting bases to upper parts of amphorae. The ceramic findings described below do not include amphorae handles with seal impressions, nor do they include ceramics from the Fatimid, Crusader and the Ottoman periods. These have been described separately (Finkielsztejn this volume and Galili et al. this volume b). The clay color of the amphorae is described based on the Munsell Soil Color Chart.

The Hellenistic Period (Third–Second Centuries BCE) Open Vessels Four types of open Hellenistic vessels were identified: Type 1 – fish bowls (Pl. 1: 1–6): Diam. 35–55cm, 20cm, oblique wall, downturned rim, ring base, usually with depression; groove in some rim fragments at beginning of downturn. Two types of glaze completely coat both inner and outer walls: 1. Black, lustrous glaze; 2. Matte glaze in various shades ranging from brown to red. Imported. Parallels: Dor, third–second centuries BCE; Tel Anafa, Athens (GuzZilberstein 1995: 291–293, Types BL4b, BL5b). Type 2 – bowl with outfolded rim (Pl. 1: 7, 8): Diam. 30–50cm, oblique wall, flattened, outward folded rim, ring base; Two types of glaze coat the interior wall and the upper part of the exterior wall: 1. Red in the center and the bottom of the wall and the rest black; 2. Brownish-red covering the entire vessel. Most bowls of this type found in the assemblage are made of a light fabric resembling that in bowls of Type 6. One is of a different fabric and contains mica. Imported vessels. Parallels: Dor, beginning in the third century BCE but mainly second century BCE (Guz-Zilberstein B. 1995: 292–293, Type BL5a). Type 3 – bowl with everted wall (Pl. 1: 9–14): Diam. 15–40cm, curved or carinated wall, everted rim, ring base. Two groups were discerned, differing in source, size of base, slip and decoration, as well as the appearance of the clay: 1. Imported vessels, broad ring base, black, oily, lustrous slip covering interior and exterior walls, stamped decoration on interior, levigated orangepink clay. Parallels: Dor, mid-fourth and mainly midthird centuries BCE (Guz-Zilberstein 1995: 291). 2. Local vessel, narrow ring base, reddish, matte slip on interior and exterior walls, except for base, light-

36

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

colored clay. Parallels: Dor, second century BCE (Guz-Zilberstein 1995: 291, Type BL7c). Type 4 – bowl with everted rim (Pls. 2: 1–8, 14; Pl. 3: 1, 2): Diam. 18–25cm, curved wall, inverted rim and ring base, slipped in shades of black, brown and red of varying quality, from thick and lustrous to thin. This type also has a group of vessels made of lightcolored clay similar in quality to the clay in Type 6 imported vessels. Parallels: Athens, beginning in the mid-fourth century (Thompson 1934: 435–437), Samaria, third and second centuries BCE (Crowfoot and Kenyon 1957: 229).

vessels. Parallels: Athens, second century BCE (Thompson 1934: fig. 29: C9). Type 4 – piriform, ribbed juglet (Pl. 27: 3–6; Figs. 2, 3): Everted rim, handle drawn from rim to shoulder, button base (Pl. 27: 4) or flat base (Pl. 27: 5, 6). Imported. Parallels: Tel Anafa, second–first centuries BCE (Berlin 1988: 170, 173–175); Dor fourth century BCE (Guz-Zilberstein 1995: 307, fig. 6.28: 8); Tel Michal, third century BCE (Fischer 1989: 184, fig. 13.3: 21).

Amphorae Cooking Bowls

The earliest amphorae recovered from the Akko Marina date from the Persian period (Pls. 35: 3; 47: 5, 6). Because of the small quantity of vessels from the Persian period, they are described together with the vessels from the Hellenistic period. Also due to their small number (ca. three fragments) they are not included in the quantitative analysis of the remains. Eighty-six body fragments of amphorae belonging to

Type 1 – (Pl. 19: 1): broad, oblique rim with ridge, handles attached to vessel wall, curved base. Imported. Parallels: Dor, mid-fourth–second centuries BCE (Guz-Zilberstein 1995: 299), Sidi Harabish (Benghazi), later than the mid-third century BCE (Riley 1979a: 242). Type 2 – (Pl. 19: 2): broad, flared rim, oblique wall, flat base. Local? Parallels: Dor and other sites in Israel not including Judea (Guz-Zilberstein 1995: 300, Type c, p. 8). Frying Pans Type 1 – (Pl. 23: 3): Gutter rim with support for lid, oblique wall, curved base. Import. Parallels: Sidi Harabish (Riley 1979a: 242–243), Tel Anafa (Berlin 1988: 84, Pl. 14: 13), Athens (Thompson 1934 E145; Rotroff 1982). Type 2 – (Pl. 23: 4): Fluted rim, thickened wall, perforated lug handle. Imported. Parallel: Paphos (Hayes 1991: 82, fig. 32: 10).

Fig. 2: Juglets with button base (IAA) Cooking Pots Type 1 – (Pl. 24: 1): Straight neck, curved wall, loop handles drawn from rim to shoulder. Local. Parallels: Dor (Guz-Zilberstein 1995: 299). Closed Vessels Type 1 – strainer jug (Pl. 25: 2): Gutter rim, strainer, black slip. Imports. Type 2 – jug with broad neck (Pl. 25: 10): Flaring rim, square, broad neck. Imports. Parallels: Ashdod, (Dothan 1967: 25 fig. 7:2; Vessberg and Westholm 1958: 60, fig. 23: 18, 19). Type 3 – black-slipped juglet (Pl. 27: 1, 2): Imported

Fig. 3: Juglets with flat base (IAA)

37

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Type 2 – bowl with broad ring base (Pl. 4: 2–5, Fig. 4): Diam. 16–27cm, upright, curved wall, tapered rim, broad ring base, thick red slip. Imported. Parallels: Hayes Type 4a, second century and mainly first century BCE (Hayes 1985: 15–16, tav. 1: 9–10).

30 different types were identified from the Hellenistic period. In addition, 44 amphorae handles were found (see Finkielsztejn this volume). The amphorae types and their parallels are presented in Pls. 35–49 and in Table 4. Most of the amphorae from the Hellenistic period (82%) were imported from the region of the Aegean Sea. The rest are local (Palestine and Syria) or imported from Egypt, North Africa and Italy.

The Late Hellenistic Period (Mid-Second Century–Late First Century BCE) Open Vessels Type 1 – bowls with epsilon handles (Pl. 2: 12, 13): Diam. 24–28cm, S-shaped wall, thin, outfolded gutter rim, narrow ring base, red slip on interior wall and on a large part of the exterior. A few examples bear black slip on the walls. Parallels: Dor, second century BCE (Guz-Zilberstein 1995: 294, BL10b). Local vessels or imported from the Eastern Aegean. Type 2 – bowl with wishbone handles (Pl. 2: 9–11, 15): Diam. 28cm, 17cm, carinated, upright wall, plain rim, ring base, gray-black or red slip that does not cover the entire vessel. Imported. Parallels: Akko, Judea, mid-first century BCE and throughout the first century CE (Bar-Nathan 1988: 196); Paphos, first century BCE–first century CE (Hayes 1991: 134, fig. 47: 39). Type 3 – bowl with ledge rim (Pl. 3: 3): Diam. 28cm (estimated), curved wall, everted ledge rim, base unknown, without slip; back and red bands decorate the rim and the interior wall. Imported. Parallels: No parallels found for form; for decoration: Dor, 125–60 BCE (Guz-Zilberstein 1995: fig. 6.41: 18).

Fig. 4: Bowl (IAA)

Type 3 – bowl (Pl. 5: 1, 2): Diam. 15cm, curved wall, tapered rim decorated in a mold, curved base, red/ brown slip. Imported. Parallels: Dor, second century BCE (Guz-Zilberstein 1996: 318–319), Hayes Type 17a, mid second–mid first centuries BCE (Hayes 1985: 21–22, tav. 3: 4). Type 4 – small bowl with ring base (Pl. 5: 4–6, Fig. 5): Diam. 15–18cm, curved wall, stylized ring base, thick, red, peeling slip. Imported. Parallels: Hayes Type 22a, late second–mid-first centuries BCE (Hayes 1985: 23–24, tav. 3:10). Type 5 – jug (Pl. 7: 8): Folded, everted rim, long, cylindrical neck, globular body, strap handle ridged on the edges from the neck to shoulder, red slip with rouletted decoration on the walls. Parallels: Hayes Type 101, first century BCE (Hayes 1985: 42, tav. 9: 2). Type 6 – jar with trefoil rim (Pl. 7: 9): S-shaped, trefoil rim, concave neck, globular body, strap handle with three ridges from beneath the rim to the shoulder, red slip. Parallels: Hayes Type 105, apparently beginning in the first century BCE (Hayes 1985: 42, tav. 9: 6).

Eastern Terra Sigillata A (ETSA) This family is characterized by light yellowish clay and a red or reddish-orange slip. In early vessels the slip is thick and lustrous and in many cases a dark band of color can be discerned in the area where the two dippings overlap. Most research indicates that Northern Syria is the source of these vessels, mainly based on their extensive distribution in that region. A relatively new study proposes a high likelihood that the clay of the vessels of Group A is identical to the clay of semi-fine vessels, which, according to their distribution, were manufactured in Phoenicia (SlaneElam et al. 1994). Type 1 – bowl with narrow ring base (Pl. 4: 1): Diam. 32cm, short, oblique wall, curved rim, short ring base, thick red slip. Imported. Parallels: Hayes Type 2b, second century BCE–mid-first century BCE (Hayes 1985: 14, tav. 1: 6).

38

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Fig. 5: Bowl (IAA) West Slope Ware Vessels (Pl. 11: 1–7) These vessels were inspired by the West Slope Ware vessels of Athens, including similarities in form and decoration, but slipped in red or brownish-red. The vessels from the Akko Marina seem to have originated in the Eastern Mediterranean, apparently in Asia Minor (Rosenthal-Heginbottom 1995: 222). The table amphorae (Pl. 11: 3) resemble amphorae discovered in Pergamon and at Tel Maresha (Kloner 1996: 26). The kantros with the decorated handle (Pl. 11: 1) resembles a vessel discovered in Ephesus (Mitsopoulos-Leon 1992: taf. 32: B38). The plates with the grooved rim (Pl. 11: 4, 5) resemble plates found in Paphos (Hayes 1991: figs. 3:3, 47: 96). The globular bowl with the long, everted rim (Pl. 11: 6) resembles plates found in Samaria (Crowfoot and Kenyon 1957: fig. 47: 9, 10).

Fig. 6: Human and animal figures (IAA)

Mold-made Bowls (Pls. 12–14) Approximately 20 fragments of mold-made bowls were found in the Akko Marina. They may be divided into three groups based on their origin, which is suggested mainly based on their fabric and the quality and color of the slip. The attribution of these bowls to the various production centers is made based on comparisons of the appearance of the fabric with descriptions appearing in various publications. Decoration clearly typical of certain regions, when present, also helps attribute a vessel to a particular geographic region. Vessels from Athens (Pls. 12: 1, 13: 4, 14: 6): late third–mid-first centuries CE (Rotroff 1982: 16, 26, no. 174). Vessels from the Eastern Aegean (Pls. 12: 4–6. 13: 2; 14: 1,2 Figs. 6, 7): Parallels: Dor, second half of second century–early first century BCE (RosenthalHeginbottom 1995: 211); Samaria, mainly second century BCE and continuing into the first century BCE

Fig. 7: Human and animal figures (IAA) (Crowfoot and Kenyon 1957: 274); Delos, late second century BCE (Laumonier 1977: Nos. 3124, 3133). Vessels from the Eastern Mediterranean (Pl. 14: 3–5): Parallels: Dor, not from sealed assemblages, but attributed to the third–first centuries BCE (RosenthalHeginbottom 1995a: 212–213, fig. 5.5: 1–10). Mortaria Mortarium Type 1 – May be divided into two types: Type 1a – mortarium with ridged rim (Pl. 18: 1): Rectangular, ridged, curled rim. Imported. Parallel: Tel Anafa (Berlin 1988: 470, Pl. 25: 35). 39

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Type 1b – mortarium with ridged rim (Pl. 18: 3):

Flared rim, broad with incised lines on the interior and exterior edges. Imported. Parallels: Athens, first half of third century BCE (Thompson 1934: 470, fig. 122: a60, a61). Mortarium Type 2 – Rectangular, ridged rim. (Pl 18: 2). Shallow Cooking Vessels Type 1 – (Pls. 19: 3): Broad, oblique rim with ridge at connection to body, upright, carinated wall, curved base. Parallels: Athens (Thompson 1934: fig. 121: E 141: Rotroff 1982). Type 2 – (Pls. 19: 4, 20: 2): Everted, folded rim, curved wall, curved base. Parallels: Tel Anafa (Berlin 1988: 73–74, Pl. 10: 12–14), Naukratis in the western Nile Delta (Coulson and Wilkie 1982: fig. 4, left). Frying Pans

Fig. 8: Globular jug, omphalos base, pinched spout (IAA)

Type 1 – Orlo bifido (Pl. 23: 1, 2): Incised, inverted rim, flat base. Imported. Parallels: Pompei (Berlin 1988: Chapter 5 n. 11); Caesarea (Oleson et al. 1994: fig. 7k11, k12); and Tel Anafa (Berlin 1988: Pl. 12: 1, 2). Closed Vessels Jugs and Juglets Type 1 – decanter (Pl. 25: 1, 3–6): Ring rim or flared rim, carinated biconical wall or oblate wall with straight shoulder, handle with oval or twisted profile, ring base, broad or narrow, cylindrical neck. Imported. Parallels: Tel Anafa (Berlin 1988: fig. 43: L6); Aegean source (Berlin 1988: 152); Paphos (Hayes 1991: 18, 19: fig. 10: 1); Crete (Hayes 1991: 18, 20, figs. 10: 14, 11: 13). Type 2 – jugs with flared neck (Pl. 25: 7): Thickened rim, flaring neck, strap handle. Imported. Parallels: Dor (Guz-Zilberstein 1995: fig. 6.29: 11, 14, 17); Paphos (Hayes 1991: fig. 43: 42); and Sidi Harabish (Riley 1979a: fig: 135: 1094, 1095). Type 3 – (Pl. 25: 13): Conical, flaring neck, strap handle. Imported (?). Parallels: Tel Anafa Stratum E (Berlin 1988: fig. 36: 5). Type 4 – spherical juglet (Pl. 25: 8, Fig. 8): Spout formed by finger pressure, ribbed wall, omphalos base. Local. Parallels: Caesarea (Oleson et al. 1994: fig. 28: D29). Type 5 – (Pl. 25: 9): Flaring, thickened rim, ribbed wall, ring base, ridged handles. Imported (?). Parallel: Tel Anafa (Berlin 1988: 146, Pls. 38, 39). Amphoriskos Type 1 – (Pl. 28: 4, Fig. 9): Elongated body, tapered base. Imported. Parallels: Coasts of Phoenicia, Palestine and Eastern Mediterranean, second– first centuries BCE (Berlin 1988: 36, 212–214).

Fig. 9: Amphoriskos with elongated body and tapered base (IAA) 40

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Bottles

Table Amphora Type 1 – (Pl 28: 6, Fig. 10): Elliptical body, broad, conical neck, ring base, two loop handles. Imported. Parallels: Tel Anafa, last quarter of second century BCE and onward (Berlin 1988: 158–159, figs. 45, 46).

Unguentarium (Pl. 29: 1–3, Fig. 11): everted or slanted, inturned rim, conical body, thin wall, long, narrow neck, narrow foot base, red slip or no slip. Imported. They can be divided into three main variants: Type 1a – bottle with carinated shoulder (Pl. 29: 1). Parallels: Dor, end of third and second centuries BCE (Guz-Zilberstein 1995: 306, Type UG2e). Type 1b – Bottle with elongated foot and neck (Pl. 29: 3). Parallels: Dor, mid-second century BCE (Guz-Zilberstein 1995: 305). Type 1c – bottle with mushroom-shaped lip, conical body, disc base (Pl. 29: 2, Fig. 12). Parallels: Dor, from second century BCE (Guz–Zilberstein 1995: 306, UG2g); Cyprus (Hayes 1991: 69, Pl. 14: 9–17).

Fig. 10: Hellenistic-period table amphora (IAA)

Fig. 12: Bottle with conical body (IAA)

Fig. 11: Bottles with long foot and neck (IAA) 41

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Fig. 13: Bowl (IAA)

Early Roman Period (Mid-First Century BCE–Mid–Second Century CE)

Type 1 – bowl (Pl. 4: 6, 7): Diam. 20–25cm, oblique, curved wall, tapered rim, broad ring base, thick red slip. Imported. Parallels: Hayes Type 5a, first century BCE–first century CE (very common mainly during the Augustan period). (Hayes 1985: 17, tav. 2: 1). Type 2 – plate (Pl. 4: 8, 9): Diam. 35–47cm, short, stepped wall, everted, sliced-cut rim, decorated in a mold, narrow ring base, thick red slip. Imported. Parallels: Tel Anafa, first half of first century CE (Gunnaweg et al. 1983: 97, fig. 21: 8), Hayes Type 9, late first century BCE (Hayes 1985: 18–9, tav. 2:7). Type 3 – bowl (Pl. 5: 3): Diam. 14cm, curved wall, tapered rim flat base, grooved, thick, peeling red slip. Imported. Parallels: Hayes Type 19a, second half of first century BCE (Hayes 1985:22, tav. 3: 6). Type 4 – small bowl (Pl. 5: 6, 2: 16): Diam. 15– 17cm, curved wall, tapered rim, ring base, thick red slip, incised or rouletted decoration on floor. Imported. Parallels: Hayes Type 22b, second half of first century BCE (Hayes 1985: 23–24, tav. 3: 12–14). Type 5 – krater (Pl. 5: 8,9): Diam. 18cm, curved wall, decorated in a mold, rim was not found, hollow foot base, thick, red slip. Imported. Parallels: Type 26, late first century BCE–first third of first century CE (Hayes 1985: 26, tav. 4:5, 8, 9). Type 6 – plate (Pl. 5: 10): Diam. (?), oblique wall, rim was not found, broad ring base, thick red slip. Imported. Parallel: Hayes Type 30, first half of first century CE (Hayes 1985: 28, tav. 4:15). Type 7 – plate (Pl. 5: 11): Diam. 14cm, upright wall, attachment to floor emphasized with ridge, thickened rim, broad, low ring base, thick red slip. Imported. Parallels: Hayes Type 35, 40–70 CE (Hayes 1985: 30, tav. 5: 8, 9). Type 8 – small bowl (Pl. 5: 12): Diam. 14cm, oblique, curved wall, flared rim, ring base, peeling red slip. Imported. Parallels: Hayes Type 43, last third of first century BCE (Hayes 1985: 33, tav. 6: 8). Type 9 – plate (Pl. 5: 13): Diam. 27cm, upright wall, tapered rim, high ring base, thick red slip. Imported. Parallels: Hayes Type 37, second half of first century CE (Hayes 1985:31, tav 5: 12, 13).

Open Vessels Type 1 – bowls with flat bases (Pl. 3: 6-9; Fig. 13) May be divided into two variants: Type 1a – (Pl. 3: 7-9). Diam. 25–40cm, oblique wall, variously worked rim, flat base, vessels are un-slipped, frequently bearing remnants of yellowish material (sealant?) on interior wall. Local production. Vessels may have been used as jar lids. Type 1b – (Pl. 3: 6). Diam. 32cm, curved wall, flattened rim, flat base (?), fabric contains a great deal of mica. Vessel bears Latin inscription. This bowl is unique in assemblage; no parallels found. Imported. Type 2 – carinated bowls (Pl. 3: 10, 11): Diam. 17–20cm, carinated wall, upper part upright, plain rim, narrow base, ring or flat. Local. Parallels: See below, ETSA Types 34–37). Type 3 – platter (Pl. 3: 14): Diam. 70cm, rather short, upright wall, tapered rim, broad, very low ring base (0.5cm high). Lustrous, thick black slip on interior, black and thin on exterior. Imported (Asia Minor). Parallels: Paphos first century BCE (Hayes 1991: 151, fig. 13: 11). Various red-slipped bowls (Pls. 3: 5, 15–18, 6: 14, 7: 6): Variously shaped red-slipped bowls; they are unique in the assemblage and no parallels were found. Eastern Terra Sigillata A Vessels For the characteristics of this family see above. The types attributed to the Early Roman period include forms inspired by the Roman ceramic tradition. The later types of this family are typically coated with a thin slip of a lighter color than the original, which frequently seems to have been painted on the vessel, as revealed by signs of brush hairs on the smooth face of the vessels.

42

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

(Hayes 1985:46 tav 10: 5). Type 16 – grooved juglet (Pl. 7: 7): Outturned, rim, globular body, incised vertical line, poor-quality orange slip, resembling slip of later types in this family; vertical line incised on wall. Parallels: Umm Aleq, Ramat Hanadiv (Silberstein 2000, fig. 18: 16), Caesarea (on display at the Institute of Archaeology, Mount Scopus). No dated parallels were found. Mold-made bowl (Pl. 14: 9-10): Curved wall, red-

Fig. 14: Bowl (IAA)

Fig. 16: Bowl (IAA) slipped, wall decorated in relief. Eastern Terra Sigillata A Ware – Late Types

Fig. 15: Bowl (IAA)

Toward the end of the first century CE, the manufacturers of Eastern Terra Sigillata A Ware were increasingly influenced by the Roman world of forms, which during that century had increasingly taken over the traditional Hellenistic forms. The later types remained in existence from the end of the first century and until the mid-second century CE. Some of the types are late variations of the early forms of this family. Type 1 – vessels with barbotine decoration (Pl. 6: 11–14, Fig. 16): The decoration appears on cups and bowls (kraters?), and includes strips of clay formed as stems bearing leaves, affixed to the wall, with red or orange slip. Imported. Type 2 – bowl (Pl. 6: 6): Rim (?), oblique wall, gradiated interior, plain rim, low ring base, orange slip. Parallels: Hayes Type 52, first third of second century CE (Hayes 1985: 38, tav 6: 21). Type 3 – bowl (Pl. 6: 7–8): Diam. 16–30cm, oblique wall, gutter rim, low ring base, orange lip. Parallels: Hayes Type 54, first quarter of second century CE (Hayes 1985: 38, tav 7: 4, 5). Type 4 – bowl (Pl. 6: 9): Diam. 17.5cm, oblique wall, everted rim, flat, incised base, red slip. Parallel: Hayes Type 57, first half of second century CE (Hayes 1985:39, tav. 7: 10). Type 5 – bowl (Pl. 6: 10): Diam. 12.5–19.5cm, oblique

Type 10 – cup (Pl. 5: 14, Fig. 14): Diam. 10–12cm, carinated wall, upper part upright, tapered rim, high, narrow ring base, thick red slip. Imported. Parallels: Hayes Type 50, second half of first century CE (Hayes 1985:37, tav. 6: 18). Type 11 – cup (Pl. 5: 15): Diam. 7–12cm, oblique wall, slightly convex, upright, incised rim, high ring base, thick red slip. Imported. Parallels: Hayes Type 47, first half of first century CE (Hayes 1985: 35, tav. 6: 15). Type 12 – bowl (Pl. 6: 1–4): Diam. 7–25cm, curved wall, upright rim, mold-made, foot base, thick red slip. Imported. Parallels: Hayes Type 48, second half of first century CE (Hayes 1985: 36, tav. 6: 16). Type 13 – small bowl (Pl. 6: 5, Fig. 15): Diam. 11cm, curved wall, tapered rim, high ring base, thin orange slip. Parallels: Hayes Type 51, 70–120 CE (Hayes 1985:37, tav. 6: 19, 20). Type 14 – jug (Pl. 7: 10): triangular rim, conical neck, pear-shaped body, strap handle drawn from neck to body, red slip on body, rouletted decoration. Parallels: Hayes Type 111, second half of first century CE (Hayes 1985: 45, tav. 10: 3). Type 15 – table amphora (Pl. 7: 11): triangular rim, short, cylindrical neck, two ridged strap handles emerging from neck, red slip, vertical lines incised on body. Parallels: Hayes Type 114, Roman period 43

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

wall, everted and ridged rim, flat, incised base, red slip. Parallels: Hayes Type 58, 59, first half of second century CE (Hayes 1985: 39, tav. 7: 11–12). Type 6 – bowl with rouletted decoration (Pl. 7: 1–5): Diam. 16–26cm, oblique wall, inverted rim, ring base, red-orange, poor-quality slip, rouletted decorations. Imported. Parallel: Hayes Late Terra Sigillata types, late second century CE (Hayes 1985: 42, tav. 8: 4–15).

Western Terra Sigillata Ware (WTS) (Pl. 9) The typological classification of these vessels has been compared to that of Hayes at Corinth (Hayes 1973) and the classification of Pucci in his summary (Pucci 1985), which mentions most of the types that have been found.3 The dating of each type is based on the dates proposed in these publications.4 Many of the WTS vessels bore the fabricant’s stamp (for WTS parallels see Table 1). (Pl. 10).

Eastern Terra Sigillata D Ware

Thin-walled Vessels (Pl. 11: 8–13)

The vessels in this group typically feature pinkishbrown clay and a very thin slip of a similar color, in some cases lacking luster, perhaps due to the conditions of their preservation. The production centers of these vessels have not yet been found. Negev (1986) proposed that vessels of this group discovered at Avdat were produced in a pottery workshop revealed at the site. Gunneweg, Perlman and Yellin proposed that the production centers were in the area of Pamphilia in Asia Minor (Gunneweg et al. 1983: 101–103, 107: fig. 24: 2). However, according to the quantity of these vessels and their many types found in Cyprus, that locale, or a place close to it on the mainland, seems to have been the origin of this group (Hayes 1991: 37). Type 1 – krater (Pl. 8: 4): Diam. 23–38cm, curved wall, grooved rim, belongs to a krater which has a tripod base with “bone-shaped” legs, thin brownpink slip, wall frequently roulette decorated. Type 2 – bowl (Pl. 8: 1): Diam. 30cm, curved wall, curved rim, broad ring base, low, thin, brown-pink slip. Parallels: Hayes Type p12, mid-first century CE–mid-second century CE (Hayes 1985: 83, tav. 19: 4, 5). Type 3 – bowl (Pl. 8: 2): Diam. 21cm, carinated wall, upper part upright, everted, gutter rim, ring base, thin, brown-pink slip. Parallels: Hayes Type p28, first century CE (Hayes 1985: 85–86, tav. 20: 7). Type 4 – bowl (Pl. 8: 3): Diam. 17cm, oblique wall, inverted rim, ring base, thin, brown-pink base. Parallels, Hayes Type p29, first half of second century CE (Hayes 1985: 84, tav. 20: 8). Type 5 – juglet (Pl. 8: 5): Conical, narrow, elongated body, flat base, strap handle drawn from rim to upper third of body. Parallels: Hayes Type 58x, late first century–second century CE (Hayes 1985: 90–91, tav. 22: 4).

This family includes vessels with thin walls originating in Italy or in the Aegean region (Hayes 1992: 59–60). Few fragments of vessels of this family were found in the Akko Marina, perhaps because of their delicate nature, which reduced their chances of surviving in the circumstances under which the marina excavation operated. Some of the vessels attributed to this family are imitations of original forms; they are differentiated from those forms by their relatively thick walls. Type 1 – cups with designed handles (Pl. 11: 11,12): Parallels: Corinth, first half of first century CE (Hayes 1973: 466, Pl. 90: 231); Paphos, second half of first century BCE (Hayes 1991: fig. 62: 31). Type 2 – spherical cup (Pl. 11: 10): Parallels: Cosa, mid-first century CE (Marabini: Form 71). Type 3 – cups of the Urnetta a Collarino type (Pl. 11: 8, 9): Parallels: Paphos, mid-second century CE (Hayes 1991: fig. 22: 17); Knossos, mid-second–midthird centuries CE (Hayes 1983: 107). Cooking Casseroles Type 1 – (Pl. 19: 9, 10): Broad rim, look handles. Locally made. Parallels: Tel Anafa (Berlin 1988: 76–77, Pl. 11: 19, 20). Type 2 – (Pl. 20: 3): Broad rim cut obliquely, emphasized attachment of rim to body. Parallels: Tel Anafa (Berlin 1988: 71–72, Pl. 9: 6–8).

1 We have chosen to direct readers to these publications because the assemblages from Corinth include most of the types found at Akko. The advantage of Pucci’s discussion is the large number of assemblages published, on which the typological division and chronological framework are based. 2 Most of the vessels from the Akko Marina are types also produced in the workshops of Arezo, which existed from the time of Augustus until the mid-first century CE, as well as the late Italian workshops that began operation in the mid-first century CE. The certain identification of the workshop, whether based on the fabricant’s import, or on the clay, will be able in the future to assist in determining more precise dates for these vessels.

44

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Shallow Cooking Vessels

Closed Vessels

Type 1 – (Pl. 19: 5, 6): Ledge rim with strip of clay adhering to it – a kind of handle – ribbed, oblique wall. Import. Parallels: Tel Mevorakh (Stern 1978: fig. 2: 14). Type 2 (Pl. 19:2): Broad, cut and oblique rim, attachment of rim to wall emphasized by ridge. Local (?) Parallel: Tel Anafa (Berlin 1988: 71–72, Pl. 9: 6–8).

Jugs and Juglets Type 1 – juglet with disc base (Pl. 27: 9): Imported. Parallel: Cyprus (Westenholm and Vessberg 1956: fig. 24: 16–18); Caesarea Harbor (Oleson et al. 1994: fig. 28: C24). Type 2 – elliptical juglet (Pl. 27: 12, Fig. 17): Everted rim, flat base. Local. Parallels: Samaria, Herodian period (Crowfoot and Kenyon. 1957: fig. 67: 5). Type 3 – juglet with rounded body (Pl. 28: 1): Plain rim, loop handles, body covered with light-colored slip with narrow incised lines. Imported. Parallels: Caesarea Harbor, first century CE (Oleson et al. 1994: fig. 29: D26).

Frying Pans Type 1 – (Pl. 23: 6): Plain rim, rounded wall, flat base. Import. Parallels: Paphos (Hayes 1991: fig. 28, 2, 4). Type 2 – (Pl. 23: 5): Triangular rim, short, oblique wall, flat base. Imported. Parallels: Tel Anafa (Berlin 1988: Pl. 13: 6–8). Type 3 – (Pl. 23: 9, 10): Thickened rim, short, oblique wall, cylindrical handles attached vertically to body or thumb-indented handles attached to the rim. Imported. Parallel: Sidi Harabish (Riley 1979a: 254–255).

Amphoriskoi Type 1 – amphoriskos with ribbed body (Pl. 28: 5): Ribbed body with tiny handles attached to lower part of neck. Imported vessel (?) No parallels. Table Amphora Type 1 – (Pl. 28: 8): Conical neck, swollen upper part. No parallels were found. Table Amphora Type 2 – (Pl. 28: 10): Piriform body, broad neck, ridged strap handles, disc base. Imported vessel. No parallels.

Cooking Pots Type 1 – (Pl. 24: 2): Rim everted downward, carinated shoulder, very thin wall. Imported. Parallels: Paphos (Hayes 1991: 82, figs. 33: 3, 64: 72). Type 2 - (Pl. 24: 9): Thickened rim, globular body, curved wall, loop handles drawn from shoulder, flat base. Local vessel. Parallel: Ben Shemen (Porath 1990: 162, fig. 3).

Fig. 17: Ovoid juglet (IAA)

Fig. 18: Pear-shaped bottle (IAA) 45

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

About half the vessels were locally produced (Syria and Palestine) or were imported from the southwestern Black Sea (distribution among the regions is almost equal). The other vessels (42%) originated in the Aegean, North Africa, Italy, southern France, and the Western Mediterranean. The origin of the remaining 8% of the vessels was unidentified.

Late Roman Period (Mid-Second–Early Fourth Centuries CE) Open Vessels Red Slip Ware The types in this family are plates or bowls, which are typically slipped in shades of red. This family includes three main groups: African Red-Slip (ARS), Group D (LRD) and Group C (LRC) which is associated with Cypriot origin. Vessels of this family replaced Terra Sigillata Ware in the import market of quality tableware during the Late Roman and the Byzantine periods. The presentation of the vessels below is based on Hayes’ (1972) accepted typological classification and dating.

Fig. 19: Pear-shaped bottle (IAA) African Red-Slipped Ware

Bottles

Type 1 – narrow bowl, Hayes Form 8A (Pl. 15: 1): Diam. 15–22cm, triangular rim with thickened ridge at its base; oblique wall, flat base, rouletted decoration on rim and bottom of wall. Date: late first–mid-second centuries CE. Type 2 – bowl, Hayes Form 9B (Pl. 15: 2): Diam. 22cm, plain rim, curved wall, ring base, two broad incised lines under rim. Date: second half of second century CE. Type 3 – lid, Hayes Form 20 (Pl. 15: 3): Diam. 22cm, oblique wall everted downward, broad rouletted decoration on edges of lid. Date: late first century–early second century CE. Type 4 – bowl, Hayes Type 23B (Pl. 15: 4): Diam. 30cm, thickened rim, wall everted slightly downward, crossed by thick ridge at base, coarsely worked at bottom of wall, base slightly convex, covered with dense concentric ribbing and soot. Date: mid-second–early third centuries CE. Type 5 – bowl, Hayes Type 27 (Pl. 15: 5): Diam. 29cm, plain rim, curved wall, thin, broad ring base. Date: mid-second–early third centuries CE. Type 6 – bowl, Hayes Type 50 A/B (Pl. 15: 6): Diam. 39cm, tapered rim, broad, slightly sunken base. Date: mid-third–mid-fourth centuries CE. Type 7 – bowl, Hayes Type 31 (?) (Pl. 15: 7): Diam. 27cm, tapered rim, oblique wall, low ring base, bands on wall decorated with rouletting . Date: first half of third century CE. Type 8 – bowl, Hayes Type 45C (Pl. 15: 8): Diam. 34cm, broad, everted rim, low curved wall. Date: first

Type 1 – piriform bottle (Pl. 29: 4, Fig. 18): Imported vessel(?). Late first century BCE. Distribution Eastern and Western Mediterranean. Parallels: Ramat Hanadiv (Silberstein 2000: Pl. 4: 21–24). Type 2 – ribbed bottle (Pl. 29: 5, Fig. 19): Everted rim, piriform body, ribbed wall, broad, spayed neck. Imported vessel. Parallels: Paphos (Hayes 1991: 84, fig. 35: 3). Type 3 – goblet with flat base (Pl. 29: 6): Slightly everted rim, ribbed wall, flat base, carelessly made. Local vessel. Flasks Type 1 – ribbed, globular flask (Pl. 29: 12): Local vessel? Parallel: Tel Anafa (Berlin 1988: 208–209, Pl. 64: 1, 2). Amphorae (Pls. 36-40) One hundred and forty-five amphorae fragments from the Roman period were identified. Most of the Roman amphorae belong to the Early Roman period, and they are presented together with the amphorae of the Late Roman period. Amphorae types and their parallels are presented in the plates and in Table 4.

46

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

half of fourth century CE. Type 9 – bowl, Hayes Type 46 (Pl. 15: 9): Diam. 28cm, broad rim everted, drawn downward, short, thin, curved wall. Date: last quarter of third–first quarter of fourth century CE. Mold-made bowl – (Pl. 14: 7): Upright wall, flat base, thin, peeling red slip. Relief of human figure, vegetation and tripod table on wall. Parallels: Corinth, mid-second–mid-third centuries CE (Spitzer 1942: Type II).

Type 2 – (Pl. 24: 6–8): Overhanging rim, arched neck pinched at base, loop handle. Local vessel. Parallel, Tel Shikmona (Algavish 1976: Pl. 15: 121, 122); Jalame (Johnson 1988, Form 22) and Samaria (Mount Gerizim) (Diez Fernandez 1983: Nos. 328, 329, 360). Closed Vessels Jugs and Juglets Type 1 – globular jugs (Pl. 26: 4, 9): Gutter rim, curved wall, spout finger-pinched. Imported vessel. Parallels: Cyprus, second and third centuries CE (Vessberg and Westholm 1956: fig. 31: 14). Type 2 – jug with flat base (Pl. 26: 1): Flared rim, ribbed wall, conical body, flat base. Local vessel. Parallels: Jalame (Johnson 1988: fig. 7.46: 697). Type 3 – strainer jugs (Pl. 26: 2): Local vessel. Parallels: Roman and Byzantine types (Diez Fernandez 1983, tipo 5.1; Johnson 1988: fig. 7.45: 670; Calderon 2000: Pl. 8: 54). Type 4 – jug with trefoil rim (Pl. 26: 3): Trefoil rim, incised neck. Imported vessel. Parallels: Corinth, first century CE (Hayes 1973, Pl. 79d); Sidi Harabish second and third centuries CE (Riley 1979a: 384– 385, fig. 138: 1145); Athens mid-third century CE (Robinson 1959: Pl. 14: k83). Type 5 – jug with long, cylindrical neck (Pl. 26: 5, 6): Grooved strap handle drawn from lower part of neck to shoulder, narrow, hollow base. Imported vessel. Parallels: Eastern Mediterranean, first–fourth centuries CE (Riley 1979b: 183–185) and summary by Peacock and Williams (1986, Class 45). Type 6 – jug with trefoil rim (Pl. 26: 8): Cupped, trefoil rim, ridged handle drawn from rim to shoulder. Local vessel. Parallels: Beit She‘arim, third–fourth centuries CE (Mazar 1973: fig. 23: 4). Type 7 – cup with flared rim (Pl. 25: 11): Local vessel. Parallels: Sidi Harabish, second–fourth centuries CE (Riley 1979a: fig. 139: 1152). Type 8 – juglets with cylindrical, long necks (Pl. 26: 10–12): No parallels found. Type 9 – juglet with pinched foot base (Pl. 27: 7, 8, Fig. 20): Flared rim, handle drawn from rim to wall, pressed foot base. All imported vessels. Parallels: Tel Shikmona, second century CE (Elgavish 1977: Pl. 11: 9); Dor, late first–mid-second centuries CE (GuzZilberstein 1995: 332–333, fig. 6: 49: 18). Type 10 – juglet with flared rim (Pl. 27: 10, 11): Conical, elongated neck. Imported. No parallels.

Shallow Cooking Bowls Type 1 – (Pl. 20: 1): Gutter rim, S-shaped wall, curving based, flat loop handles. Imported vessel (?). Parallels: Kefar Hananya (Adan-Bayewitz 1993: 119–124). Type 2 – (Pl. 19: 7): Grooved rim, curved wall, slightly convex base. Local vessel. Parallel: Kefar Hananya (Adan-Bayewitz 1993: 88–91). Type 3 – (Pl. 19: 8): Thickened, double-incised rim, curved wall, ridged loop handle. Local vessel. Parallel: Kefar Hananya (Adan-Bayewitz 1993: 88–91, Type 1B). Type 4 – (Pl. 21: 1, large exemplar, Pl. 21: 2, small exemplar): Broad, oblique rim, curved, ribbed wall, ridged loop handle. Imported vessel. Parallels: Knossos (Hayes 1983: 105, fig. 5: 59–63). Large vessels: Knossos (Hayes 1983: fig. 6: 67–71). Type 5 – shallow cooking vessels – varia. These may be divided into three variants:

Type 5a – carinated krater (Pl. 21: 4, 5). Everted rim, handles drawn from below rim to wall. Imported. Type 5b – carinated krater (Pl. 21: 3). Rim everted downward. Imported vessel. Parallel: (Hayes 1983: fig. 7: 81–89, 90). Type 5c – small carinated krater (Pl. 21: 6). Everted rim, ridged handles pulled from rim. Imported vessel. Parallel: (Diez- Fernandez 1983: 210, T 11.4). Frying Pans Type 1 – (Pl. 23: 7): Ledge rim, oblique wall, convex base. Imported vessel. Parallels: Knossos (Hayes 1983: 107–108, fig. 9: 103–109, Type 2). Type 2 – (Pl. 23: 13): Ledge rim, curved wall, elongated handle with three incised lines and perforation at end connected to rim. Imported vessel. Parallels: Sidi Harabish (Riley 1979: fig. 105: 528–529). Cooking Pots

Bottles

Type 1 – (Pl. 24: 3–5): Grooved rim, upright neck, thin wall. Local. Parallels: Kefar Hananya (AdanBayewitz 1986).

Type 1 – elongated bottle (Pl. 29: 7–10, Fig. 21): Narrow neck, ridged at its base, button base, brown

47

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Fig. 20: Juglets with foot base (IAA) slip. Imported vessel, fifth–seventh centuries CE (Hayes 1971). Flasks

the latter attributed to a Cypriot origin.

Type 1 – flask (Pl. 29: 11): Unique vessel with hanging handles and additional ring base on one of the outer walls. Imported. No parallels were found.

Type 1 – plate, Hayes Form 58 (Pl. 15: 10): Diam. 30–43cm, flat, everted rim, curved wall, sunken base. Date, first three quarters of the fourth century CE. Type 2 – plate, Hayes Form 59A (Pl. 15: 12): Diam. 32cm, flat rim, double-incised, curved wall, decorated with vertical lines not densely spaced. Date: fourth century CE. Type 3 – plate, Hayes Type 61A (Pl. 15: 13): Diam. 28cm, plain rim, oblique wall, sunken base. Date: last three quarters of fourth century CE. Type 4 – plate, Hayes Type 67 (Pl. 15: 14): Diam. 31cm, triangular rim, oblique, carinated wall. Date: mid-fourth–mid-fifth centuries CE. Type 5 – plate, Hayes Type 68 (Pl. 15: 15): Diam. 32cm, triangular rim, oblique wall. Date: last quarter of fourth–first quarter of fifth centuries CE.

African Red Slip Ware

Byzantine Period (Second Third of the Fourth Century–First Third of Seventh century) Open Vessels Red Slip Ware Two more groups belonging to this family in addition to the ARS Ware are known during the Byzantine period, beginning in the Roman period: Late Roman Group D (LRD) and Late Roman Group C (LRC),

48

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Type 6 – bowl, from Hayes Form 91B (Pl. 15: 16): Diam. 26cm, wide, pulled-down rim,, curved wall. Date: mid-fifth–mid-sixth centuries CE. Late Roman D Ware Type 1 – bowl from Hayes Form 2 (Pl. 16: 1, 2): Diam. 18–22cm, triangular rim, curved wall, flat base. Date: late first–early sixth centuries CE. Type 2 – bowl from Hayes Form 7.1 (Pl. 16: 3): Diam. up to 30cm, triangular, thickened rim incised three times, oblique wall, decorated with a few rouletted bands on wall. Date: mid-sixth–mid-seventh centuries CE. Type 3 – bowl from Hayes Form 9B (Pl. 16: 4): Diam. 24–27cm, thickened inverted rim, oblique wall, flat base, decorated with rouletted bands on wall, broken incised line on rim. Date: late sixth– seventh centuries CE. Type 4 – bowl from Hayes Form 10 (Pl. 16: 5): Diam. 25–31cm, triangular, double-ridged rim, oblique wall, flat base, decorated with rouletted bands on wall. Date: mid-seventh century (?) CE. Type 5 – bowl from Hayes Form 11 (Pl. 16: 6): Diam. 30cm, outfolded rim, curved, ribbed wall, flat base. Date: mid-fifth–mid sixth centuries CE. Late Roman C Ware Type 1 – bowl from Hayes Form 8 (Pl. 16: 7): Diam. 24cm, convex rim, ridged, down-turned, curved wall. Date: second half of fifth century CE. Type 2 – bowl from Hayes Form 3 (Pl. 16: 8, 9): Diam. 19–20cm, concave rim, everted downward, ridged, curved wall. Date: beginning of fifth–midsixth centuries CE. Type 3 – bowl from Hayes Form 10A (Pl. 16: 10): Diam. 22cm, square rim, oblique wall. Date: late sixth–early seventh centuries CE. Varia – two ring bases of bowls (Pl. 17: 1, 2): Attributed to the Late Roman Ware according to quality of fabric and floral and geometric patterns in concentric frames. No parallels found. Mortaria Type 1 – Syrian mortarium (Pl. 18: 6–8, Figs. 22, 23): Thick rim drawn downward, spout formed as part of rim. Imported vessel: Parallels: El-Bassit, end of second–fifth centuries CE (Hayes 1967); Caesarea (Riley 1975: 41, no. 67, fig. 4). Type 2 – mortarium (Pl. 18: 4, 5): Square ledge rim with spout, flat base. Imported vessel. Parallels: Athens

Fig. 21: Vertical bottle (IAA)

49

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Fig. 22: Syrian mortarium (IAA) Imported vessel. Parallel: Kelia, Egypt (Egloff 1977: Types 118–120). Frying Pans Type 1 – (Pl. 23: 8): Horizontal, ridged handle. Local vessel. Parallels: Tel Shikmona (Elgavish 1977, Pl. 4:20); Jalame (Johnson 1988: fig. 7. 43: 630). Cooking Pots Type 1 – (Pl. 22: 4, 5, 7, 9): Loop handles drawn from rim to shoulder, ribbed body. Parallels: Ramat Hanadiv, Jalame (Adan-Bayewitz 1993: 132–135, form 4E2; Calderon 2000: 108, Pl. 7:32).

Fig. 23: Seal impression on mortarium (IAA)

Closed Vessels

(Robinson 1959: Pl. 14: k90); Jalame, first half of fourth century CE (Johnson 1988: 183, figs. 7.31: 489, 490).

Jugs and Juglets

Shallow Cooking Vessels Type 1 – (Pl. 26: 7): Rim folded inward, ribbed wall, small spout, local vessel.

Several members of the krater family have also been attributed to the Roman-Byzantine period. Most were found sporadically in the assemblage. The type that appears in significant quantities is the krater with horizontal handles, below, Type 1. Type 1 – krater with horizontal handles (Pl. 22: 1): Horizontal handles, local vessel. Parallels: Caesarea (Magness 1992: 132). Type 2 - (Pl. 22: 2): Sunken rim, crescent-shaped lug handle. Parallels: Cosa (Marabini Moevs 1978: figs. 66: fc23, 67: fc24). Type 3 (Pl. 22: 3): Square rim, ribbed wall, imported vessel. Parallel: Carthage (Fulford and Peacock 1984: Type 26.6). Type 4 - (Pl. 22: 6): Sunken rim, horizontal handles.

Table Amphorae Type 1 – (Pl. 28: 9): Rim slightly everted, strap handles, biconical body, ring base. Type 2 – (Pl. 28: 7): Broad rim, broad neck, strap handles with two ridges, narrow, flat base. Imported vessel. Amphorae (Pls. 41–46) Two hundred and thirty-five amphorae fragments from the Byzantine period were counted. The

50

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

various types and parallels are described on the plates and in Table 4. The source of 64% of the amphorae is local production (Syria and Palestine) or imports from the Aegean. About one tenth (9%) of the amphorae were imported from the western and southern parts of the Black Sea, from North Africa and from the Western Mediterranean. The origin of the remaining vessels was not identified. Notably, during the Byzantine period, the number vessels grows but the number of types declines. A grouping containing a few dozen amphorae of two types belonging to the Byzantine period were found in Area C near the entry to the western basin.

is a ridge higher than the shoulder. The lamps in the assemblage have an oblique shoulder typical of the later variant attributed to 220–140 BCE. At Dor, lamps of a similar shape were attributed to an Attic source in assemblages dating from 250 to 150 BCE (Rosenthal-Heginbottom 1995: 235, Type 9, figs. 1:13, 2: 1–5). Type 4 – ‘Ephesus’ lamps (Howland Type 49) (Pl. 30: 6, 8): Lamps of this type are characterized by gray, micaceous clay and black slip. It was produced in Asia Minor beginning in the second century BCE. Imitations of the type in other workshops continue to appear until the end of the first century BCE (Rosenthal and Sivan 1978:12). Number 6 features a radial decoration with a relief of the face of Silenus (?) on the nozzle and two lug handles flanking the nozzle. An identical lamp was discovered in Samaria (Crowfoot and Kenyon 1957: fig. 87:6). Number 8 has a collar separating the sunken discus, in which there are three perforations. The discus is decorated with a wreath. An identical lamp was found in Athens (Howland 1958: fig. 55). Type 5 – mold-made lamp (Pl. 31: 1): Rim and sides decorated with dots in relief, discus is sunken and undecorated, a strap handle features two grooves, with a prominent knob on the side. The type was produced in Italy in the last part of the first century BCE (Rosenthal and Sivan 1978: 16, no. 32). Type 6 – mold-made lamps with large discus decorated in relief (Pl. 31: 2–4; Fig. 25): These lamps can be attributed to one of three Broneer Types: XXI– XXIII, dating to the end of the first century BCE–early second century CE (Broneer 1977). These types differ from one another in the form of the nozzle and the handle. No remnants of nozzles were found among

Oil Lamps from the HellenisticByzantine Periods Type 1 – lamp with elongated socket carried with a stick (Pl. 30:1): This type of lamp is the product of the Attic industry, dated to the fourth–early third centuries BCE. A fragment of one of this type was found in the Akko Marina, made of yellowish clay with red slip typical of the vessels from the late Hellenistic period. The fragment goes together with a ring base and it seems that the perforation of the lower end of the nozzle was done after firing. This may have been a (late?) imitation of the Attic type, but it may also not have been a lamp and its similarity in form to the above type of lamp may be coincidental. Type 2 – lamp with curved side and ring base (Pl. 30: 2): Parallels of local production, without ring base (RosenthalHeginbottom 1995: 235, Type 6, fig. 5.14: 1–7). Type 3 – lamp with carinated shoulder (Pl. 30: 3–5) (Fig. 24): Perforated lug handle, raised base, sunken in middle. Rim has two sub-types: 1. Rim is sunken and surrounded by a narrow groove; 2. Rim

Fig. 25: Mold-made lamp with large discus (featuring human face wearing helmet) (IAA)

Fig. 24: Lamp with carinated shoulder (IAA) 51

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Fig. 26: Mold-made lamps with rounded nozzle (IAA) the fragments of discus discovered in the assemblage and therefore it is difficult to determine to which of the three types they should be attributed. The single fragment of nozzle that was discovered (no. 4) belongs to a lamp of Type 23. Type 7 – round, mold-made lamp with rounded nozzle (Pl. 31: 5-7, Fig. 26): This type begins to appear in the mid-first century CE and its main distribution is during the second century CE. It continues to appear in the early third century CE as well. Missing from our assemblage is the Syrian-Palestinian variant, characterized by a decoration of double axes on the rim and schematic volutes flanking the nozzle. Apparently all lamps of this type are imported. Type 8 – ‘ovoid’ lamp with large discus decorated in relief (Pl. 31: 6; Fig. 27): Lamps of this type were discovered at Dor where they were dated to the third century CE (Rosenthal-Heginbottom 1995: 246, Type 27). Type 9 – Deba‘al lamp, mold-made (Pl. 31: 8): The discus is decorated with concentric grooves, with oblique grooves flanking the nozzle. The base is flat and emphasized with two concentric grooves. This type is dated to the third century CE and it is quite common in the Eastern Mediterranean (Rosenthal Higenbottom 1995: 274, Type 29). Type 10 – round lamp with rounded nozzle and rim decorated with a row of knobs (Pl. 32: 3): Belongs to a type found in Palestine and attributed to the third– fifth centuries CE (Magness 1993: 6, Form 1). Type 11 – ovoid lamp with lug handle and small filling hole in center of discus (Pl. 32: 6): This type was found mainly in northern Palestine and Lebanon

Fig. 27: ‘Ovoid’ lamp with large discus decorated in relief (IAA) during the entire Byzantine period (Rosenthal and Sivan 1978: 123). Type 12 – (Pl. 32: 5) – lamp with broad rim, elongated lug handle attached to shoulder: North African from Hayes Type II, dated to the mid-fifth– sixth centuries CE (Hayes 1972: 310). Type 13 – round lamp with round nozzle and sunken discus separated by a very prominent ridge (Pl. 32: 1): Similar lamps were discovered in the Yassi Ada shipwreck (Bass and Van Doorninck 1992: Pl. 25: 123, 125) and in Sarachane (Istanbul), Turkey, where they were attributed to the sixth–early seventh centuries CE (Hayes 1992: Pl. 25: 123, 125). 52

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Vessels with the same form but of a different fabric were frequently found, showing that it is impossible to unequivocally match form to geographic origin. Among the lids of the cooking vessels are those whose form resembles Pompeian Red Ware (Pl. 33: 1) which was imported to Palestine from central Italy (Berlin 1993: 37–38). Another type, featuring ribbing and a hollow button handle (Pl. 33: 6), apparently served as cooking kraters of a type found in Caesarea (Adan-Bayewitz 1986: fig. 4: 8, 9). There are also ribbed lids with a non-hollow button handle (Pl. 33: 3) or a lug handle with a flat top bearing string-cut signs (Pl. 33: 5). At Dor, these lids were identified as cooking krater lids and they are found in assemblages from the Hellenistic period (Guz-Zilberstein 1995: 302, fig. 6.42:4–9). The tiny lid with a pinched lug handle (Pl. 33: 8) resembles lids found in Paphos and in Sidi Harabish (Hayes 1971: fig. 17:485; Riley 1979a: fig. 119: 800). Amphorae lids were mainly handmade and are finished rather carelessly (Pl. 33: 7–10, Fig. 28). Nevertheless, some of these lids have parallels at other sites and it seems that their production was inspired by other types on the market. Small, flat lids with reliefs of crosses on their faces and with an applied cube-like handle (Pl. 33: 10) resemble lids found at Sidi Harabish in an assemblage that was mainly Hellenistic–Early Roman (Riley 1979b: 176, fig. 81: 213) and at Paphos (Hayes 1991: Pl. 23: 7, 8). Lid No. 7 on Pl. 33 has double edges and perforation that were found filled with resin that coated the inner side of the lid. Despite the complexity of the lid, it, too, is made quite carelessly. The last use of the lid, perhaps after the perforations were stopped up, may have been different from its original purpose. One lid has an unusual shape – it is a kind of piriform bottle with a broadening spout whose edges are upright, at the bottom of which is a narrow foot (Pl. 33: 2, Fig. 29).

Hayes surmises that the lamps of this type, which were quite widely distributed in the region of Turkey originated in Knidos. Type 14 – channel lamps (Pl. 32: 4, 7): The lamps with the grooved decoration are common in Palestine in the seventh–eighth centuries CE (Magness 1994, Form 4B); those with a geometric design appear in the eighth–tenth centuries CE (Magness 1993, Form 5). Type 15 – lamp with ivy leaves on the shoulder (Pl. 32: 2): Hellenistic period, import. Lamps – varia: A mold-made lamp with shoulder decorated with vegetative designs (?) on the nozzle, an amphora and on the base, rosette with eight petals, alternating between heart-shaped and pointed (Pl. 30: 7). Rosette decoration on the base is also seen on Holander’s lamp Type 50 (Rosenthal and Sivan 1978: no. 38). Lamps – summary: Ten lamps are attributed to the Hellenistic period, all of them apparently imported. The source of the Hellenistic lamps is Athens (?) and Asia Minor. The most common lamps in Palestine in this period, such as lamps with radial decoration and round, wheel-made lamps, are missing in the assemblage, which is not an assemblage typical of Palestine in this period. Also in the assemblage are 13 lamps from the Roman period, mostly imported from Italy and from the provincial workshops in the Eastern Aegean. The assemblage does not feature typical lamps for the period in Palestine, such as round lamps with a double ax. The small number of lamps in general and the lack of local lamps in particular, indicates that the lamps found in the Akko Marina came from ships anchored there and were apparently used on board and not intended for trade. Four lamps are attributed to the Byzantine period – two local and two imported. The decline in activity during this period is also reflected both in the small number of lamps in general and the small number that may reasonably be assumed arrived with ships passing through the port. Only two of the lamps are attributed to the Early Islamic period. Even these, of types common in the country at the time, might not necessarily be connected to commercial activity in the port. The general picture in terms of lamps thus shows meager trade activity in the area of the port during the Early Islamic period. Lids (Pl. 33) Dozens of lids and fragments of lids were discovered in the assemblage, among them lids of cooking vessels and lids of jars and amphorae. The amphora lids featured the remains of resin used as a sealant. Some of the lids were handmade and feature no special characteristics that allow them to be classified into types. The rest of the lids were classified by form.

Fig. 28: Lid in the form of a piriform bottle (IAA) 53

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Because the base types are of such a wide variety, they can be used to fill out the general picture of the amphorae as it emerges from the amphorae fragments that have been presented above chronologically. Dozens of amphora handles with stamps were also recovered. These are studied in a separate article (see Finkielsztejn this volume). Varia 1 – Stand for a cooking pot (Pl. 33: 11, 12). 2 – Clay face mask with red slip (Pl. 33: 13).

Graffiti Some vessels and vessel fragments found in the assemblage bore incised, engraved or drawn graffiti (see also Friedman et al. this volume). The engravings were done after the production of the vessels and therefore it seems likely that they were incised by their owners rather than their potters. Almost all the graffiti appear on high-quality vessels that were imported to Akko from elsewhere, apparently by sea. Most of the graffiti were found on Eastern Terra Sigillata vessels, mainly on bowls of Form 22. Some of the graffiti were found on Western Terra Sigillata vessels defined in this report as Type 6. One graffito was found on a Late Roman Ware C bowl and another was found on an imported bowl from an unidentified origin. Graffiti were also found on two Hellenisticperiod lamps (Pl. 30). The graffiti were usually found incised in hidden areas of the vessels: on the bottom or on the lower outer wall. An unusual engraved graffito – AKPATON – was found on the shoulder of a decanter. The inscription was deciphered by Vassilios Tzaferis (pers. comm.) as "new wine" (Fig. 30). Apparently the inscription was intended to identify the contents of the decanter. The graffiti may be divided into two categories: 1. Combinations of letters that make full names such as LEON, CIMW or abbreviated names; 2. Forms, usually carelessly incised crossed lines, apparently as some sort of mark, or possibly for no defined purpose. Two unusual drawings were found: One is of an anchoring ship with its bow tied to a tree and the upper part of the anchor rope depicted stretched from the stern downward (Pl. 34: 18, Fig. 31). Alongside the ship is the monogram ПP, which may indicate the city of Paphos, whose ceramic assemblage from the Hellenistic and Early Roman periods very closely resembles the contemporaneous assemblage from the Akko Marina (Galili et al. this volume b). Another unique engraving includes five, two-arm anchors of the type in use in the Roman period. The drawing was found on a vessel of a type attributed to this period (Fig. 32). The only drawing incised on the visible inner side of wall of a vessel includes the feet of a bird, on which spurs can clearly be seen (Pl. 34:17). This may have been intended as a further decoration of the vessel.

Fig. 29: Lid in the shape of a piriform bottle with narrow foot (IAA) This type is common at various sites along the Mediterranean coast (Oleson et al. 1994: 122, fig. 44: K80 and see further examples there). Its unusual form has led some scholars to conclude that this kind of lid was used to measure some liquid that was stored in an amphora and our lid may have had a similar purpose. In addition to the amphora lids, bowls were discovered bearing the remains of an unidentified coating on the inner side (above, Pl. 3: 7, 8). These could have served as lids and measuring devices. Amphora Bases (Pls. 46–49) From a morphological point of view it is difficult to precisely identify the amphora bases discovered separately rather than attached to their vessels. Because there are no complete profiles of amphorae in the assemblage and only one intact amphora was found, the bases are presented separately. The bases are important in that they are made of a particularly thick fabric and are usually preserved on the seabed. 54

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Fig. 30: Shoulder of decanter with inscription “new wine” (IAA)

Fig. 31: Graffito of an anchoring ship moored to a tree, at bottom of clay vessel (IAA) Fig. 32: Bowl with graffiti of anchors (IAA)

Summary of Ceramic Finds The ceramic finds from the Hellenistic, Roman and Byzantine periods are shown in detail in Tables 1–6.

2). With regard to cooking vessels (Table 3), jugs and juglets (Table 4), the number of these vessels increased in the Roman period and declined in the Byzantine period. Notably, a gradual increase over time is seen in the quantity of amphorae (Table 5). The assumption is that the earlier the vessels are, the lesser the chance of their preservation and discovery (they are found deeper beneath the seabed). Thus, the

Changes in the Quantity of Vessels over Time During the Hellenistic period the quantity of bowls was very large and declined from then on (Table

55

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

the relative number of bowls did drop during the Roman and Byzantine periods. This trend may be explained in the change in the countries from which the bowls originated: During the Hellenistic period bowls were imported mainly from the northeastern Mediterranean (the Aegean, southern Asia Minor and northern Syria), which are not far from Akko and were also under the political control of the Seleucids. However, during the Roman and Byzantine periods, the bowls came mainly from southern Italy and North Africa, which were much farther from Akko. We can also observe that the decline in the relative quantity of bowls in the assemblage led to an increase in the relative numbers of amphorae (Table 6).

decline in numbers of vessels during the Byzantine period apparently shows a significant downswing in activity at the port. Unusual in this respect are the amphorae, whose numbers increased during the Byzantine period (Table 5). Relative Quantities of Imported Vessels in the Assemblage The large quantity of imported vessels stands out. As noted, there may have been imported vessels that were not identified as such, and so their relative numbers may be even greater. The percentage of vessels is especially great in terms of amphorae, which by their nature involve import activity. The percentage of imported bowls among all bowls is also great, because there were bowls in every period among the main high-quality vessels brought to the country. During the Byzantine period imported bowls of the Late Roman Ware family constitute quite a large percentage of all bowls discovered in assemblages in Israel, and it seems that they were very common in the markets in this country, even those at some distance from Israel. A special aspect of the assemblage from the Akko Marina was observed with regard to jugs and cooking vessels. The study of assemblages from our region, that included jugs and cooking vessels, shows that these vessels were not imported for sale locally. However, in the assemblage from the Akko Marina, these vessels constitute around 50% of imports during the Hellenistic and Roman periods. These vessels must have originated in ships anchoring in the harbor. During the Byzantine period, in contrast, the percentage of imported jugs and cooking vessels dropped. This decline conforms to the general trend in this country during the Byzantine period and is true for amphorae as well. It is possible that the decline reveals a downturn in commerce at this time in Akko. These finds may indicate a decline in the usability of the harbor in that period. This hypothesis is supported by other finds from the marina, e.g., the broken iron anchors recovered (Galili and Rosen this volume b). However it should be noted that 50 fragments of amphora handles with stamps were also recovered in the marina. These are studied in a separate article (see Finkielsztejn this volume) and are not included in the statistics of the current study.

Distribution of the Findings in the Akko Marina As noted, the deepened part of the marina was divided into six excavation areas (Fig 1): Area A was a large area east of the marina; area B was a very small area north of the entrance to the marina; Area C was at the entrance to the marina and Areas D, E and F were inside the marina. It is also noteworthy that 38 barges removed material from the marina (Areas D, E and F) as opposed to 19 barges from Area A outside the marina and 24 barges from Area C at the entrance to the marina. It is reasonable to assume that the large quantity of material removed from Areas D, E and F is partly responsible for the larger number of findings from these areas. An analysis of the distribution of the material in the marina by period shows that more than 70% of the vessels dating from the Hellenistic and Roman periods were found in the inner part of the marina. The relative quantity of vessels from this part declines significantly during the Byzantine period. In contrast, a gradual rise was discerned in the number of findings from Area C over time, and by the Byzantine period findings from Area C constitute nearly half of all findings attributed to that period. This seems to indicate that the center of activity during the Byzantine period had shifted, possibly because of the inner port had silted up. However, it cannot be ruled out that most of the amphorae from the Byzantine period at the entrance to the marina came from one or more shipwrecks in this area. The lack of Crusader ceramics in the western basin and the small quantity of findings from that period found outside the entrance to the marina indicate that the area of the modern marina may have not been in use at this time. It is possible that at this time ships were anchoring in the eastern basin (which was not excavated in the course of the recent project) or in the open sea during storms, when anchoring in the eastern basin was dangerous (See Galili and Rosen this volume e).

Make-up of the Assemblages of Vessels An examination of the assemblages from the various periods (Table 6) shows that amphorae constituted a large part of all the findings in all periods. The imported nature of the assemblage can be seen by the numerous amphorae that served as containers for imported merchandise, and the bowls that were the preferred vessel of import in all periods. However,

56

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Number of tools

Discussion and Conclusions The ceramic assemblage from the Akko Marina sheds light on the interrelationships of Akko with other urban centers in the Mediterranean and on the activities there over a long period. These activities, mainly consisting of marine commerce, obviously made use of various imported vessels, among them amphorae and jars that served as containers for merchandize, especially food products. These were quality vessels that were traded on their own, used on a daily basis on ships or belonged to sailors (whether brought from their home port or purchased during their voyages). The findings surveyed above show that the main activity in the port began in the third century BCE. Findings earlier than this time are meager, consisting of a single jug fragment that may perhaps be attributed to the Iron Age, a single rim fragment that may belong to a hole-mouth jar from the Persian period, and the base of a jar that apparently belongs to a Persian basket-handle jar. The rarity of jars from the Persian period, otherwise common along the coast, and the lack of imports bearing quality black slip characteristic of imported fourth-century BCE Attic vessels, show an apparent lack of intensive activity in the port area prior to the third century BCE. But for the 1,000 years of the Hellenistic, Roman and Byzantine periods, the port saw significant activity. Thereafter, in the eighth–tenth centuries CE, findings dwindle greatly, reduced to a few Late Roman bowls, a few amphorae, an Egyptian bowl and a lamp or two. It seems that after the Byzantine period, activity in the marina area declined, and the commercial and economic activity was greatly reduced, until it rose again during the Crusader period (Galili et al. this volume b). Amphorae were the main and most common containers for marine transport and international trade in liquid merchandise (wine, oil, fermented fish) or other products (dried fruit, cheese, preserved grain, etc.). These products required sealable, durable and inexpensive packaging that could be produced in commercial quantities. Each group of amphorae typically consisted of vessels of the same size and volume and their form corresponded to storage conditions and the rocking of the ship. They featured a tapered base, made of especially thick material so that the vessels could be held fast and withstand shocks. Most of the amphorae that were discovered in the Akko Marina date to the time span between the early Hellenistic period (late fourth century BCE) and the Byzantine to Early Muslim periods (seventh– eighth centuries CE). Most of the vessels come from well-known trading and production centers in the Mediterranean, the Aegean, the Adriatic and a small number of sites in the Marmara Sea and the Black Sea (Fig. 33). The proportion of local vessels (from the coast of Syria and Palestine) in the assemblage

Fig. 33: Distribution of the amphorae from the Akko Marina by period and origin in the Mediterranean: A –diagram, B – Map is small compared to the imported vessels. In addition to the Hellenistic, Roman and Byzantine amphorae, a few fragments of amphorae dated to the Persian period (fifth century BCE) were found. The discovery of significant quantities of non-local amphorae in the Akko Marina indicates the extensive nature of the imported merchandise to local and regional markets, where they were in demand. The number of amphorae from the Hellenistic period in the assemblage is relatively small compared to the number of amphorae from the Roman and Byzantine periods. However, the amphorae and handles from the Hellenistic period show that there was significant activity in the port beginning in the second century BCE and perhaps even in the third century BCE. Among the amphorae attributed to the Hellenistic period, particularly ousttanding is the large quantity of amphorae and handles of Aegean origin compared to the small number of local vessels, a trend that is seen though much of history (see Finkielsztejn this volume). It is possible that the small number of local vessels stems from the fact that the Akko Port was located in northern Palestine, relatively close to the Mediterranean markets, and that commercial activity there included extensive import of products from the Greco-Roman world. In the third century BCE, many Greeks migrated eastward, and brought with them products and foodstuffs from their homeland, and thus encouraged trade with Greece and its environs (Dan 1990: 95). Akko served local trade and to ship merchandise from Greece and Asia Minor to Ptolemaic Egypt. The Zenon Papyri and the Reiner Papyrus attest to

57

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

(Cornell and Matthews 1982: 168–174). The increase in the number of locally made jars shows growing local trade in agricultural produce (wine, olive oil, sesame oil and fish) during the fifth century CE, reaching throughout regions under Byzantine control, especially the Eastern Mediterranean coast (Riley 1979b: 221). Trade typical of the Levantine coast in the fifth century continued into the latter part of the period of Byzantine control as well and the beginning of Arab rule in Palestine. Only a few fragments of jars were discovered in the marina from the Early Muslim (Fatimid) period; other vessels from this period were also surprisingly meager. Notably, excavations in the city of Akko have also revealed only small quantities of ceramics from the Early Muslim period (E. J. Stern pers. comm.). In the first half of the eighth century CE, the cities along Palestine’s coast maintained their cosmopolitan nature and continued to be exposed to direct Byzantine Christian influence. A number of the coastal cities, including Akko, continued trading and Christian pilgrimage ties persisted with the commercial and cultural centers in the Mediterranean (Sharon 1986: 87). This may be the reason for the small number of jars and other ceramics typical of the Early Muslim period in the Akko Marina (E.J. Stern 2013) and in the city, relative to Byzantine-period types. Another possible reason for the small quantity of Early Muslim material in the assemblage is that the areas of anchorage and activity in this period were in the eastern basin, which was not deepened in the project and thus the excavation did not extend to those areas (Galili this volume b).

the export of merchandise from Syria and Palestine to Egypt and trade that passed through Palestine (Tcherikover 1934: 37–44). The findings from the Akko Marina strongly bear out the historic evidence of the strengthening of economic ties between the East (including Palestine) and the Greek world following the conquest by Alexander the Great. The quantity of Roman amphorae in the Akko Marina is twice that of the Hellenistic period, apparently showing significant growth in trade from the Hellenistic to the Roman periods. The variety of types and places of origin of the Romanperiod amphorae reveal commercial ties with many Mediterranean trade centers. It seems that the Akko Port’s trade ties in the Roman period proliferated throughout the Mediterranean basin as a result of political and economic changes that turned the Mediterranean into a single political and commercial entity under Roman rule. As for amphorae from the Byzantine period, more than three quarters belong to three types. About a third belong to Type LR13 (Pls. 41–45), discovered in the smaller Area C near the entrance to the marina (Fig. 34). It is possible that this was the cargo of a ship that was wrecked at that spot or that a good deal of the cargo was somehow damaged and the fragments of the amphorae were intentionally thrown overboard there. Thus the assemblage might not be representative, quantitatively speaking, of the types of vessels that passed through the port in the Byzantine period. The large quantity of amphorae from the Aegean and of local amphorae (Fig. 33) stem from the division of the Roman Empire and the move of the center of economic activity eastward

Fig. 34: Distribution of amphorae in the Akko Marina excavation areas by periods: Hellenistic, Roman, Byzantine 58

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Dothan M. 1967. Ashdod I: The First Season of Excavation 1962. ‘Atiqot 7. Jerusalem.

Acknowledgments The authors wish to thank the Israel Antiquities Authority for the institutional support, M. Feinberg Vamosh for the English editing, T. Sagiv and C. Amit for the photographs and N. Zeevi, M. Smelansky, H. Rozen -Tahan and C. Hersh for the drawings.

Diez-Fernandez F. 1983. Ceramica Comun Romana

de la Galilea. Madrid.

Dressel H. 1899. Corpus Inscriptionum Latinarum 15.1. Berlin.

References

Fischer M. 1989. Hellenistic Pottery Strata III–V. In Z. Herzog, G. Rapp and O. Negbi (eds.). Excavations at Tel Michal. Tel Aviv.

Adan-Bayewitz D. 1986. The Pottery from Late

Byzantine Building (Stratum 4) and its Implications. Excavations at Caesarea Maritima (1975, 1976, 1979) Final Report. L.I. Levine and E. Netzer (eds.).

Finkielsztejn G. This volume. Amphora Stamps from the Akko Marina.

(Qedem 2). Pp. 90–129. Jerusalem.

Fulford M.G. and Peacock D.P.S. 1984. Excavations at Carthage: the British Mission. Sheffield.

Adan-Bayewitz D. 1993. Common Pottery in Roman Galilee. A Study of Local Trade. Ramat Gan.

Galili E., Rosen B., Zviely D., Silberstein N. and Finkielsztejn G. 2010. The evolution of the Akko harbour and its trade links revealed by recent underwater and coastal archaeological research. Journal of Island and Coastal Archaeology 5: 191–211.

Elgavish J. 1977. Shikmona, the Ceramics from the Roman Period. Haifa. (Hebrew). Bar-Nathan R. 1988. The Pottery of Jericho in the Hasmonean Period and the Time of Herod. (M.A. thesis, Hebrew University of Jerusalem). Jerusalem. (Hebrew).

Galili E., Sharvit J., Dahari U., Bahat-Zilberstein N., Finkielsztejn G., Stern E.J., Kool R., Kahanov, Y. and Rosen B. 2002. Akko Port, Underwater Surveys. Hadashot Arkheologiyot, 114: 13–16, 12*–15* (Hebrew).

Bass G.F. and Van Doorninck, H. Jr. 1982. Yassi Ada 1. College Station, Texas.

Galili E. This volume a. The Akko Marina Archaeological Project – Introduction.

Berlin A.M. 1988. The Hellenistic and Early Roman Common-Ware Pottery from Tel Anafa (Ph.D. diss., University of Michigan). Ann Arbor.

Galili E. and Rosen B. This volume e. The Akko Marina Archaeological Project – Summary.

Berlin A.M. 1993. Italian Cooking Vessels and Cuisines of Tel Anafa. IEJ 43: 35–44.

Galili E. and Rosen B. This volume b. The Anchors.

Broneer O.T. 1977. Terracotta Lamps. Princeton N.J.

Galili E., Rosen B. and Sharvit J. This volume b. The Fatimid, Crusader, Mamluk and Ottoman ceramics from the Akko Marina.

Calderon R. 2000. Roman and Byzantine Pottery. In Y. Hirschfeld (ed.). Ramat Hanadiv Excavations, Final Report of the 1984-1998 Seasons. Pp. 91–165. Jerusalem.

Gunneweg J., Perlman I. and Yellin J. 1983. The

Cornell T. and Matthews J. 1982. Atlas of the Roman World. Oxford.

Provenience, Typology and Chronology of Eastern Terra Sigillata. (Qedem 17). Jerusalem.

Coulson W.D.E. and Wilkie N.C. 1986. Ptolemaic and Roman Kilns in the Western Nile Delta. Annual of the American Schools of Oriental Research 263: 61–75.

Guz-Zilberstein B. 1995. The Typology of Hellenistic Coarse Ware and Selected Loci of the Hellenistic and Roman Periods. In Excavation of Dor Final Report, Vol. IB: Areas A and C: The Finds. E. Stern, J. Berg, A. Gilboa, B. Guz-Zilberstein, A. Raban, R. Rosenthal-Heginbottom and I. Sharon (eds.). (Qedem Reports 2). Pp. 289–483. Jerusalem.

Crowfoot J.W., Crowfoot G.M. and Kenyon K.M. 1957. Samaria Sebaste III: The Objects. London.

Hayes J.W. 1967. North Syria Mortaria. Hesperia 36: 337–347.

Dan Y. 1990. Local and Foreign Trade in Palestine in the Second Temple Period. In B. Kedar, T. Dothan and Z. Safrai (eds.). Commerce in Palestine throughout the Ages. Pp. 91–107. Jerusalem. (Hebrew).

Hayes J.W. 1971. Four Early Roman Groups from

59

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Knossos. Annual of the British School of Athens 66: 249–276.

Oleson J.P., Fitzgerald M.A., Sherwood A.N. and Sidebotham S.E. 1994. The Harbours of Caesarea Maritima. Results of the Caesarea Ancient Harbour Excavation Project 1980–1985, Vol. II: The Finds and the Ship. (BAR International Series 594). Ann Arbor.

Hayes J.W. 1972. Late Roman Pottery. London. Hayes J.W. 1973. Roman Pottery from the Stoa at Corinth. Annual of the British School of Athens 74: 418–470, Pls. 77–93.

Peacock D.P.S. and Williams D.F. 1986. Amphorae and the Roman Economy. London.

Hayes John W. 1983. The Villa Dionysos Excavations, Knossos: The Pottery. The Annual of the British School at Athens 78 (1983): 97–169.

Pucci G. 1985. Terra Sigillata Italica, Enciclopedia dell’Arte Antica: Atlante delle forme ceramiche romane, II: 350–406. Rome.

Hayes J.W. 1985. Sigillata orientali. Enciclopedia dell’Arte Antica: Atlante delle forme ceramiche romane, Vol. II: 1–96. Rome.

Riley J.A. 1975. The Pottery from the First Season of Excavation in the Caesarea Hippodrome BASOR 218: 28–54.

Hayes J.W. 1991. Paphos, Vol. III: The Hellenistic and Roman Pottery. Nicosia.

Riley J.A. 1979a. Excavations at Sidi Khrebish Benghazi (Berenice), Vol. II: The Coarse Pottery from Berenice. Tripoli.

Hayes J.W. 1992. Excavations at Sarachane in Istanbul, Vol. II: The Pottery. Princeton.

Riley J.A. 1979b. Amphoras: The Coarse Pottery from Berenice. Excavation at Sidi Khrebish Benghazi (Berenice), Vol. II (Supplements to Libya Antiqua V). Pp. 112–236.

Howland R.H. 1958. Greek Lamps and their Survivals, the Athenian Agora, Vol. IV. Princeton.

Robinson H.S. 1959. The Athenian Agora, Vol. 5. The Pottery of the Roman Period. Princeton.

Johnson B.L. 1988. The Pottery. In E. Weinberg (ed.).

Excavation in Jalame, Site of a Glass Factory in Late Roman Palestine. Pp. 137–226. Columbia, Mo.

Rosenthal-Heginbottom R. 1995. Imported Hellenistic and Roman Pottery. In E. Stern, J. Berg, A. Gilboa, B. Guz-Zilberstein A. Raban R. RosenthalHeginbottom and I. Sharon (eds.). Excavation of Dor Final Report, Vol. Ib: Areas A and C: The Finds. (Qedem Reports 2). Pp. 138–288. Jerusalem.

Kloner A. 1996. Maresha Guide. Jerusalem. Laumonier A. 1977. Exploration archéologique de Délos 31. La céramique hellénistique à reliefs 1: Ateliers ioniens. Paris.

Rosenthal R. and Sivan R. 1978. Ancient Lamps in the Schloessinger Collection. (Qedem 8). Jerusalem.

Magness J. 1992. Late Roman and Byzantine Pottery, Preliminary Report, 1990, Caesarea Papers – Straton Tower, Herod’s Harbour and Roman and Byzantine Caesarea. Journal of Roman Archaeology Supplement Series 5. R.L. Van (ed.). Pp. 125–153. Ann Arbor.

Rotroff S.I. 1982. Hellenistic Pottery, Athenian and Imported Moldmade Bowls, Vol. XXII: The Athenian Agora. Princeton.

Magness J. 1993. Jerusalem Ceramic Chronology Circa. 200-800 CE. Sheffield.

Sciallano M. and Sibella, P. 1994. Amphores: comment les identifier? Aix en Provence.

Mazar B. 1973. Beth Shearim Report on the Excavations during 1936-1940. New Brunswick, N. J.

Sharon M. 1986. “The Towns of Eretz Israel under Islamic Rule.” Cathedra 40: 83–120. (Hebrew).

Marabini Moevs M.T. 1978. The Roman Thin Wall Pottery from Cosa (1948–1954). Rome.

Silberstein N. 2000. Roman and Byzantine Pottery. In Y. Hirschfeld (ed.). Ramat Hanadiv Excavations – Final Report of the 1984-1998 Seasons, Pp. 91–165. Jerusalem.

Mitsopoulos-Leon V. 1992. Die Basilika am Staatmarketin in Ephesos Kleinfunde 1. Teil Keramikhelenistischer und romischer Zeit. Wein.

Slane-Elam J. M.G., Lasscock D. and Neff H. 1994. Compositional Analysis of Eastern Sigillata A and Related Ware from Tel Anafa (Israel). Journal of Archaeological Science 21: 51–64.

Negev A. 1986. The Late Hellenistic and Early Roman Pottery of Nabatean Oboda. (Qedem 22). Jerusalem.

60

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Spitzer D. C. 1942. Roman relief bowls from Corinth. Hesperia 11: 162–192. Stern E.J. 2013. The Fatimid, Crusader and MamlukEarly Ottoman Ceramic Finds from the Akko Marina: Some Insights into Medieval Maritime Activity. ‘Atiqot 76: 139–168. Stern E. 1978. Excavations at Tel Mevorakh (19731976), Part 1: From the Iron Age to the Roman Period. (Qedem 9). Jerusalem. Tcherikover A. 1934. The Land of Israel in Light of the Zenon Papyri. Tarbiz 5: 44–37. (Hebrew). Thompson H. 1934. Two Centuries of Hellenistic Pottery. Hesperia 3: 311–477. Vessberg O. and Westholm A. 1956. The Swedish Cyprus Expedition, Vol. IV Part 3: The Hellenistic and Roman Periods in Cyprus. Stockholm.

61

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 1: Bowls (IAA)

1

2

4

3 6 5 8 7 9

10

11

12 13

14

62

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Plate 1: Bowls

No. Vessel

No. in Area Excavation

Squares

Material

1

Fish bowl 319

C

81. 82. 87, 93, 99, 104

Brown, levigated, peeling black-brown slip

2

Fish bowl 1249

F

95. 96. 100, 101, 105, 106

Brown, levigated, peeling black slip, brownish-red patches on exterior

3

Fish bowl 61

B

80, 81, 86, 92–94

Orange, levigated, thick black slip

4

Fish bowl 477

C

87, 92, 93, 95, 98

Pink, levigated, thick brown slip

5

Fish bowl 49

E

110, 111, 113

Brown, tiny black inclusions, peeling red slip

6

Fish bowl 209

F

95, 96, 100, 101, 105, 106

Brown, tiny black inclusions, red slip with paint drips

7

Bowl

566

F

95, 96, 100, 101, 105, 106

Black, tiny light inclusions, black-brown peeling slip on interior and over rim

8

Bowl

1496

F

95. 96, 100, 101, 105, 106, 110, 112, 115–117

Brown, sandy, red slip, darker at edges, on interior and over rim from exterior

9

Bowl

476

F

95, 96, 100, 105, 106

Red with tiny white inclusions and gold mica. Uneven, brown-red slip, groove around bottom

10

Bowl

223

F

115–117, 123, 129

Orange, levigated, thick, black-brown slip

11

Bowl

54

F

102, 104, 107, 109, 110, 111

Pink, levigated, lustrous black slip, turning brown in center

12

Bowl

320

C

81, 82, 87, 93, 99, 104

Brown, containing mica, brown-red slip, pair of circumferential grooves and small circumferential groove in center

13

Bowl

374

F

115–117, 123, 129

Light pink, sandy, red-brown slip, circumferential crack in center of bottom

14

Miniature bowl

1281

F

95, 96, 100, 101, 105, 106

Orange, sandy, red-brown slip

63

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 2: Bowls (IAA)

2

1

3

4 8 7 5

6

10

11 9

13 12

15 14

16

64

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Plate 2: Bowls

No.

Vessel

No. in Excavation

Area

Squares

Material

1

Bowl

1201

F

95, 96, 100, 105, 106, 110

Orange, levigated clay, black slip turning orange under rim on exterior

2

Miniature bowl

1946

F

95, 96, 100, 105, 106, 110

Orange, levigated clay, brown slip, partly missing on bottom

3

Bowl

45

E

117, 123, 124, 131, 132

Brown, levigated clay, black-brown slip on interior and over rim from exterior

4

Bowl

71

E

117, 123, 124, 131, 132

Orange, sandy clay, brown slip on interior and over rim on exterior

5

Bowl

454

C

87, 92, 93, 95, 98

Brown, sandy, red slip on interior and over rim on exterior

6

Bowl

246

F

95, 96, 100, 101, 105, 106

Orange, levigated clay, red slip

7

Bowl

342

C

87, 92, 93, 95, 98

Orange clay, brown-red slip on interior and exterior

8

Bowl

365

F

115–117, 123, 129

Orange, levigated clay, orange interior, turning brown over rim on exterior

9

Bowl

312

F

115–117, 123, 129

Brown-gray, sandy clay, black slip, peeling on exterior

10

Bowl

1956

F

95, 96, 100, 105, 106

Black, levigated, black slip, does not cover the band under the rim from exterior Brown-red, sandy clay, remnants of red slip on exterior upper part of wall

11

Bowl

1062

F

95, 96, 100, 105, 106, 110, 112, 115, 116, 117

12

Bowl

1950

F

95, 96, 100, 105, 106

Brown-black clay, brown-gray, peeling slip

13

Bowl

1534

F

107, 108, 110, 111, 113, 114

Yellowish sandy clay, red slip, missing on exterior lower part

14

Miniature bowl

93

E

117, 123, 124, 131, 132

Gray, levigated clay, black-brown peeling slip

15

Skyphos

1212

F

95, 96, 100, 105, 106

Brown levigated clay, thick black slip, missing on bottom

16

Bowl

283

F

95, 96, 100, 105, 106

Black levigated clay, brown peeling slip, unclear incising on interior

65

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 3: Various red-slipped bowls and other vessels (IAA) (Nts = not to scale)

66

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Plate 3: Bowls and Various Red-slipped Vessels

No.

Vessel

No. in Excavation

Area

Squares

Material

1

Bowl

1142

F

95, 96, 100, 101, 105, 106

Light brown, levigated, thick black slip, in center turning dark brown, narrow band without color around base

2

Bowl

1248

F

95, 96, 100, 101, 105, 106

Orange, levigated, black slip, turning brown in center

3

Bowl

1562

F

Brown-pink, levigated, mica, red bands on rim and interior, charred rim

4

Bowl

299

F

107, 108, 110, 111, 113, 114 95, 96, 101, 105, 106

5

Bowl

*85/1





Brown-pink, levigated, remnants of red slip over entire vessel

6

Bowl

2

E

117, 120, 123, 124, 131, 132

Dark brown, levigated, mica

7

Bowl

1524

C

94, 96, 98

Gray, orange surface

8

Bowl (Fig. 2)

1529

F

107, 108, 110, 111, 113, 114

Black, white inclusions, remnants of red slip

9

Miniature bowl

423

F

95, 96, 100, 101, 105, 106

Brown-red, black inclusions, remnants of whitish-yellowish material on interior

10

Miniature bowl

1177

F

95, 96, 100, 101, 105, 106, 110, 113,115–117

Orange-red, remnants of whitish material on interior

11

Cup



Haltern 8, Hayes 16

Hayes 1973:449

10 BCE–30 CE

12

Platter

1640

F

107, 108, 110, 111, 113, 114

Black-brown, mica, thick, lustrous black slip

13

Bowl

1089

F

95, 96, 100, 101, 106, 110, 113, 115–117

Brown, levigated red slip

14

Platter

F1640

F



Black-brown, mica, thick, lustrous, black slip

15

Miniature bowl

1058

F

95, 96, 100, 101, 106, 110, 113, 115–117

Brown, levitated red to brown slip

16

Miniature bowl

898

F

95, 96, 100, 101, 105, 106

Brown, levigated, red slip

17

Bowl (?)

1470

F

95, 96, 100, 101, 106, 110, 113,115– 117

Orange, many white inclusions, remnants of red slip

18

Bowl/ krater

564

C

67, 68, 75, 79, 85, 91, 97, 102, 103

Pink, black inclusions, remnants of red slip

*Found in previous survey, precise location unknown

67

Brown, levigated, pinkish surface

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 4: Eastern Terra Sigillata A vessels (IAA)

68

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Plate 4: Eastern Terra Sigillata Ware A

No.

Vessel

No. in Excavation

Area

Squares

1

Plate

1776

F

107, 108, 110, 111, 113, 114

2

Plate

18

E

117, 120, 123, 124, 131, 132

3

Plate

1048

F

95, 96, 100, 101

4

Plate

427

F

115–117, 123, 129

5

Plate (Fig. 3)

425

F

95, 96, 100, 101, 105, 106

6

Bowl

342

F

115–117, 123, 129

7

Bowl

344

C

87, 92, 93, 95, 98

8

Plate

1252

F

95, 96, 100, 101, 105, 106

9

Plate

346

F

115–117, 123, 129

69

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 5: Eastern Terra Sigillata A vessels (IAA)

70

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Plate 5: Eastern Terra Ware A

No.

Vessel

No. in Excavation

Area

Squares

1

Miniature bowl

1189

F

95, 96, 100, 101, 105, 106

2

Miniature bowl

1197

F

95, 96, 100, 101, 105, 106

3

Miniature bowl

3

DBF

87–89, 93, 94, 109, 112, 114, 117–124

4

Miniature bowl

1970

F

95, 96, 100, 101, 105, 106

5

Bowl

261

F

95, 96, 100, 101, 105, 106

6

Bowl

251

C

69, 81, 82, 87, 93, 99, 104

7

Bowl

87/1

F

95, 96, 100, 101, 105, 106

8

Krater

1039

F

95, 96, 100, 101, 105, 106

9

Krater

86/2*





10

Plate

1506

F

95, 96, 100, 101, 105, 106, 110, 113, 115–117

11

Plate

1531

F

107, 108, 110, 111, 113, 114

12

Bowl

930

F

95, 96, 100, 101, 105, 106

13

Plate

79

F

110, 111, 113, 114

14

Cup

39

F

110, 111, 113, 114

15

Cup

347

F

115–117, 123, 129

*Found in previous survey, precise location unknown

71

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 6: Cups, plates and kraters (IAA)

1

2

3

4

5

6

7 8

9

11

10

12

14

13

72

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Plate 6: Cups, plates and kraters

No.

Vessel

No. in Excavation

Area

Squares

1

Cup

232

F

115–117, 123, 129

2

Cup

250

C

69, 81, 82, 87, 93, 99, 104

3

Cup

38

F

110, 111, 113, 114

4

Cup (Fig. 8)

620

F

110, 113, 115–117

5

Cup (Fig. 9)

340

C

87, 92, 93, 95, 98

F6F

Plate

283

B

84, 87–89, 93, 94, 99

7

Plate

27

F

95, 96, 100, 101, 105, 106

8

Plate

224

F

115–117, 123, 129

9

Plate

1955

F

95, 96, 100, 101, 105, 106

10

Cup

1078

F

95, 96, 100, 101, 105, 106, 110, 113, 115–117

11

Cup (Fig. 10)

489

C

87, 92, 93, 95, 98

12

Krater

1404



95, 96, 100, 101, 105, 106, 110, 113, 115–117

13

Krater

43

F

110, 111, 113, 114

14

Cup

60

E

117, 120, 123, 131, 132

73

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 7: Eastern Terra Sigillata A vessels and jugs (IAA)

1 2

3

4

5 6 7

8

9

10

11

74

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Plate 7: Eastern Terra Sigillata A and Eastern Terra Sigillata A jugs

No.

Vessel

No. in Excavation

Area

Squares

1

Miniature bowl

155

C

95, 97, 100, 102, 107

2

Bowl

247

F

115–117, 123, 129

3

Bowl

1415

F

95, 96, 100, 101, 105, 106, 110, 113, 115–117

4

Bowl

1097

F

95, 96, 100, 101, 105, 106, 110, 113, 115–117

5

Bowl

1589

F

92, 102, 103, 107, 108, 110, 111

6

Krater

418

F

95, 96, 100, 101, 105, 106

7

Decanter

866

F

95, 96, 100, 101, 105, 106

8

Decanter

737

F

107, 108, 110, 111, 113, 114

9

Decanter

1516

F

107, 108, 110, 111, 113, 114

10

Decanter

1548

F

107, 108, 110, 111, 113, 114

11

Table amphora

903

F

95, 96, 100, 101, 105, 106

75

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 8: Eastern Terra Sigillata C and D vessels (IAA)

1

2

3

4

5

76

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Plate 8: Eastern Terra Sigillata D

No.

Vessel

No. in Excavation

Area

Squares

1

Plate

320

F

115–117, 123, 129

2

Krater

199

F

69, 81, 82, 87, 93, 99, 104

3

Bowl

618

F

110, 113, 115–117

4

Bowl

908

F

95, 96, 100, 101, 105, 106

5

Jug

84/1*





*Found in previous survey, precise location unknown

77

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 9: Western Terra Sigillata vessels (IAA)

78

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Plate 9: Western Terra Sigillata Ware

No.

Vessel

No. in Excavation

Area

Squares

1

Cup/miniature bowl

1949

F

95, 96, 100, 101, 105, 106

2

Plate

856

F

95, 96, 100, 101, 105, 106

3

Plate

1399

F

95, 96, 100, 101, 105, 106, 110, 113, 115–117

4

Plate

52

F

110, 111, 113, 114

5

Plate

2011

F

95, 96, 100, 101, 105, 106

6

Plate

52

F

110, 111, 113, 114 – decoration of no. 4

7

Plate

1945

F

95, 96, 100, 101, 105, 106

8

Plate

1397

F

95, 96, 100, 101, 105, 106, 110, 111, 113, 114

9

Plate

1063

F

95, 96, 100, 101, 105, 106, 110, 111, 113, 114

10

Bowl

650

F

105, 100, 110, 111, 113

11

Bowl

896

F

95, 96, 100, 101, 105, 106

12

Cup

810

F

115, 116, 117, 123, 129

13

Cup

917

F

95, 96, 100, 101, 105, 106

14

Cup

877

F

95, 96, 100, 101, 105, 106

15

Plate

985

F

95, 96, 100, 101, 105, 106

16

Cup

1193

F

95, 96, 100, 101, 105, 106

17

Cup

1428

F

95, 96, 100, 101, 105, 106, 110, 111, 113, 114

18

Plate

80

F

110, 111, 113, 114

79

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 10: Stamps on West Slope Ware vessels (IAA)

1

2

4

5

7

6

8

11

9

12

10

13

14

15

80

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Plate 10: Stamps on West Slope Ware vessels

No.

No. in Excavation Area

Stamp

Vessel Shown on Plate

1

218

C

LNONIL



2

1399

F

AE.CVI

9:3

3

37

B

CL.I



4

1397

F



9:7

5

312

C

L.GEL



6

1506

F





7

1193

F



9:15

8

653

F

-iLLI



9

810

F

C.SEN

9:12

10

863

F

FR?



11

311

C

L.GEL



12

1066

C

LVMF



13

2011

F

CNAEIZOH

9:4

14

80

F

I.II

9:17

15

71

F

P.CLO.PR

9:5

81

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 11: Thin-walled vessels in the style of West Slope vessels (IAA)

82

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Plate 11: West Slope vessels, thin-walled vessels

No.

Vessel

No. in Excavation

Area

Squares

Material

1

Krater

1282

F

44, 48

Pink, black-gray slip, white decoration

2

Krater

1028

F

44, 48

Orange, brown-slipped exterior, interior orange

3

Table Amphora

1990

F

66, 67

Pink, brown-reddish slip, pink decoration

4

Miniature bowl

42

B

75, 77, 78

Pink, red slip, pink and white decoration

5

Miniature bowl

40

B

75, 77, 78

Pink-purple, brown-red slip

6

Miniature bowl

419

F

44,48

Yellowish, slip decoration in black and white

7

Krater

693

F

50

Brown-orange

8

Cup

146

C

63

Brown, orange, white inclusions, friable

9

Cup

889

F

63

Pink

10

Cup

1925-4

F

44, 48

Pink

11

Juglet

214

F

69, 70

Brown-pink

12

Cup

940

F

44, 48

Orange-yellowish, remnants of red slip

13

Cup

127

C

Unknown



83

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 12: Mold-made bowls (IAA)

84

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Plate 12: Mold-made bowls

No.

No. in Excavation

Area

Squares

1

1916+1917

F

66, 67

2

86/1





Brown-pink, black-brown slip

3

36

F

52

Orange, brown-red slip

4

88

F

52

Orange, red slip

5

1395

F

48, 54

Orange, brown slip on exterior, red interior

6

1977

F

44, 48

Orange, black band of slip on upper part of wall

7

81

F

52

Orange, black-brown slip

Material Orange, black slip

85

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 13: Mold-made bowls (IAA)

86

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Plate 13: Mold-made bowls

No.

No. in Excavation

Area

Squares

Material

1

12

E

80

Orange, black-brown slip

2

34

F

66, 67

Pink-yellowish, black-brown slip

3

45

F

52

Pink-yellowish, red-brown slip

4

813

F

44, 48

Orange, brown slip

87

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 14: Mold-made bowls (IAA)

88

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Plate 14: Mold-made bowls, varia

No.

No. in Excavation

Area

Squares

Material

1

2001





Orange, brown-red slip

2

96

E

80

Orange, black slip

3

1915

F

66, 67

Pink-yellow, red-black slip

4

1997

F

44, 48

Brown-pink, brown-red slip

5

821

F

68

Yellowish (ETSA), red slip

6

1988

F

66, 67

Orange, black slip

7

35

F

52

Orange, red slip

9

313

C

55

Pink-yellow, red slip

10

634

F

66, 67



89

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 15: Late Roman vessels (IAA)

90

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Plate 15: Late Roman vessels

No.

Vessel

No. in Excavation

Area

Squares

Material

1

Bowl

934

F

44, 48

Pink

2

Bowl







Orange

3

Lid

1394

F

48, 54

Orange

4

Bowl

691

F

50

Orange, charred on exterior

5

Bowl

360

F

66, 67

Orange

6

Flat bowl

937

F

44, 48

Orange, red slip

7

Bowl

1033

F

44, 48

Pink, brown red slip

8

Bowl

145

A

40

Pink

9

Bowl

1929

F

44, 48

Orange, remnants of red slip

10

Flat bowl

542

F

49

Orange

11

Decorated rim

809

F

39, 43

Orange

12

Flat bowl

359

C

5

Orange

13

Flat bowl

149

A

40

Orange, white and red inclusions

14

Bowl

88

B

72, 73, 74

Orange, few inclusions

15

Bowl

398

A

36

Orange, red slip

16

Bowl

618

C

6

Orange

17

Bowl

1193

C

6

Pink, yellowish, red slip

91

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 16: Late Roman vessels (IAA)

92

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Plate 16: Late Roman vessels

No.

Vessel

No. in Excavation

Area

Squares

Material

1

Bowl

615

C

6

Brown-pink, red slip

2

Miniature bowl

608

C

6

Pink-yellowish, red slip

3

Bowl

20

C

70, 71

Brown-pink

4

Bowl

272

B

9

Brown-pink

5

Bowl

135

A

42

Pink, remnants of slip

6

Bowl

1447

C

24

Brown, remnants of brown slip

7

Bowl

1445

C

24



8

Bowl

401

A

36

Pink, red slip

9

Bowl

510

C

4

Pink, red slip

10

Bowl

1228

C





93

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 17: Late Roman vessels (IAA)

94

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Plate 17: Late Roman/Byzantine periods

No.

Vessel

No. in Excavation

Area

Squares

Material

1

Plate

418

C

4

Pink, red-painted decoration

2

Plate

1236

C

6

Pink, red-painted decoration

3

Bowl

47

F

80

Brown-pink, painted black on red

4

Bowl

1586

F

30

Orange

95

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 18: Mortaria/basins (IAA)

1 2

3

5

6

4

7

8

96

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Plate 18: Mortaria/basin

No.

Vessel

No. in Excavation

Area

Squares

Material

1

Mortarium

5

B

75, 77

Black, small white inclusions

2

Mortarium

37

E

80

Yellowish, red inclusions

3

Mortarium

959

F

48, 54

Pink, levigated

4

Mortarium (?)

1852

F

31

Brown-pink, red inclusions

5

Mortarium

1984

F

48, 54

Brown-pink, red inclusions

6

Basin (?)

1478

C

24

Gray, red inclusions

7

Mortarium

177

F

39, 43

Brown-purple, white inclusions

8

Mortarium

500

F

48, 54

Pinkish, red and black inclusions

97

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 19: Cooking casseroles (IAA)

98

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Plate 19: Cooking casseroles

No.

Vessel

No. in Excavation

Area

Squares

Material

1

Cooking casserole

646

F

51

Brown, red inclusions

2

Cooking casserole

396

F

66, 67

Brown-red

3

Cooking casserole

111

F

52

Brown

4

Cooking casserole

1141

F

44, 48

Brown-red, white inclusions

5

Cooking casserole

1057

F

48, 54

Orange, purple core, inclusions

6

Cooking casserole

995

F

44, 48

Orange, purple core, inclusions

7

Cooking casserole

1324

F

66, 67

Orange

8

Cooking casserole

41

A

57

Orange

9

Cooking casserole

1724

F

44

Brown-red

10

Cooking casserole

1110

F

48, 54

Orange, few white inclusions

99

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 20: Cooking casseroles (IAA)

1

2

3

100

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Plate 20: Cooking casseroles

No.

Vessel

No. in Excavation

Area

Squares

Material

1

Cooking casserole

241

A

34

Pink, yellowish, remnants of red slip (?)

2

Cooking casserole

872

F

44, 48

Black

3

Cooking casserole

453

F

39, 43

Orange, brown core

101

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 21: Cooking casseroles (IAA)

1

2

3

4

5 6

102

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Plate 21: Cooking casseroles

No.

Vessel

No. in Excavation

Area

Squares

Material

1

Casserole

1523

F

31

Brown, white inclusions

2

Casserole

428

F

66, 67

Brown

3

Casserole

747

F

46

Orange

4

Casserole

147

C

63

Gray, sandy, large inclusions

5

Casserole

1621

F

34

Orange

6

Miniature casserole

1213

F

44, 48

Brown

103

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 22: Cooking casseroles and pots (IAA)

1

2

3 4

5 6

7 8

9

104

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Plate 22: Cooking casseroles and pots

No.

Vessel

No. in Excavation

Area

Squares

Material

1

Krater

1524

C

29

Brown-gray

2

Cooking pot

694

F

50

Brown, red inclusions

3

Cooking pot

1179

F

48, 54

Brown, many large white inclusions

4

Miniature pot

310

C

71

Gray

5

Pot (?)

1442

F

48, 54

Orange

6

Cooking pot

454

A

11

Brown, many inclusions

7

Cooking pot

991

F

44, 48

Orange

8

Funnel mouth

95

A





9

Casserole

249

F

66, 67

Orange

105

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 23: Frying pans (IAA)

2

1

3

5

4 6

7 8

9 11 10

12

13

106

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Pate 23: Frying pans

No.

Vessel

No. in Excavation

Area

Squares

Material

1

Frying pan

281

F

44, 48

Brown-red

2

Frying pan

822

F

68

Brown-red

3

Frying pan

142

F

48 54

Brown, white inclusions

4

Frying pan

102

F

52

Pink-yellowish, tiny black inclusions

5

Frying pan

544

F

49

Orange, inclusions

6

Frying pan

82

F

80

Brown-gray

7

Frying pan

1180

F

48, 54

Black, inclusions

8

Frying pan

1499

F

48, 54

Brown/orange

9

Frying pan

1547

F

31

Black to brown-red

10

Frying pan

253

F

44, 48

Black, inclusions

11

Frying pan

91

B

72, 73, 74

Brown-red, inclusions

12

Frying pan

1649

F

28

Orange

13

Frying pan

532

F

49

Orange

107

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 24: Cooking pots (IAA)

1

2 3

4 5

6

7

8

9

108

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Plate 24: Cooking pots

No.

Vessel

No. in Excavation

Area

Squares

Material

1

Cooking pot

1863

F

31

Brown

2

Cooking pot

222

F



Brown-red, tiny white inclusions

3

Cooking pot

969

F

48, 54

Orange

4

Cooking pot

1937

F

44, 48

Orange

5

Cooking pot

1081

F

48, 54

Orange

6

Cooking pot

1541

F

31

Orange

7

Cooking pot

1958

F

44, 48

Orange

8

Miniature cooking pot

1120

F

48, 54

Orange

9

Cooking pot

314

F

44, 48

Purple (?)

109

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 25: Jugs (IAA)

110

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Plate 25: Jugs

No.

Vessel

No. in Excavation

Area

Squares

Material

1

Jug

1214

F

44, 48

Pink

2

Decanter

242

C





3

Decanter

62

E

80

Brown-pink

4

Decanter

17

E

80

Brown with remnants of sealant on interior

5

Decanter

1025

F

44, 48

Pink-yellowish

6

Decanter base

2010+370

F

66, 67

Pink, decorated with red bands

7

Jug

482

C

35

Pink, inclusions

8

Jug

429

F

66, 67

Brown-red, inclusions

9

Jug

2002

C

35



10

Jug

315

F

44, 48

Brown-gray

11

Jug

432

F

44, 48



12

Jug

10

B

80



13

Jug



C

69, 70

Orange

111

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 26: Jugs (IAA)

2

3

1

5

6

4 7 8

10 9

11

112

12

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Plate 26: Jugs

No.

Vessel

No. in Excavation

Area

Squares

Material

1

Jug

2000

C

35

Gray (?)

2

Jug

305

F

66, 67

Brown, white inclusions

3

Jug

410

F

44, 48

Orange, white inclusions

4

Jug

558

F

44, 48

Gray

5

Jug

726

F

31

Brown

6

Jug

79

C

70, 71

Brown

7

Jug

1280

F

44, 48

Brown, white inclusions

8

Jug

1633

F

31

Black

9

Jug

234

C

69, 70

Pink, remnants of red slip

10

Jug







Brown-gray, white inclusions

11

Jug

1610

C

21

Brown-red

12

Jug

1835

F

39, 43

Pink, inclusions

113

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 27: Jugs (IAA)

1

2 3

5

6

4

7

8 9

10

12

11

114

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Plate 27: Jugs

No.

Vessel

No. in Excavation

Area

Squares

Material

1

Juglet

1215

F

44, 48

Brown, black slip

2

Juglet

1278

F

44, 48

Orange, black-brown slip

3

Juglet

420

F

66, 67

Brown-gray

4

Juglet

119

F

31

Brown-gray

5

Juglet

97

E

80

Brown-gray, remnants of red slip

6

Juglet

1

E

80

Brown-gray

7

Juglet

196

F

52



8

Juglet

118

F

66, 67

Orange

9

Juglet

127

F

66, 67

Orange

10

Juglet

66

F

44, 48

Gray

11

Juglet

128

F

48, 54

Remnants of red slip, remnants of sealant on rim

12

Juglet

62

F

52



115

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 28: Jugs (IAA)

1

2

3

5 4 6

8

7 9

10

116

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Plate 28: Jugs

No.

Vessel

No. in Excavation

Area

Squares

Material

1

Juglet

195

F

44, 48

Brown, yellowish slip (?)

2

Juglet

58

52

Orange, red slip

3

Jug

380

C

66, 67

Brown-gray

4

Amphoriskos

96

F

75, 77, 78

Brown-gray

5

Amphoriskos

1401

B

31

Pink-yellowish

6

Table amphoriskos

2003

F

44, 48

Orange-red

7

Table amphoriskos

321

C

70, 71



8

Amphoriskos

1143

F

44, 48

Pink

9

Amphoriskos

430

F

66, 67

Orange, sealant on interior

10

Amphoriskos

1913

F

31

Brown-orange

117

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 29: Bottles and flasks (IAA)

118

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Plate 29: Bottles and flasks

No.

Vessel

No. in Excavation

Area

Squares

Material

1

Bottle

126

F

31

Brown-red slip

2

Bottle

38

E

80

Yellowish

3

Bottle

129

E

44, 48



4

Bottle

676

E

48

Brown-pink

5

Bottle

1974

E

44, 48

Brown-pink

6

Bottle

1

E

44, 48

Orange

7

Bottle

1811

F



Orange, brown slip, remnants of sealant

8

Bottle

631

C

6

Pink, remnants of sealant

9

Bottle

1712

C



Light gray

10

Bottle

1711

C

6

Light orange

11

Flask

1752

C



Yellowish-pink, lustrous

12

Flask

1811

F

31

Brown

119

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 30: Lamps (IAA)

120

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Plate 30: Lamps

No.

Vessel

No. in Excavation

Area

Squares

Material

1

Lamp

278

F

44, 48

Pink-yellowish, red slip

2

Lamp

246

C

69

Brown friable

3

Lamp

123

F

44, 48

Pink, red slip

4

Lamp

1978

F

44, 48

Brown-pink, brown-red slip

5

Lamp

623

F

52

Pink, brown slip

6

Lamp

*178-4





Dark material

8

Lamp

1521

F

31

Black, black slip

*Found in Ma‘agan Survey (report 158)

121

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 31: Lamps (IAA)

122

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Plate 31: Lamps

No.

Vessel

No. in Excavation

Area

Squares

Material

1

Lamp

191

F

44, 48

Yellowish, thick red slip

2

Lamp

120

F

44, 48

Brown-yellowish, brown slip

3

Lamp

1992

F

44, 48

Orange, thick orange slip

4

Lamp

1959

F

44, 48

Remnants of red slip

5

Lamp

1985

F

66, 67

Brown

6

Lamp

1995

F

48, 54

Yellowish

7

Lamp

2004

F



Brown-pink

8

Lamp

1991

F

44, 48

Orange

123

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 32: Lamps (IAA)

124

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Plate 32: Lamps

No.

Vessel

No. in Excavation

Area

Squares

Material

1

Lamp

196

A

36

Remnants of red slip

2

Lamp

1423

F

48, 54

Remnants of red paint

3

Lamp

1918

F

66, 67

Orange

4

Lamp

329

C

5

Orange, inclusions

5

Lamp

1751

C

5

Orange

6

Lamp

1724

C

5

Light brown

7

Lamp

331

C

3



125

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 33: Lids and varia (IAA)

126

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Plate 33: Lids and varia

No.

Vessel

No. in Excavation

Area

Squares

Material

1

Cooking pot lid

1131

F

44, 48

Orange

2

Lid/measuring vessel

1127

F

48, 54

Brown inclusions

3

Lid

1301

F

44, 48

Orange-pink

4

Lid

346

C

35

Remnants of sealant

5

Lid

269

F

44, 48

Brown

6

Lid

887

A



Brown-black

7

Lid

239

C

69, 70

Pink-yellowish

8

Lid

32

E

80



9

Lid

338

C

35



10

Lid

78

F

52

Yellowish

11

Stand

1660

C

21



12

Stand

96

B

72, 73, 74



13

Face mask

86

F

52

Brown-orange, red slip

14

Lid

2789

A

31



127

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 34: Graffiti (IAA) (drawings not to scale)

128

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Plate 34: Graffiti

No.

No. in Excavation

Area

Squares

1

1999

F

44, 48

2

260

F

44, 48

3

1393

F

48, 54

4

94

F

52

5

78

F



6

1921

F

44, 48

7*

1399

F



8

177

B

10

9

1270

F

44, 48

10*

52

F



11*

86/2





12*



E

80

13*

1024

F



14

14

F

44, 48

15*

1228

C



16

478

E

35

17

54

E

80

18*

18

F



19*

347

F



20

1922

F

44, 48

21

29

B

75, 77, 78

22*

1970

F



23

1256

F

44, 48

24*

476

F



25*

1397

F



26

994

F

44, 48

27

1309

F

44

*On vessels on other plates

129

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 35: Amphorae from the Persian and Hellenistic periods (IAA)

130

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Plate 35: Amphorae from the Persian and Hellenistic periods

No.

Cat. No.

Area

Description

Origin

Date

Parallels and type

1

1725

F

Brown 7.5yr5/3 Black inclusions

Local

Hellenistic

Guz-Zilberstein 1995, type JR1

2

517

F

Local

Hellenistic

Guz-Zilberstein 1995, type JR3

Local

5th–4th cent. BCE

Stern 1973:110 fig. 149, type 3 H

Italy

2nd–1st cent. BCE

Peacock and Williams 1986 class 3

Knidos

2nd cent. BCE

Rosenthal-Heginbottom 1995: fig. 5.0:10

Italy GrecoItalian

2nd–1st cent. BCE

Peacock and Williams 1986 class 3

Egypt (Nile Delta) Crete

3rd–2nd cent. BCE

Coulson and Wilkie 1986: figs. E11.124:10, E11:69

F

Dark gray 2.5y4/1 White inclusions, resembles the material in Pl. 2:3

Egypt

3rd–2nd to early 1st cent. BCE

Coulson and Wilkie 1986: figs. E11.124:10, E11:69

F

Yellowish red 7/5yr4/4 Levigated with much mica

Sicily

2nd cent. BCE

Romano 1994

3

113

F

4

1778

C

5

2

DBF

6

666

F

7

8

9

91

1906

819

E

Gray 10yr5/1 Gray 10yr5/1 Tiny, dark and light inclusions Light brown 7.5yr6/3 Mica Light brown 7.5yr6/4 Black inclusions Brownish red 5yr5/3 Red and black inclusions Dark red 2.5yr4/6 Quartz inclusions, light gray interior 5yr7/2

131

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 36: Amphorae from the Roman period (IAA)

132

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Plate 36: Amphorae from the Roman period

No.

Cat. No.

Area

Description

Origin

Date

Parallels and type

1

1727

F

Light gray brown 10yr6/2

Italy

1st cent. BCE–1st cent. CE

Peacock and Williams 1986 class 4

2nd–4th cent. CE

Peacock and Williams 1986 class 53; Raban et al. 1993: fig. 136:8 Sciallano and Sibella 1994

2

103

C

Dusky red 2.5yr4/2: Edges: 2.5yr5/6 many Egypt white inclusions and a few black inclusions

3

1905



Brownish red 5yr5/3 Interior: pale yellow 5y8/2

North Africa Egypt

1st cent. BCE–1st cent. CE

Sciallano and Sibella 1994

Egypt

2nd–4th cent. CE

Peacock and Williams 1986 class 53

4

1872

F

Dark grey 7.5yr4/1 Interior: reddishbrown 5yr 4/4 many gray inclusions

5

1905

F



Southeast Italy

2nd–1st cent. BCE

Sciallano and Sibella 1994 (amphora de Brindes), p. 29

6

1827

F

Reddish-brown 5yr5/4 much mica, black inclusions

Aegean (Kos)

2nd–1st cent. BCE

Peacock and Williams 1986 class 10

Aegean

2nd–1st cent. BCE

Guz-Zilberstein 1995: fig. 6.42: 6

7

80

C

Light brownish- gray 10yr6/2 Black and brown inclusions

8

436

F

Gray 5y5/1 Light interior

Northern Italy Brindisi

Mid-1st cent. BCE–mid-1st cent. CE

Cipriano and Carre 1989: 70, 74, fig. 4 tipo III–IV

9

159



Yellowish light gray 10yr6/4

Aegean Late Rhodian

Late 1st cent. BCE–early 2nd cent. CE

Peacock and Williams 1986 class 9

10

157

F

Gray 10yr5/1 Dark core

Southern France

1st cent. CE

11

1205

C

Yellowish red 5yr6/6



1st cent. CE

12

836

F

North Africa

Mid-2nd –4th cent. CE

Peacock and Williams 1986 class 40A

13

1505

F

North Africa

Mid-2nd –4th cent. CE

Peacock and Williams 1986 class 40B

14

1247

F

Yellowish-red 5yr4/6 Brown edges Many gray inclusions

North Africa

1st–4th cent. CE

Peacock and Williams 1986 class 37A



Dark reddish gray 2.5yr4/1-4/2 Quartz, gray inclusions

North Africa

1st–4th cent. CE

Peacock and Williams 1986 class 36

15

1339

Dark red 2.5yr4/6 Quartz, gray inclusions Red 2/5yr5/6 Black and white inclusions

133

Peacock and Williams 1986 class 30 Peacock and Williams 1986 class 13

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 37: Amphorae from the Roman period (IAA)

1

2

4 5 3

7 6 8

9

10 13

11

12

134

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Plate 37: Amphorae from the Roman period

No.

Cat. No.

Area

Description

Origin

Date

Parallels and type

1

1005

F

Red 10r5/6 Mica, black inclusions

North Africa

3rd–4th cent. CE

Peacock, Bejaoni and Belazreg 1989: fig. 16:9

North Africa

3rd–4th cent. CE



2

1824

F

Dark grayishbrown 10yr4/2 Light, levigated interior

3

551

F

Gray 10yr6/, sandy

Crete

1st–mid2nd cent. CE

4

604

F

Light brown 7.5yr6/4

Crete/Greece

1st–3rd cent. CE

5

18

F

Gray 10yr5/1, levigated

Southern France

1st cent. CE

6

1604

F

Crete

1st–3rd cent. CE

7

741

F

Crete or Knidos

Mid-2nd cent. CE

Black Sea coast Cilicia

1st cent. CE

Oleson 1994: fig. 41 A83 class I

Black Sea coast

1st–3rd cent. CE

Hayes 1983, type 14

Black Sea coast

1st–3rd cent. CE

Hayes 1983, type 14

Italy and very similar variants also made in Spain

Late 1st cent. BCE–1st cent. CE

Joncheray 1976: Plate 7, no.75 “Lamboglia 2” Sciallano and Sibella 1994: Diverses Amphores Grecques, Epave de la Tradelière p. 95 Joncheray 1976: Pl. 9, no. 95 Cylindrical amphorae

8

1910

F

9

302

C

10

482

F

11

12

13

1359

889

641

Brownish-red 5yr5/4, Levigated Yellowish-red 5yr5/6 Gold mica Dark gray 7.5yr4/1 Light interior, quartz, many inclusions Brown 7.5yr5/3 Light interior, numerous inclusions Dark gray 10yr4/1 Light interior, inclusions

F

10r5/6 Red Quartz inclusions

C

7.5yr5/1 Gray Core: 7.5yr5/4 Brown

Aegean Sea Greece

1st cent. BCE–2nd cent. CE

F

5yr5/6 Yellowish-red, levigated, light interior

Tunisian North Africa and similar variants made in Spain and France

1st–4th cent. CE

Peacock and Williams 1986 class 39 Peacock and Williams 1986 class 41 Peacock and Williams 1986 class 29 Peacock and Williams 1986 class 41* Empereur, Kritzas et Marangou 199 amph. AC2

*Identification based on Empereur, Marangou and Papadakis 1992: fig. 6, where it is identified with group AC

135

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 38: Amphorae from the Roman period (IAA)

1

2

3

4

5

6

7

8

9

136

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Plate 38: Amphorae from the Roman period

No.

1

Cat. No.

1232

Area

Description

Origin

Date

Parallels and Type

F

Dark gray 7.5yr4/1 Reddish-brown 5yr4/3

Aegean, Black Sea coast, Cilicia

First half of 1st cent. CE

Empereur and Picon 1989: fig. 21

Aegean, Black Sea coast

2nd–3rd cent. CE

Riley 1979 type MR5

2

2000

F

Yellowish red 5yr5/6 Core: gray 10yr5/1 Small inclusions

3

947

F

Brown 7.5yr5/3 Small amount of gray inclusions

Crete

1st–3rd cent. CE

Peacock and Williams 1986 class 41

4

965

F

Brown 7.5yr5/3

Northeast Italy

3rd–4th cent. CE

Peacock and Williams 1986 class 42

Northeast Italy

1st cent. BCE –3rd cent. CE



5

2



Grayish brown 10yr5/2 Few large, white and brown inclusions, very small white inclusions

6

1004

F

Red 10r5/6 Light and dark inclusions

Italy

2nd cent. CE

Panella 1989: fig. 9, Ostia III: 369–370

448

F

Yellowish red 5yr5/6 Small amount of mica, gray inclusions

Central Italy

2nd cent. CE

Panella 1989: fig. 1

8

431

F

Light brown 7.5yr6/4 Large quantity of inclusions

Italy

1st cent. CE

Peacock and Williams 1986 class 7

9

140

C

Reddish brown 5yr5/4 Red inclusions, black gravel

Aegean

2nd–3rd cent. CE

Hayes 1983: fig. 24:67

7

137

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 39: Amphorae from the Roman period (IAA)

138

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Plate 39: Amphorae from the Roman period

No.

Cat. No.

Area

Description

Origin

Date

Parallels and tType

1

1635

F

Yellowish red 5yr5/4 Red inclusions, white gravel

Eastern Aegean

2nd–3rd cent. CE



2

1701

F

Red 2.5yr6/8 Large brown inclusions

Eastern Aegean

2nd–3rd cent. CE

Hayes 1983: fig. 25:78

3

1856

F

Red 2.5yr5/8 Gray inclusions

Eastern Aegean

2nd–3rd cent. CE

Hayes 1983: fig. 25:78

4

114

A

Brown 7.5yr4/3 Gray edges, quartz, tiny inclusions

Northern Palestine

2nd–3rd cent. CE

Getsov, in press (Ḥorvat Zanita)

5

285

F

Yellowish-red 5yr4/6 Levigated

Palestine Late Phoenician

2nd–3rd cent. CE

Peacock and Williams 1986 class 46*

6

189

B

Dark gray 7.5yr4/1 Brown, thick core 7/5yr5/4 Levigated

Palestine Gaza

1st cent. CE

Peacock and Williams 1986 class E

Peacock and Williams 1986 class 17



7

241

C

Pale yellow 2.5y8/4 Tiny inclusions

Southern Spain

Late 1st cent. BCE – early 2nd cent. CE

8

1732

F



Spain



139

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 40: Unidentified amphorae from the Roman period (IAA)

1

4

2

3

5

6

7

8

11

12 9

10

140

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Plate 40: Unidentified amphorae from the Roman period

No.

Cat. No.

Area

Description

Origin

Date

Parallels and type

1

1693

C

Reddish brown 5yr4/1-4/3 Quartz

Palestine

1st–2nd cent. CE

Peacock and Williams 1986 class 46

2

935

A

Brown, orange core, white inclusions

Palestine

1st–2nd cent. CE

Peacock and Williams 1986 class 46

3

1728

F

Pale yellow 5y7/3 Numerous inclusions



1st–2nd cent. CE

Oleson 1994: fig. 6: A39 Hayes 1991: fig. 79:7

C

Remains of light slip

Anemurium/ Cyprus*

2nd–3rd cent. CE

Williams 1989: fig. 54 type A Hayes 1991: Pl.24:1,3 Roman Amph. type III

Anemurium/ Cyprus

2nd–3rd cent. CE

Williams 1989: fig. 54 type A Hayes 1991: Pl. 24: 1, 3 Roman Amph. type III

4

22

5

292

F

Dark gray-brown, 7.5yr4/1-4/3 Levigated, quartz

6

1230

F

Grayish-red 5yr4/2 Orange core, a few light inclusions

North Africa





7

1570

F

Dark gray 2.5Y4/1

Spain

2nd–3rd cent. CE

Riley 1979: fig. 88:298 Bar Nathan and Adato 1986: fig. 1:17

F

Brown-gray, friable, similar to amphorae materials from the Black Sea coast



2nd–3rd cent. CE

Peacock and Williams 1986, imitation of above type.

8

1810

Peacock and Williams 1986: 193–195, class 47 Hollow foot “amphora”

9

1857

F

Rough orange-red, 2.5yr5/8

Aegean Sea area

3rd–4th cent. CE

10

67

--

Very pale brown 10yr7/4 Black and white gravel

Unknown origin

3rd–4th cent. CE

Peacock and Williams 1986: 199–200, class 50

North Africa or Spain



Unidentified





Unidentified

11

1579

F

Red 10r5/6 Brown edges, mica with a few large, light and dark inclusions

12

23

B

Brown10r5/6 Friable, mica

*This type was manufactured in Anemurium and in Cyprus. The material from which the amphorae in the Akko assemblage was made is not certain to be identical to the material from which the amphorae from Cyprus were made, as described by Williams (1989) and/or Hayes (1991).

141

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 41: Amphorae from the Byzantine period (IAA)

1

2

3

5

4

6

7

8

10

9

11

12

13

14

142

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Plate 41: Amphorae from the Byzantine period

No.

Cat. No.

Area

Description

Origin

Date

Parallels and Type

1

729

F

Red 2.5yr5/6 Brown edges, levigated

North Africa Spain

3rd–4th cent. CE

Peacock and Williams 1986 class 33

2

1532

C

Red 2.5yr5/8 Quartz, small inclusions

North Africa

Late 3rd– mid-5th cent. CE

Keay 1984: type XXV Peacock and Williams 1986: class 37B

3

930

A

Red 2.5yr5/8 Numerous gray and white inclusions

North Africa

Late 3rd– mid-4th cent. CE

Peacock and Williams 1986 class 34C

4

349

A

Dark red 2/5yr4/8 Dark gray edges, a few white inclusions

North Africa Balkans

3rd–5th cent. CE

Keay 1984: type IV

5

1450

C

Red 2.5yr5/6 Brown edges, levigated

North Africa

5th–6th cent. CE



6

1195

C

Dark gray 2.5y4/1 Levigated

Algiers

3rd–4th cent. CE

Peacock and Williams 1986 class 38



3rd–4th cent. CE

Peacock and Williams 1986 class 38 Sciallano and Sibella 1994: Dressel 30

7

1

B

Reddish brown 5yr4/4 Sandy, light inclusions

8

1137

C

Light gray 5y7/1 Exterior, pale 5Y8/2 Interior, light pink

Palestine

5th–6th cent. CE

Peacock and Williams 1986 class 46

9

354

C

Yellowish red 5yr5/6 Red inclusions

Palestine

5th–6th cent. CE

Peacock and Williams 1986 class 46

10

2/1

Dark gray 5yr4/1 Levigated, quartz

Palestine coast, RomanLebanon Byzantine

11

843

F

Very dark grayishbrown 10yr3/2 Quartz, black inclusions

Lebanon

5th–6th cent. CE

Riley 1979: LR II

12

831

F

Grayish-brown

Lebanon

5th–6th cent. CE

Riley 1979: LR II

B

5yr5/6 Yellowish-red Inclusions of quartz or mica

North Africa Tunisia

Late 4th cent. CE.

Robinson 1959: 109 Reynolds 1995: fig. 89

F

5y8/4 Pale yellow Dark, quartz, inclusions appearing outside the vessel, friable and airy

Eastern Mediterranean

Early 5th– Peacock and Williams mid-7th 1986 class 44b cent. CE

13

14

294

1766

143

Safrai, On Jars and Wine, Kibbutz Sa’ar Museum

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 42: Amphorae from the Late Roman, Byzantine and Post-Byzantine period (IAA)

1

2

4

3

6

5

7

8

9A 9B

10

11

144

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Plate 42: Amphorae from the Late Roman, Byzantine and Post-Byzantine period

No.

Cat. No.

Area

Description

Origin

1

167

F

Very pale brown 10yr8/3, dark inclusions

Iberian Peninsula 3rd–5th cent. Algerian North CE Africa Late Roman

Peacock and Williams 1986 class 23

2

1848

F

Yellowish-red 5yr5/6 Mica, light inclusions

Iberian Peninsula 3rd–5th cent. Algerian North CE Africa Late Roman

Peacock and Williams 1986 class 23

3

262

B

Dark gray 10yr4/1 Quartz

Iberian Peninsula 3rd–mid-5th Algerian North cent. CE Africa Late Roman

Peacock and Williams 1986 class 23

4

1227

F

Dark gray 5yr4/1 Mica, black inclusions

3rd–5th cent. Iberian Peninsula CE Byzantine

Peacock and Williams 1986 class 23

5

1708

C

Red 2.5yr6/8 Interior, pink 5yr7/4 Quartz

North Africa Spain

6th–7th cent. CE Byzantine

Peacock and Williams 1986 class 51

6

357

C

Gray 2.5y5/1 Core: pinkish- gray 7.5yr6/2

North Africa

6th–7th cent. CE Byzantine

Peacock and Williams 1986 class 51

7

817

F

Brown 10yr5/3 Quartz, a few white and gray inclusions



Early 5th cent. CE

Hayes 1992: fig. 22: 2 type 2

Brown 7.5yr4/2 Friable, white inclusions, mica, gold on exterior of vessel



6th cent. CE

Fulfold and Peacock 1984: fig. 38: 47

Core: brown 5yr4/4: Quartz, light inclusions

Palestine

6th–9th cent. CE

Peacock and Williams 1986 class 46

B

Dark reddish- brown 5yr3/3 Quartz, numerous dark and light inclusions

Cyprus Italy

6th–7th cent. CE

Oleson 1994: fig. A36 class K

F

Dark reddish- brown 7.5yr3/3 Numerous white inclusions



Early 6th cent. CE

Bass 1982: fig. 8 20: p. 80

8

1880

F

9a

1125

C

9b

1022

C

10

11

32

730

145

Date

Parallels and type

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 43: Amphorae from the Byzantine and post-Byzantine periods (IAA)

146

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Plate 43: Amphorae from the Byzantine and post-Byzantine periods

No.

Cat. No.

Area

Description

Origin

Date

Parallels and type

Pale yellow, pink 2.5y8/3 Pink 7.5yr7/4 Sandy

Istanbul

8th cent. CE

Riley 1979 type 35

Istanbul area

10th cent. CE

Hayes 1992: fig. 24: 2–4 type 54 Riley 1979: LR13

1

1387

C

2

1141

C

3

1352

C

Yellowish- red 5yr5/6 Large white and black inclusions

Black Sea

6th–7th cent. CE

Hayes 1992: fig. 23: 2, 3, type 29 Bass 1982 type 2

Black Sea

6th–7th cent. CE



Reddish-brown 5yr4/4

4

798

F

2.5tr4/6 Dark reddish- gray, numerous white inclusions

5

648

F

7.5yr6/3 Light brown

Northern Syria

5th–mid7th cent. CE

Peacock and Williams 1986 class 44

--

Antioch, Northern Syria Cyprus

5th–7th cent. CE

Peacock and Williams 1986: Class 44 Calderon 2000: 132–136

North Africa Tunisia (?)

5th cent. CE 6th cent. CE

Peacock and Williams 1986: Class 35, type LX

North Africa Tunisia

5th cent. CE 6th cent. CE

Peacock and Williams 1986: Class 35, type LX

6

7

8

1008

724

1202

F

C

2.5yr4/6 Dark reddish- gray, light interior (quartz?) 7.5yr5/2 Brown, orange interior

147

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 44: Amphorae from the Late Roman and Byzantine periods (IAA)

1

2

4

3

5

6

7

148

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Plate 44: Amphorae from the Late Roman and Byzantine periods

No.

Cat. No.

Area

Description

Origin

Date

Parallels and type

Central Italy Ostia



Kapitan 1972: fig. 8, 249 Dressel 29

1

1808

F

5yr5/6 Yellowish- red, small quantity of mica, black inclusions

2

1188

F

7.5yr5/1 Gray, levigated, mica

Central Italy Ostia



Kapitan 1972: fig. 8, 249 Dressel 29

C

2/5yr4/1 Dark reddish-gray, light inclusions, quartz

Southern France

1st–3rd cent. CE

Peacock and Williams 1986 class 27

F

2.5yr5/6 Red, black inclusions

North Africa (?)

4th cent. CE

Sciallano and Sibella 1994: Africaines Cylindriques, Epave B de la pointe

Aegean Sea (Argolid region of the Peloponnese) Black Sea North Africa

4th–late 6th cent. CE

Peacock and Williams 1986 class 43, (Chartage LR2, Benghazi LR2) Keay 1984 type LXV

3

4

1751

1597

5

252

C

7.5yr8/4 Pink Interior 10YR,8/4 Very pale brown

6

650

A

2.5yr4/6 Red, many gray inclusions

North Africa

5th cent. CE

Keay 1984 type LV

F

10yr5/4 Yellowish-brown, mica and white inclusions

North Africa

5th–6th cent. CE

Keay 1984 type XXXV: B

7

1888

149

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 45: Unidentified amphorae from the Byzantine period (IAA)

150

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Plate 45: Unidentified amphorae from the Byzantine period

No.

Cat. No.

Area

Description

1

92

B

5yr5/6 Yellowish-red, quartz or mica inclusions

2

601

F

— 2.5y4/1 Dark reddish-gray, levigated, dark gray inclusions Reddish-gray Dark gray 7.5yr

3

1572

C

4 5

1545 1603

C F

6

765

F

7.5yr5/4 Strong brown

7

1685

F

7.5yr 4/3-4/4 Brown, white inclusions

8

1245

C

2.5yr 4/6 Red

9

1357

C

2.5yr 5/6 Red

10

1159

C

5yr 5/4 Reddish-brown

151

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 46: Amphorae bases (IAA)

3 1

4

2

5

7

8

6

12 9

10 11

13

15

14

152

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Plate 46: Amphorae bases

No.

Cat. No.

Area

1

6

F

Description

Origin

Date

Parallels and type

Kos

2nd–early 1st cent. BCE

Zemer 1977: Pl. XII, no. 35

Western Italy

3rd–early 2nd Sciallano and Sibella cent. BCE 1994: 31

2

151

F

5yr5/8 Yellowishred, black inclusions and mica

3

80

F



Cyprus (Kourion)

3rd cent. BCE

Zemer 1977: Plate XI, no. 32

F

7.5yr6/4 Light brown, black inclusions

Rhodes Aegean Sea

4th–3rd cent. BCE

Sciallano and Sibella 1994: 88 Smaller variant of a Rhodian jug

Greece

6th–5th cent. BCE

Sciallano and Sibella 1994: 26, Amphorae IonioMassaliete Zemer 1977: Plate VIII, no. 23

4

64

5

466

F

2.5yr5/6 Red

6

40



7.5yr6/2 Pinkish-gray

Greece Kos

7

293

F

7.5yr4/2 Brown

Knidos

8

581

F

2.5yr5/6 Red

Italy





Italy (Campagnia, Etruria, Latium), possibly also Spain

Peacock and Williams 1986: Late 2nd– Class 4, fig. 28, pp. 89–90; early 1st cent. Sciallano and Sibella BCE 1994: Type Dressel 16, p. 41

Southern Spain and apparently also Catalonia

Late 1st cent. BCE

Peacock and Williams 1986: Class 26A, fig. 50, pp. 117– 118; Keay and Jones 1982: fig. 6.1, 4

1st cent. BCE–2nd cent. CE

Sciallano and Sibella 1994: 59, Dressel 38/Beltran II A

9

527

First half of 4th cent. BCE 2nd–1st cent. BCE 2nd–first half of 1st cent. BCE

Zemer 1977: Pl. X, no. 30 Zilberstein 2000: 442, fig. 8 Peacock and Williams 1986: Class 3, fig. 26, p. 86

10

496

F

2.5yr5/6 Red

11

1324

F

2.5yr5/8 Red, black inclusions

Southern Spain

12

849

F

7.5yr5/1 Gray

Kos

13

2007





Western Italy

14

1718

F



Southern Spain

1st cent. BCE–1st cent. CE

Sciallano and Sibella 1994: 55, Dressel 8,

15

239





Southern Spain

Early 2nd cent. CE

Peacock and Williams 1986 class 18, fig. 54, pp.122-123

153

1st cent. BCE–1st cent. CE Early 2nd– mid-1st cent. BCE

Sciallano and Sibella 1994: Dressel 5 Sciallano and Sibella 1994: 90, Dressel 1A, p. 32

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 47: Amphorae bases (IAA)

3 2

1

6

5

4

7 8

9

154

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Plate 47: Amphorae bases

No.

Cat. No.

Area

Description

Origin

Date

Parallels and type

North Africa (Tunisia)

Late 2nd– late 4th cent. CE

Peacock and Williams 1986 class 34C, fig. 81: 156 Panella 1973 type Africana IIC

1

157

C

5yr5/6 Yellowish-red, black inclusions

2

811

F

2.5yr5/8 Red, dark inclusions, quartz

North Africa (Tunisia)

Late 4th– 8th cent. CE

Sciallano and Sibella 1994. Type Africaines Cylindriques, p. 82

3

661

F

2.5yr5/1 Dark reddish-gray

North Africa

Late 3rd– early 5th cent. CE

Sciallano and Sibella 1994 type Almagro 50, p. 66

4

444

F

2.5yr5/1 Dark reddish- gray, light red inclusions

Eastern Mediterranean

1st–4th cent. CE

Williams 1989: fig. 555, 556, pp. 91–93

5

391

C

10yr7/4 Very pale brown

Eastern Mediterranean

1st–4th cent. CE

Williams 1989: fig. 555, 556, pp. 91–93

6

572

A

10r5/4 Yellowish- brown, mica

North Africa

5th–6th cent. CE

Fulford and Peacock 1984: fig. 43

7

1341

F

2.5y4/1 Dark gray

North Africa

Late 5th cent. CE

Keay 1984: fig. 32, nos. 2-3 Type LXI.A

8

952

F

5yr5/4 Reddish-brown

North Africa

4th–mid5th cent. CE

Keay 1984: fig. 29, nos. 3-4 Type XXXVI.B

4th–5th cent. CE

Peacock and Williams 1986: Class 49, fig. 116, pp. 198–199 Sciallano and Sibella 1994 type Late Roman 4, p. 63 Keay 1984: fig. 28, no. 3 Type LIV

9

1620

C

5yr5/6-10yr5/3 Yellowish- red-brown, shell fragments

Israel (Gaza) Egypt

155

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 48: Amphorae bases (IAA)

1

3 2 4

5

7 6

8

11 10

9

13

15

12

14

16

17 18

19

156

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Plate 48: Amphorae bases No.

Cat. No.

Area

Description

Origin

Date

Parallels and type

1

56



2.5yr5/8 Red

Portugal

Late 3rd–5th cent. CE

2

52

F

10yr7/4 Very pale brown

Spain Portugal

3rd mid5th cent. CE

3

1235

F



Southern Spain

4

1352

F

2.5yr5/1 Dark reddish-gray

Israel (Gaza) Egypt

1st cent. BCE 1st cent. CE

Sciallano and Sibella 1994 type ALMAGRO 50 – Port Vendres 1, p. 68 Peacock and Williams 1986 class 23, fig. 62, pp. 132–133; Sciallano and Sibella 1994 type ALMAGRO 51/C- Planier 7 (Marseille)

2.5yr5/1 Gray

Antioch Northern Syria

7th–8th cent. CE

10yr7/4 Very pale brown, levigated 10yr6/4 Light yellowish- brown 2.5yr6/6 Red, red inclusions 5yr5/6 Yellowish-red 7.5yr4/1 Dark gray, quartz, black inclusions 10r8/2 Creamy white core and interior wall 10r5/8

Cilicia Anemurium Cilicia Anemurium Unknown Common in the western Mediterranean

1st–4th cent. CE 1st–4th cent. CE

Peacock and Williams 1986: Class 44, fig. 104B, pp.185–187; Calderon 2000 type 1, pp. 132– 133 Williams 1989: fig. 55, nos. 554, 558 Williams 1989: fig. 55, nos. 554, 558

3rd–4th cent. CE

Peacock and Williams 1986 class 42, fig. 99, pp. 180–181

4th–7th cent. CE 4th cent. BCE 1st cent. CE

Fulford and Peacock 1984: fig. 36, no. 20



4th–5th cent. CE

Williams 1987 type 2, p. 235

Aryana, Tunisian coast

6th–7th cent. CE

Riley1979: 226–227 Panella 1974: fig. 32, pp. 609– 630

4th–6th cent. CE

Sciallano and Sibella 1994: Late Roman 3, p. 88

1st–6th cent. CE

Peacock and Williams 1986 class 45, fig. 107, pp. 187–189

1st–6th cent. CE

Peacock and Williams 1986 class 45, fig. 107, pp. 187–189

1st–6th cent. CE 1st–6th cent. CE 4th–7th cent. CE 3rd–4th cent. CE

Peacock and Williams 1986 class 45, fig. 107, pp. 187–189 Peacock and Williams 1986 class 45, fig. 107, pp. 187–189 Sciallano and Sibella 1994: Late Roman 7, p. 95 Peacock and Williams 1986 class 47, pp. 193–195,

5

1821

F

6

52

B

7

395

A

8

710

F

9

1955

F

10

147

F

11

212

B

12

245

A

5yr5/4 Reddishbrown, quartz

13

293

A



14

783





15

1819

F

2.5y4/1 Dark gray

16

444

F

17

258

F

18

592



19

1843



— 2.5y5/1 Gray, quartz, brown inclusions — Orange-red 2.5yr5/8

4th–6th cent. CE

— Balearic Islands

Asia Minor Greece and the Aegean Black Sea Asia Minor Greece and the Aegean Black Sea Asia Minor Greece and the Aegean Black Sea Western Asia Minor Sardis Aegean Western Asia Minor Sardis Aegean Egypt Eastern Aegean

157

Sciallano and Sibella 1994 type Dressel 9 Sciallano and Sibella 1994 type Late Roman 4 , p. 63

Sciallano and Sibella 1994: Roman, no. 18

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Plate 49: Unidentified bases (IAA)

158

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Plate 49: Unidentified bases

No.

Cat. No.

Area

Description

Origin

Date

1

671

F

7.5yr5/1 Gray, mica, white inclusions

North Africa (Tunisia)



North Africa

5th cent. CE





2

1/12

F

10yr3/2 Very dark grayishbrown, sandy

3

993

F

10yr3/1-5/2 Very dark graygrayish brown

159

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Table 1: Parallels and dating of Western Terra Sigillata vessels

Vessel

Pl.

Type

Parallels

Date

Cup/small bowl

Pl. 9: 1

Haltern 7, Pucci XX

Pucci 1985: 388–389

Last quarter of 1st cent. BCE– early 1st cent. CE

Plate

Pl. 9: 2

Pucci IV, Haltern 4, Hayes 6

Pucci 1985: 381– 382 Hayes 1973: 439–440

Last quarter of 1st cent. BCE– second half of 1st cent. CE.

Plate

Pl. 9: 3

Pucci IX, Haltern 3, Hayes 11

Pucci 1985: 383, Hayes 1973: 449

First half of 1st cent. CE

Plate

Pl. 9: 4-5

Pucci X, Haltern 5

Pucci 1985: 385

Last third of 1st cent. BCE–early 2nd cent. CE.1

Plate

Pl. 9: 7-9

Pucci X, Haltern 2, Hayes 4

Hayes 1973: 449

10 BCE–30 CE

Bowl

Pl. 9: 10,11,15



Pucci 1985: 387

Second quarter of 1st cent. CE (?)

Bowl

Pl. 9: 18

Pucci XIX, Hayes 13

Pucci 1985: 388; Hayes 1973:439– 441

Second half of 1st cent.–early 2nd cent. CE

Cup

Pl. 9: 12

Hayes 232



Like nos. 4–5

Cup

Pl. 9: 13

Pucci XXXVII, Haltern 12, Hayes 21

Pucci 1985:396– 397

Early–late 1st cent. CE–early 2nd cent. CE

Cup

Pl. 9: 14

Haltern 8, Hayes 16

Hayes 1973: 449

10 BCE–30 CE

Cup

Pl. 9: 16

Pucci XLI, Haltern 16, Camolodunum 18

Pucci 1985:397– 398

Second quarter of 1st cent. CE

Cup

Pl. 9: 17

Pucci XXIX, Haltern 15

Pucci 1985:392– 393

1st cent.–mid. 2nd cent. CE

1 According to the distribution of this type at Pompeii, it was apparently popular during the time of Claudius at the earliest

(40–53 CE) (Hayes 1973: 440)

2 This type appears together with the plate (1: 4–5) and is of similar date.

160

Hellenistic, Roman and Byzantine Ceramics from the Akko Marina

Table 2: Summary of bowl finds

Period

Total

Certainly imported

Area A

Area B

Area C

Area E

Area F

Hellenistic (3rd–1st cent. BCE)

225

158 (70)

4 (1.7)

9 (4)

16 (7.1)

11 (4.9)

145 (64.4)

Roman (late 1st cent. BCE– early 4th cent. CE

107

105



3 (2.8)

13 (12.1)

3 (2.8)

88 (82.2)

Byzantine (4th–7th cents. CE)

92

92

18 (19.6

5 (5.4)

35 (38)



29 (31.5)

Table 3: Summary of cooking-vessel finds

Period

Total Imported

Area A

Area B

Area C

Area E

Area F

Hellenistic (3rd –1st cent. BCE)

37

17 (45.9)





2 (5.4)

1 (2.7)

34 (91.9)

Roman (late 1st cent. BCE– early 4th cent. CE

71

31 (43.6)

4 (5.6)

3 (4.2)

5 (7)

1 (1.4)

53 (74.6)

Byzantine (4th–7th cents. CE)

46

5 (10.9)

4 (8.7)

3 (6.5)

21 (45.7)

1 (2.2)

14 (30)

Table 4: Summary of finds of jugs and juglets

Period

Total

Imported

Area A

Area B

Area C

Area E

Area F

Hellenistic (3rd –1st cent. BCE)

53

29 (54.7)

4 (7.5)

1 (1.8)

5 (9.4)

5 (9.4)

36 (67.9)

Roman (late 1st cent. BCE–early 4th cent. CE

81

37 (45.7)

6 (7.4)

5 (6.2)

17 (21)

7 (8.6)

61 (75.3)

Byzantine (4th–7th cents. CE)

10













1 Because examinations were not undertaken to determine the source of the vessels, we cannot be certain in all cases whether

they were imported. The vessels that were certainly imported include types whose origin is known and vessels whose clay can be identified as a type that does not exist locally, for example, clay containing mica.

161

Na‘ama Silberstein, Ehud Galili and Jacob Sharvit

Table 5: Summary of amphorae finds

Period

Total

Local

Imported

Area A

Area B

Area C

Area E

Area F

Hellenistic (3rd –1st cent. BCE)

86

7

79 (91.9)

1 (1.2)

2 (2.3)

5 (5.8)

13 (15.1)

54 (62.8)

176

21

144 (81.8)

8 (4.5)

6 (3.4)

30 (17)



108 (61.4)

235

68

104 (44.3)

23 (9.8)

18 (7.7)

109 (46.4)

1 (0.4)

56 (23.8)

Roman (late 1st cent. BCE–early 4th cent. CE Byzantine (4th–7th cents. CE)

Table 6: Summary of ceramic finds in the Hellenistic, Roman and Byzantine periods

Bowls

Cooking Vessels

Jugs and Juglets

Amphorae

Total

Area A

Area B

Area C

Area E

Area F

Hellenistic

225 (56.11)

37 (9.23)

53 (13.21)

86 (21.44)

401

9 (2.24)

12 (2.99)

28 (6.98)

30 (7.48)

269 (67.08)

Roman

107 (24.54)

71 (16.28)

81 (18.58)

177 (40.6)

436

18 (4.13)

17 (3.9)

65 (14.9)

11 (2.52)

310 (71.1)

Byzantine

92 (24.02)

46 (12.01)

10 (2.61)

235 (61.36)

383

40 (10.44)

27 (7.05)

170 (44.39)

2 (0.52)

100 (26.11)

162

Chapter 4

Amphora Stamps from the Akko Marina Gerald Finkielsztejn Israel Antiquities Authority Key words: marine archaeology, underwater archaeology, Hellenistic pottery, Amphora stamps, Acre

Fifty fragments of amphoras bearing stamps were uncovered during the underwater excavation in the Akko Marina project (Galili this volume a), together with many other unstamped ones. Although both groups belong to the same category of finds, the stamps are presented separately because of the significance of the information they provide, especially concerning the chronology. The collection comprises 47 Hellenistic Rhodian stamped fragments (including three necks with both handles stamped), one Hellenistic Knidian (neck with one handle stamped), one Byzantine and one medieval. As is generally the case in the surface excavations in Israel for the Hellenistic period, the Rhodian production of the Hellenistic period is by far the best represented.

Catalogue Rhodes Many works have been published on the topic of the Rhodian stamps and their contributions cannot be detailed here. In the following catalogue, the commentary concerning this class of amphora stamps is minimal, mainly to justify the dates. Some unusual dies are more detailed. The chronology of the stamps is according to this writer’s revision, in which the dates of the eponyms in office before the middle of the second century BCE according to V. Grace are lowered by about 12 years (Periods I to IV; see Grace 1985: 42 and Empereur 1990: 207). This suggested chronology – referred to as ‘the lower chronology’ in the catalogue – is detailed in the publication of the relevant chapters of the writer’s Ph.D. dissertation and further research published in Finkielsztejn 2001a (including updates of a less detailed list of eponym dates that was published in preliminary publications: Finkielsztejn 1993; 1995a; 1995b, with demonstrations of the validity of the lower

chronology). Table 1 below displays the comparison between both “higher” and “lower” chronologies. The ‘lower chronology’, although not fully refined (as explained in the book), is now accepted in its general frame by specialists and some adjustments are suggested that contribute to its improvement. The dates provided in the catalogue are according to Finkielsztejn 2001a, with some mild adjustments. In some cases a period will be provided instead of the date as a year that appears in the book, since the latter often seem prematurely too precise, because the cycle of the use of the ‘intercalary month’ (Finkielsztejn 2001a: 178) was not established in the book. Partial but significant improvements on historical and institutional questions (some of which are listed in Finkielsztejn 2001a: 44–46) were provided by Nathan Badoud’s Ph.D. dissertation, which was recently published (Badoud 2015). However, for the dating of the present contexts the general frame of the 1 lower chronology is accurate enough. The finds from the Akko Marina provide us with additional associations of eponyms with fabricants on necks with their two stamped handles still attached. A comparison of the finds from the marina and those from excavations inland will be presented in the large catalogue of the finds from excavations at the Hospitaller Citadel and at the Law Court location. Not much can be said on the significance of the finds from such a context as the random dredging of a port. The stamped finds should in any case be compared with the unstamped ones and the rest of the pottery (Silberstein, et al. this volume) and the coins. The results are presented on two graphs. The first one (Fig. 1) compares the chronological breakdown of the Rhodian eponym stamps from the marina dredging with finds from the main excavated areas in the city of Akko, i.e., the Old City (Hospitaller Citadel), close to the marina, and the “lower city,” i.e., extending from the Old City in the west eastward

1 The present chapter was completed in 1995. The dates of the Rhodian stamps were subsequently updated on the basis of Finkielsztejn 2001a. The reader is invited to check the literature relevant to the chronology that appeared since 2001, primarily the publications by the author and Nathan Badoud.

163

Gerald Finkielsztejn

Table 1. Higher and lower Rhodian eponym chronologies.

Period

See Finkielsztejn 2001a:196-197, Tables 22.1 and 22.2 Grace’s chronology (1985) with Empereur’s refinements (1990)

Finkielsztejn’s Lower Chronology (2001a)

Ia

c. 300-c. 280

c. 304-c. 271

Ib Ic IIa IIb IIc IIIa IIIb IIIc IIId IIIe IVa IVb Va Vb Vc VI VIIa VIIb

c. 279-c. 270 c. 269-c. 240 c. 239-c. 225 c. 224-c. 206 c. 205-c. 202 c. 201-c. 194 c. 193-c. 188 c. 187-c. 182 c. 181-c. 175 c. 174-c. 156 c. 155-c. 146 c. 145-c. 108

c. 85-Augustus

c. 270-c. 247 c. 246-c. 235 c. 234-c. 220 c. 219-c. 210 c. 209-c. 199 c. 198-c. 190 c. 189-c. 182 c. 181-c. 176/174 c. 175/173-c. 169/167 c. 168/166-c. 161 c. 160-c. 153 c. 152-c. 146 c. 145-c. 133 c. 132-c. 121 c. 120-c. 108 c. 107-c. 88/86 c. 85-c. 40 c. 40-Augustus

Fig. 1: Breakdown of Rhodian eponym stamps from the Akko Marina, Akko’s Old City and lower city, compared with the Rhodian production

Chronological Tables in Finkielsztejn 2001a (with pages) P. 55 (partial list; see also Lawall 2007:31-42) P. 188, Table 17

P. 191, Table 18

P. 192, Table 19

P. 193, Table 201 P. 195, Table 21

P. 161-162, Tables 13-14 P. 162, Tables 14-15; Finkielsztejn 2000.

Fig. 2: Breakdown of the Rhodian amphora stamps from the Akko Marina including the eponym stamps and the fabricant stamps 164

Amphora Stamps from the Akko Marina

to the ancient mound. The second one (Fig. 2) presents the breakdown of the eponym and fabricant stamps, separately and together. This illustrates a methodological point, i.e., the fact that only the eponym stamps, much better dated than the fabricant ones (since the latter were producing amphoras during several eponym years) are taken into account for the analysis of periods of imports, which is supposed to reflect the periods of activity in the harbor. The graph shows that the impact on the general profile of the breakdown (total or average number of the stamps) is minimal, taking into account that the five-year span chosen in the career of a given fabricant takes into account those represented by eponym stamps. The fact that the impact of the production in Rhodes is minimal is illustrated by the profile of the stamps found on the island itself on Fig. 1 (see Finkielsztejn 2001b). The Rhodian stamps cover the whole second century BCE to the transition to the first century BCE. The third century BCE is completely absent. The bestrepresented period is that of the apex of the power of Rhodes in its history (ca. 188–166 BCE, more than 50% of the finds). Contrary to the picture in Israel and the Eastern Mediterranean in general, the second half of the second century BCE is relatively poorly represented (about 25%). Compared to other areas of the city of Akko-Ptolemais, the Old City (excavations in the Crusader Hospitaller Citadel show that it was settled much earlier but the significant activities were very strong in the last 15 years of the third century BCE (probably following the battle of Raphia in 217 BCE, in preparation for a possible renewed attack by Antiochus 3rd; see Finkielsztejn 2001b: 190–191). However, it does not seem that the activity took place in the area of the marina studied in this report. Activity then continued at a relatively reduced volume, except for the period of Alexander Balas’ reign (150–145 BCE). As for the lower city, extending from the foot of Tel Akko to the Old City, settlement started early in the Ptolemaic period, augmented only in the last years of the third century BCE and became significant after the beginning of the Seleucid period, when the evidence starts to appear in the marina. The explanation of the excavation’s results could be that, on the one hand, the upper layers (from the first century BCE onward) had previously been removed, and that, on the other hand, the dredging did not reach the earlier layers (from the third century BCE downward). Another explanation may be that the harbor indeed was not built or enlarged before the Seleucid period, i.e., in the course of the first quarter of the second century BCE, to become the main harbor of the Southern Levant (see the relevant chapters on the contexts in this volume). In the catalogue, all dates are BCE, unless otherwise stated.

First Quarter of the Second Century BCE 1 – 31/93.F/1755 – Rectangular stamp. Complete handle with rim and shoulder.

1:1 Σωκράτευς Torch ↑ On the evidence of combinations with eponyms, Grace dated Σωκράτης 2nd’s career to ca. 212–185 (Grace 1974: 200). It may be understood that he endorsed amphoras dated by eponyms from Κλέαρχος or Παυσανίας 2nd down to Νικασαγόρας 1st (Grace 1985: 9). Σωκράτης 2nd is linked with the fabricant Ὄλυμπος by the torch device (21–22) and both endorsed amphoras dated by Αἰνησίδαμος 2nd (Lund 1993: 360, fig. 1; see Giveon 1963: 20–21; see 12 and 16). This evidence may point to an association of both fabricants in the same workshop. This author is considering addressing elsewhere the topic of Rhodian workshops more in detail. Date: ca. 200–172/170. 2 – 31/93.E/5 – Rectangular stamp.

1:1 Νικαγίδος The career of this fabricant is dated in Period III, at least before the building of the Middle Stoa at the Athenian Agora (Grace 1968: 177, no. 12; 1985; Finkielsztejn 2001a: 176–177). Νικαγίς endorsed an amphora dated by the eponym Καλλικρατίδας 2nd (dated ca. 175–173; see Grace 1985: 9) found among the offerings of a tomb in Marissa (see Regev 1991). See Nilsson 1909: 59–60, 101–103 on the fabricant’s female name; but see also Masson 1986: 40. Date: ca. 200–175/173. 3 – 31/93.? – Rectangular stamp.

Not to scale Ε .[---------] 165

Gerald Finkielsztejn

Only one letter is legible. The profile of the handle seems to point to a date around the turn of the third– second centuries.

The stalk of the caduceus crosses the initial phi of the fabricant’s name. Reading of the month is thanks to the first letter visible here, confirmed by a better-impressed stamp from the same die in V. Grace’s file (my thanks to A. Dimoulinis). The spelling of the month with a zeta (note the form of the zeta: Ι with longer horizontal lines) instead of the sigma is not rare (see Fisher 1989: 185, no. 6, to be corrected). Both our dies are unusual as the fabricant Φιλαίνιος usually appears with a wreath or without any device (6–8). The eponym Ἱέρων 1st is known on the same type with months Ἀγριάνιος, Πεδαγείτνυος and Ὑακίνθιος (Porro 1916: 117, nos. 111, 5, 7; V. Grace’s cards in the American School of Classical Studies; see also IG XII, 1: no. 1149, 8). The type of the stamp naming the fabricant is already known on an amphora, perhaps dated by the eponym Κρατίδας. Indeed, the latter’s die bears the wreath as a device but lacks the month (Calvet 1972: 27, nos. 44, 28–29, 50; this unusual association is based on the context of find, the same for both handles; this suggestion should have required confirmation, as the month would thus be missing, but, unfortunately, Y. Calvet informed me that the finds are no longer available for examination; 1978: 227, no. 44, pl. LV). The eponym Ἱέρων 1st appears to have succeeded Κρατίδας (Nicolaou and Empereur 1986: 516, no. 2, with dating in 199 and 198, according to Grace’s chronology). Date: ca. 186.

4 – 31/93.F/549 – Rectangular stamp.

1:1 Ἀγοράνακτος Δαλίου This is the third and last style used by the fabricant Ἀγορᾶναξ for his own stamps (Finkielsztejn 1993: 104–105). His successor, Μαρσύας (9), used the same style at the beginning of his career. Both put the month with their own name and a head of Helios on the eponym stamp (Grace and Pétropoulakou 1970: 302, E1; see 12). Date: ca. 195–187 (this style; eponyms Σώδαμος to Κρατίδας; see Grace 1974: 200). 5 – 31/93.F/433 –Rectangular stamps. Neck with top of two handles.

6 – 31/93.E/29 – Rectangular stamp. Handle with rim.

1:5

1:5

1:1

1:1 Φιλαινίου

Fabricant stamp Top of Φιλαινίου Caduceus ← Frame.

Same fabricant as 5. At the beginning of his career, he used the wreath on rectangular stamps (CIL VIII, Suppl. 3: 2190, no. 22639, 2, a–b, eponym Ξενοφάνης; Nilsson 1909: 495, no. 424, 7, eponym Φιλόδαμος 2nd; see Τιμασαγόρας, 10) and on circular stamps (Pâris 1913: 162, no. LVI, 5; 1914: 316, no. LXXXIX, eponyms Κλεώνυμος 2nd and Φιλόδαμος 2nd; see 11). The unusual caduceus is discussed above (5). The plain type is the latest, used during most of his career (Badalianz 1980: 165, eponym Ἀγέστρατος 2nd, 20; see Nicolaou and Empereur 1986: 518–519, no. 4). According to the associations with eponym names, the career of the fabricant Φιλαίνιος covered the period ca. 189–161 at least. Date: ca. 178/176–161 (this type).

1:1 Eponym stamp Ἐπ’Ἱέρων ̣[ος] Caduceus → {Ζ̣}μινθίου ̣ Frame.

166

Amphora Stamps from the Akko Marina

7 – 31/93.F/1292 – Rectangular stamp.

11 – 31/93.F/760 – Rectangular stamp.

Not to scale 1:1

Φιλαινί[ο]υ Ἐ̣πὶ Φιλ ̣ο̣ δ ̣ά ̣μου

Same fabricant as 6. 8 – 31/93.F/1747 – Rectangular stamp.

1:1 Φιλαινίου Same fabricant as 6. 9 – 31 – 7/93.E/168 – Rectangular stamp.

There appear to have been three eponyms of the name Φιλόδαμος (Badoud 2015). Φιλόδαμος 1st was in charge in Period Ib (Finkielsztejn 2001a: 188). Φιλόδαμος 2nd appears on all the types of the fabricant Φιλαίνιος (Finkielsztejn 2001a: 120–121; see 5–8), therefore he is the penultimate eponym of the group dating amphoras in the Villanova deposit (Maiuri 1924: 259, no. XXXI; see 5-6 and 10). Since the month name is not mentioned on our eponym stamp, it normally appeared on the associated fabricant stamp. Rare stamps of Φιλαίνιος do bear his name with the month, but no association is known to me. Date: c. 183 BCE (see also Badoud 2015: 257). 12 – 31/93.F/547 – Rectangular stamp.

1:1 Μαρσύας Σμιν̣θ̣[ίου] This is the first style of stamp (with non-retrograde nu) used by this fabricant, at the beginning of his career when he succeeded Ἀγορᾶναξ (4). The combined stamps of eponyms – bearing the head of Helios as a device – with the same script are early in the fabricant’s career (from Ἱέρων 1st to Αἰνησίδαμος 2nd, see 5 and 12). Date: ca. 186–179/177. 10 – 31/93.E/87 – Rectangular stamp.

1:1 Head of Ἐ̣πὶ Αἰ{ν} Helios ησιδά ̣[μ](ου)

Nu retrograde. This amphora was endorsed by the fabricant Μαρσύας who used this style a few years after the beginning of his activity (see 4 and 9). For the associations of Αἰνησίδαμος 2nd with the fabricants Σωκράτης 2nd and Ὄλυμπος, see 1 and 21–22. Date: ca. 179/177.

1:1 Ἐπὶ Τιμασαγόρα Βαδρομίου Τιμασαγόρας belongs to the group of eponyms dating amphoras in the Villanova workshop (Maiuri 1924: 258–259, nos. XXIX–XXX; see Brugnone 1986: 39, no. 55, for associations with the fabricant Φιλαίνιος, with wreath as a device, nos. 5–6). It can be deduced that it fits into the slot of ca. 196, according to Grace’s chronology, the only one not attributed to any other eponym in the literature (Empereur 1990: 207). Date: ca. 184.

167

Gerald Finkielsztejn

14 – 31/93.F/1533 – Rectangular stamp.

13 – 31/93.F/552 – Rectangular stamps. Two complete handles with neck and shoulder.

1:1 Ἐπὶ Καλ λικράτευ̣[ς] The profile of the handle and the script of the stamp date this amphora to the term of Καλλικράτης 2nd. Date: ca. 177/175 or 176/174. 15 – 31/93.E/7 – Rectangular stamp.

1:1 Star Ἐπὶ Κλευ κρ ̣ά ̣τευς

1:5

This stamp dated an amphora endorsed by the fabricant Ἀριστείδας 2nd, on the basis of the star device, and the month was mentioned on the stamp naming the fabricant. Therefore, the eponym Κλευκράτης 1st is named here. Note that eponyms of Period II with a similar star device may rather have dated amphoras endorsed by the fabricant Ἱεροκλῆς 1st. Indeed, the profile of the latter’s handles is less angular (see Macalister 1901: 40, no. 119, pl. 2, 50; Grace and Pétropoulakou 1970: 309, E 24; Ariel 1990: 63, S 302, corrected in Finkielsztejn 1994). Date: ca. 175/173–173/171.

1:1 Fabricant stamp Star Star Ἀ[ρ]ιστάρχ[ο]υ Star [Star]

16 – 31/93.C/73 – Rectangular stamp.

1:1

1:1

Eponym stamp Ἐπὶ Αἰνήτο ρος Δ{ε}οσθύου

Ἐπὶ Συμμ άχου Ἀρταμιτ(ί)ο(υ) The eponym Σύμμαχος is known, associated with an early rose type secondary stamp by the fabricant Δαμοκράτης 1st (Grace 1968: 175, no. 3; 1985: 9; and so is Κλευκράτης 1st, 15; Finkielsztejn 2001a: 117). He is associated with the fabricant Σωκράτης 2nd (1; Giveon 1983: 20–21, pl. 8). Date: ca. 175/ 173–173/171.

Month name with epsilon instead of iota. This association is already known (see Brugnone 1986: 47, no. 66). Date: ca. 178/176.

168

Amphora Stamps from the Akko Marina

Second Quarter of the Second Century BCE 17 – Unnumbered – Circular stamps. Complete neck with top of the two handles. Not to scale Secondary square stamp Β Same dies as the stamps in Grace 1985: 9, 46, 3, a–b, pl. 1, which date an amphora endorsed by the fabricant Ἱπποκράτης. This same association of names appears on 17, above, with a different month. 19 – 31/93.E/42 – Circular stamp.

1:5 Fabricant stamp (not illustrated) Ἱπποκράτε[υ]ς Rose Eponym stamp (not illustrated) [Ἐπὶ Θε]α̣ιδήτου Δα̣λ ̣[ίου] Rose This pair of names is already known (see 18). It is noticeable that no secondary stamp is associated with the main stamps, a fact that should be taken into account for understanding the meaning of this additional stamping. The most likely hypothesis for the practice is that the potter who actually made the amphora imprinted it, probably to count the number of vessels he produced for payment purposes (Finkielsztejn 2001a: 114–116; Palaczyk 1999, 2000). J.-Y. Empereur suggested that it may be related to the branch of a network of workshops managed by the same fabricant but did not elaborate on the reasons for the need for such information (Nicolaou and Empereur 1986: 521, under no. 6; note that the secondary stamp IC, that appear on that example, was also used by the fabricants Ἀριστοκλῆς 2nd and Τιμόξενος, relatives of Ἱπποκράτης; Finkielsztejn 2001a: 118–119). Date: ca. 171/169.

1:1 Ἐπὶ Ἀρατ(ο)φάνευς Δαλίου Rose Frame. This stamp is dated by the eponym Ἀρατοφάνης 1st. Note the similarity between the style of this stamp and that the stamp naming the fabricant Ἀριστοκλῆς 2nd (32), who may have endorsed this amphora too. It is not clear, of course, whether both our handles belonged to the same amphora. Such an association is already known (Nicolaou and Empereur 1986: 520, no. 5). For the homonym Ἀρατοφάνης 2nd see Ariel 1988: 31–32 and Ariel and Finkielsztejn 1994: 194, SAH 16 (note that he should be dated now ca. 109). Date: ca. 169/167–165/163. 20 – 31/93.F/408 – Rectangular stamp.

18 – 31/93.F/1507 – Circular stamp.

1:1 Ἐπὶ Ἀγεστράτου Πανάμου

Not to scale Main stamp Ἐπὶ Θ̣[ε]αιδήτου Πανάμου Rose Frame.

I suggest that the eponym Ἀγέστρατος 2nd may have been the last one represented in the Pergamon deposit, i.e., Grace’s Period III, which may now be dated ca. 198–161 (Grace and Pétropoulakou 1970: 290–291; Grace 1985: 8–9, 42; Finkielsztejn 1995b: 281; 2001a: 192). Date: ca. 161.

169

Gerald Finkielsztejn

21 – 31/93.F/144 – Rectangular stamp.

24 – 126/92.F/47 – Rectangular stamp. Handle of a half-capacity amphora.

1:1

1:1

Ἀθανο Cornucopia ↑ δότου

Ὀλύμ̣[πο]υ Rose ↑ The fabricant Ὄλυμπος uses the torch more often than the rose as a device (see Grace 1934: 306, no. 5, from a different die). His name is combined with the eponym Αἰνησίδαμος 2nd (12) and appears six times in the Pergamon deposit (Schuchhardt 1895: 475, no. 1160; Pridik 1926: 329, 331). The torch device links Ὄλυμπος with the fabricants Σωκράτης 2nd (1), and Ῥόδιππος (see Ariel 1990: 57, S 239, S 242, 58, S 256). The only likely restorations of the eponym name combined with our type in Gentili 1958: 37, no. 22, are Δαμοκλῆς 2nd or Σωσικλῆς (33). The former’s term (ca. 176/174) is close in date to that of Αἰνησίδαμος 2nd, and the latter’s term would suggest a career of about 25 years for the fabricant (down to ca. 154/155), a less likely possibility. Date: ca. 179/ 177–176/174 at least.

The fabricant Ἀθανόδοτος was associated with eponyms from Κλευκράτης 1st (15) (Levi 1964: 272, nos. 332–333) to Ἀριστείδας 2nd (Lund 1993: 360, fig. 1). Date: ca. 174/172–168/166. 25 – 31/93.F/181 – Rectangular stamp. Complete handle with neck and shoulder.

22 – 31/93.F/1777 – Rectangular stamp.

1:5

1:1 Ὀλύμπο[υ Rose or Torch] ↑ Same fabricant as 21. 1:1 Ἀμύντα Wreath

23 – 31/93.F/761 – Rectangular stamp.

This is the most common type used by the fabricant Ἀμύντας. He may have been succeeded one or two years later by the fabricant Βρόμιος who also used the wreath as a device, apparently at the beginning of his career (Finkielsztejn 2001a: 121–123; see Γόργων, 34). Ἀμύντας was associated with eponyms from Αἰνησίδαμος (12; Empereur and Hesnard 1987: 60, no. 11) and Ἡραγόρας (neck of an amphora from the Studium Biblicum Franciscanum collection in Jerusalem, courtesy of M. Piccirillo; see Nachtergael 1978: 28–30, and Fisher 1989: 185, no. 1). Date: ca. 179/177–ca. 157/156.

1:1 [Ἀγαθ]οκλεῦς It is not clear (although unlikely) whether the same person is to be identified on earlier stamps bearing the same name but with the month mentioned on the fabricant stamp (see Ariel 1990: 38–39, S 58–60, without month, S 61–65, with month). Name connections with eponyms like Αἰνήτωρ (13) or Ἀγέστρατος 2nd (20), put the career of Ἀγαθοκλῆς 2nd in the second half of Period III. Date: ca. 178/176–161. 170

Amphora Stamps from the Akko Marina

26 – 31/93.E/83 – Rectangular stamp. Complete handle with neck.

stamp with caduceus (27). In any event, if the (heavy) restorations of the names on the amphora IG XIV, no. 2393, 7 are correct, this type without caduceus was used earlier than the other one, which is at least ca. 188 (eponym Πρατοφάνης). The profile of the handle seems in favor of a later date in Period III for the activity of this Ἀντίμαχος, and the restoration of the name of Ἀρατοφάνης 1st (19) could better fit the association just noted. Date: 160s–150s. 29 – 31/93.F/766 – Rectangular stamp.

1:5 1:1 Ἀριστογείτου 1:1 Ἀμύντα Wreath Same fabricant as 25. 27 – 31/93.E/78 – Rectangular stamp.

This fabricant is named on stamps with and without the mention of the month (Sztetyłło 1976: 38, no. 59; Finkielsztejn 1990: vol. II, 35, no. 82 = Palestine Archaeological Museum 33.2434, from Samaria-Sebaste). Four examples with his name in the Pergamon deposit (Schuchhardt 1895: 447, no. 893) and the angular profile of his handles put Ἀριστόγειτος’ activity in late Period III. Date: second quarter of the second century BCE. 30 – 31/93.E/23 – Rectangular stamp. Half neck with one complete handle.

1:1 Ἀντι̣μ̣ά̣χου Caduceus → This type of stamp with caduceus is known in 13 examples in the Pergamon deposit. It is not clear whether this fabricant Ἀντίμαχος is the same person as the following one (28). This type was associated with stamps naming Ἀθανόδοτος (Grace 1985: 10), Σωσικλῆς (Deshayes 1963: 33; 34, no. 32, pl. LXVI, 13, LXVIII, 1–2) and Ἀλεξίμαχος (Pridik 1926: 309, 321, 331). Date: ca. 171/169–147 at least. 1:5 28 – 31/93.F/187 – Rectangular stamp.

Not to scale 1:1 Ἀντιμάχου It is not clear whether this type of stamp names the same fabricant as the one endorsing the preceding

Νάνιος The handle was eventually broken after discovery. The career of the fabricant Νᾶνις is dated in the second quarter of the second century by Grace and Pétropoulakou 1970: 304, E11, a dating still valid.

171

Gerald Finkielsztejn

33 – 31/93.F/144 – Circular stamp.

31 – 31/93.F/548 – Rectangular stamp.

1:1 Αἰσώπου Statue of Rhodos (?)

1:1 Ἐπ̣ὶ Σωσικλ̣[εῦς Ἀρτ]α̣μιτ̣ίο̣υ Rose

This fabricant seems to have been active at the end of the period of the Pergamon deposit, i.e., Period III, since only two of his stamps were found there (Schuchhardt 1895: 443, nos. 854 and 855, which may be from the same die as ours). I do not know any association with eponym. Αἴσωπος may have worked in association with (or have been succeeded by) Νύσιος who used the same statue type (among others) and the stamps of whom are more commonly found. The latter was active from the end of Period III to the beginning of Period V (see 1895: 458–459, nos. 1003–1004, to be corrected; Brugnone 1986: 71, no. 112). For the identification of the statue as that of the nymph Rhodos, daughter of Helios, see Boussac 1992: 120–123. Date: second quarter of the second century BCE.

Frame. This amphora may have been endorsed by the fabricant Ἀριστοκλῆς 2nd (32), according to the aspect of the rose which is known for the stamps of an amphora from Marissa associating this fabricant with the eponym Παυσανίας 3rd (Ariel and Finkielsztejn 2003: 138–140, Rh 1). The eponym Σωσικλῆς is linked – by rare secondary stamps of the fabricants Μαρσύας (9) and Ἐπίγονος 2nd – to the eponym Ἡραγόρας whom he may have succeeded (Finkielsztejn 1993: 239–240; 2001a: pl. VIII, 152, 156, X, 189; VIII, 158; 125, with n. 143). Date: ca. 155/154. 34 – 31/93.F/567 – Rectangular stamp.

32 – 31/93.E/67 – Circular stamp. 1:1 Ἐπὶ Γ ̣ό[ρ] γωνος ̣ [month] The eponym Γόργων was in office in Period IV (Börker 1974: 44, no. 35). He is known on rectangular stamps with head of Helios without mention of the month, and not on lozenge-shaped ones, by the fabricant Θεύμναστος (36). He thus may have been the first eponym to date an amphora by the fabricant Βρόμιος, still putting a wreath on his own stamp (Jalabert and Mouterde 1959: 103, no. 2188, B; Finkielsztejn 1993: 239–241, 256–257; see Ἀμύντας, 25–26). Date: ca. 154/153 (perhaps 153).

1:1 Ἀριστοκ[λε̣]ῦς Rose Frame. The fabricant Ἀριστοκλῆς 2nd is very commonly found and a member of a dynasty of Rhodian fabricants (Grace 1985: 10). Ἀριστοκλῆς 2nd may have endorsed amphora 19, dated by the eponym Ἀρατοφάνης 1st. He was associated with eponyms Ἀθανόδοτος (1985: 10), Τεισαγόρας (Börker 1985: 398) and Ἀναξίβουλος (my revision of Landau and Tsaferis 1979, combining their no. 28, p. 156, fabricant Ἀρ[ιστοκλῆς] 2nd and fragments of the same handle nos. 31–32, p. 157, eponym Ἀναξίβ[ου]λος). The associations with eponyms seem to support the lower chronology, as the fabricant’s career thus covers about 32 years instead of the less likely 44 according to Grace’s chronology. Date: ca. 171/169–ca. 140/139.

Third Quarter of the Second Century BCE 35 – 31/93.F/762 – Rectangular stamp.

1:1 Ἐπὶ Εὐ δάμου Ὑακινθίου 172

Amphora Stamps from the Akko Marina

Εὔδαμος is one of the eight eponyms dating lozenge-shaped stamps endorsed by the fabricant Θεύμναστος (36; Finkielsztejn 2001a: 125, 127). The group is dated in the very end of Period IV and one of them––Τιμόδικος––not known on stamps from Carthage nor probably from Corinth (Gentili 1958: 27–28; Grace and Pétropoulakou 1970: 311, E31; Brugnone 1986: 14, no. 15, 41, no. 56) could be the first eponym of Period V. Connections with other fabricants contribute to the refinement of the chronology of the group. Date: ca. 150–148.

38 – 31/93.C/194 – Rectangular stamp. Handle with neck and rim.

1:5

36 – 31/93.F/723 – Rectangular stamp.

1:1

1:1

Θ̣ευμν̣άστου Branch →

Ἐπὶ Ἀρ ̣ιστάκο υ Ὑα[κ]ι̣νθίου

This fabricant is better noticed for his use of stamps in the shape of a lozenge, dated by eight eponyms (see 35; Gentili 1958: 27–28; Zeitoun et al. 1998; Finkielsztejn 2001a: 125, 127). Date: ca. 152–145. 37 – 31/93.F/763 – Rectangular stamp.

1:1

Note that the eponym Ἀρίστακος should be identified with the so-called eponym *Ἀρίσταρχος 2nd of Period V in several publications with faulty readings (see Grace 1934: 219; 1948: 145; Kontorini 1989: 132–133, nos. 53, 144–145, nos. 60, 195). According to the style of the stamps with his name, he appears to have been in office in early Period V (see Nachtergael 1978: 40, n. 1). Date: ca. 137/136–134/133. 39 – 31/93.E/70 – Rectangular stamp.

Ἐ̣π ̣ὶ Ἀστ[υμή] ̣ δε̣[υ]ς ̣ Π̣α̣νάμου This stamp has been twice imperfectly stamped so that none of the two impressions give the whole reading of the inscription. In addition, the latter may in any case have been faulty. The only possibility for the restoration of the eponym name is Ἀστυμήδης 2nd. The aspect of the handle and the script indicate that the second of three homonyms dates this amphora. Date: ca. 144–142.

1:1 [Ἐ]π ̣ὶ [Τιμο]θ ̣ ̣έ ̣ ου Ἀ̣γριανίου If the reading of the legible letters is right, this is the only possible restoration of the eponym name. It fits the aspect of the handle since the term of Τιμόθεος fell in Period V (see Ariel and Finkielsztejn 1994: 218, SAH 86–87). Date: ca. 130–128. 40 – 31/93.F/967 – Rectangular stamp.

1:1 Ἐπ ̣ὶ Λ[ε]οντ ̣ίδ̣ ̣α Ἀρταμιτίου 173

Gerald Finkielsztejn

Last Quarter of the Second Century BCE

Retrograde reading. A complete amphora from Tel Istabah provides the association of a stamp from the same die with a stamp of the fabricant Σωσίφιλος (Retrograde reading with cornucopia; my thanks to G. Mazor and R. Bar-Nathan for entrusting me with the publication of the amphoras from their excavations in Bet She’an; Finkielsztejn 1993: 343–344, no. 31). Λεοντίδας is combined with the fabricant Εὐφράνωρ 2nd on circular stamps with head of Helios and a specific secondary stamp (Finkielsztejn 2001a: 138–140). According to the evidence of other types he seems to have been succeeded by Κληνόστρατος who himself seems to have preceded Πολυάρατος 2nd (Finkielsztejn 1993: 288, 309–310). Date: ca. 127.

43 – 31/93.F/1750 – Rectangular stamp. Top of handle with rim.

1:5

1:1

41 – 31/93.F/454 – Rectangular stamp. Διο̣δ̣ο̣ Grape του cluster

The fabricant Διόδοτος 2nd is named here. An earlier homonym, fabricant Διόδοτος 1st, was active in the end of Period III to the end of IV at least (Schuchhardt 1895: 458, no. 1001; combination with the eponym Αὐτοκράτης 1st – probably dated ca. 146 – on an amphora from Marissa; my thanks to A. Kloner for entrusting me with the publication of the amphoras from that site). Διόδοτος 2nd was active in the second half of the second century, according to the type and style of the stamp (Le Roy 1984: 311, no. 12, correct to second century; Finkielsztejn 1989: 229). Date: ca. last two decades of the second century BCE.

1:1 Δ̣[ρ]α̣κοντ ̣ίδα ̣ Anchor ← The combinations of stamps endorsed by the fabricant Δρακοντίδας date his career in Period V (Nachtergael 1978: 40, n. 6; the present author agrees with Nachtergael’s suggestion to emend the eponym’s name in Gentili 1958: 36, no. 19, from Ἀρχοκράτης to Αὐτοκράτης 1st; there are several unpublished associations with eponyms – among them Πολυάρατος 2nd – on amphoras from Marissa; my thanks to A. Kloner for entrusting me with the publication of the amphoras from that site). Date: ca. 146–125.

44 – 31/93.E/75 – Rectangular stamp.

42 – 31/93.C/19 – Rectangular stamp. 1:1 [Ἐπὶ Ν]ι̣κ ̣ομά χου Πανάμ̣ο̣υ̣ The eponym Νικόμαχος is known on a combination with the fabricant Ἀπολλοφάνης 1st (45). He should be dated in early Period VI, on the basis of the style of his stamps, and because his name does not appear on any amphora stamps from Period V contexts (see Grace and Pétropoulakou 1970: 317, E 46; Ariel and Finkielsztejn 1994: 210–211, SAH 66–68). Date: ca. 107–100.

1:1 Λυσί[ωνος] Anchor ← This fabricant is combined with five eponyms that date his amphoras to Period V (Pogwisch 1859: 65–73, pl. 3, Ἀνδρόνικος; unpublished amphora from Tel Istabah, Καλλικράτης 3rd; my thanks to G. Mazor and R. Bar-Nathan for entrusting me with the publication of the amphoras from their excavations at Bet She’an). Date: ca. 132–130.

174

Amphora Stamps from the Akko Marina

45 – 31/93.E/38 – Rectangular stamp.

The suggestions for restoration of the month are quite likely. There seems to have been a double stamping according to the lower edge of the stamp which is thickened and bears unclear letters. The aspect of the handle is not typical enough to help date this amphora. The possible restorations of the eponym name are Σωσικλῆς (ca. 155–154; 33) or (rather unlikely) Σωκράτης (very beginning of the first century).

1:1 [Ἀ]πολλο [φ]ά̣ν̣ευς Grape cluster On the fabricant Ἀπολλοφάνης 1st see Ariel and Finkielsztejn 1994: 193, SAH 13–14, SAH 21–22. He is known on a combination with the eponym Νικόμαχος on an amphora from the Hecht Museum, University of Haifa (see 44). The fabricant Ἀπολλοφάνης 2nd used circular stamps with monograms or rose and was active in the beginning of the first century (Grace 1965: 15, F–G). Ἀπολλοφάνης 1st was active in late Period V and early Period VI. Date: last quarter of the second century.

Knidos Only one Hellenistic stamped amphora fragment of non-Rhodian origin has been found, a proportion that is relatively low but not surprising in the Eastern Mediterranean. It is from Knidos. On the class, see Grace and Pétropoulakou 1970: 317–354; Grace 1985: 13–18, 31–35. 48 – 31/93.DBF/78/2 – Rectangular stamp. Neck with top of both handles.

46 – 31/93.B/58 – Rectangular stamp.

Not to scale [Ἐπὶ -----] ν ̣[α]κ̣τ ̣[ο]ς [Ἀγρια]νίου The reading is tentative but quite likely. The profile of the handle and the script date this amphora in the last quarter of the second century and only two names may be restored: Ἀριστᾶναξ 2nd (ca. 112) or Ἀγορᾶναξ (ca. 108).

1:5 The same stamp on each handle:

Unclear Date 47 – 31/93.1/9 – Rectangular stamp.

1:1 Ἐπὶ Σ ευ ̣ς ̣ (?) Δ̣[αλίου] or Δ̣[ιoσθύ]ο[υ]

1:1 Ἐ[πὶ φρου] ράρχου Ἀρ ̣ι̣σ τίω̣̣ [ν]ο ̣ς ̣Δα ν ̣αοῦ ̣ Κνι(δίον) Retrograde reading, restored thanks to Ph. Matheson, who also provided the Knidian type number of V. Grace’s corpus, KT 1888. Lunate sigma, cursive omega. Ἀριστίων was phrourarch in Period IVAa (Grace 1985: 31). Δαναός is the fabricant. Date: early in the period 188–167.

175

Gerald Finkielsztejn

Byzantine period Medieval period

49 – 31/93.C/850 – Incuse stamp in the shape of a duck (?) at bottom of handle.

50 – 31/93.C/181 – Circular stamp.

1:5

1:1

1:5

Circle of incuse triangles with large plain center. Anepigraph. The stamp diameter is 2.6cm. The handle is very flat (L preserved 5cm, W 6.2cm, Th 0.9cm) and connected to the top of the thickened lip of a cylindrical neck (D 10cm). The fabric is gray toward the inside, light reddish-brown toward the outside, and the surface is light reddish-yellow; there are very rare white inclusions. Similar but not identical stamps were studied by Nergìs Günsenìn in her Ph.D., of which I was provided with figures only (Günsenìn 1990: pl. XXVI, 4a–c, and table of fig. 43, line 5, columns 1 [the closest], 2–4 and 9). Günsenìn kindly provided the following additional references by email (translated from French by the present author): “The illustrated amphora is located in the Sinope Museum. It belongs to Type Günsenin 1. Similar stamps are known in the Edirne Museum, in Istanbul (Demangel and Mamboury 1939: fig. 201: 53, 57) and in the Serçe Limani shipwreck (Van Doorninck 1989: fig. 3: 27)”. However, the fragment of our handle does not match that of the Type Günsenin 1 amphoras and no parallel is known to me. The date of Type Günsenin 1 amphoras is 10th–11th century CE.

1:1 Κω (inside a duck ←) Only part of the head, the complete eye and the front part of the body with one leg of the animal appear. The complete form of a bird, probably a duck, is restored thanks to a better example from the same die, found in Bethany (Saller 1957: 216, no. 541, pl. 123, b, 3). Both stamps appear on the lower part of the handle of a jar. This type of incuse stamps representing animals is very common on Late Roman Red Slip Ware. These parallels, together with the ribbed body still attached to the handle from Jerusalem, suggest a date in the Byzantine period. The small bar above the second letter (here omega) to indicate an abbreviation is also very common in that period. If a person is named here – a magistrate or a potter – then the name Κωνσταντῖνος is the most common possibility. On the other hand, the abbreviation could refer to either the place of manufacture or the city responsible for the control of either the pottery-making or the trade of the content. In that case the name of the capital of the Byzantine empire, Κωνσταντινούπολις (Constantinople, former Byzantium), comes to mind. Only the lower vertical branch of the handle is preserved (H preserved 16cm, W max. 5.5cm, Th max. 3.4cm, both latter at top break) and the stamp is set at the bottom, near the lower attachment on the shoulder or upper body. It was detached from the body at the bottom and broken not far below the bend at the top. The fabric is reddish-yellow and the surface is brown to reddish-brown; there are very few white inclusions and copper-colored fine mica.

Acknowledgments The author wish to thank the Israel Antiquities Authority for the institutional support, M. Feinberg Vamosh for the English editing and T. Sagiv and C. Amit for the photographs.

References Ariel D.T. 1988. Two Rhodian Amphoras. Israel Exploration Journal 38: 31–35. Ariel D.T. 1990. Excavations at the City of David 1978–1985 II. (Qedem 30). Jerusalem. Ariel D.T. and Finkielsztejn G. 1994. Stamped Amphora Handles. In S. Herbert (ed.). Tel Anafa

176

Amphora Stamps from the Akko Marina

I, i. Final Report on Ten Years of Excavation at a

Hellenistc and Roman Settlement in Northern Israel (Journal of Roman Archaeology Supplementary Series 10). Pp. 183–240. Ann Arbor.

Ariel D.T. and Finkielsztejn G. 2003. Amphora Stamps and Imported Amphoras In A. Kloner, Maresha Excavations Final Report I. Subterranean Complexes 21, 44, 70. (IAA Reports 17). Pp. 137– 151. Jerusalem.

Finkielsztejn G. 1989. Tell Abu Hawam: réexamen des périodes hellénistique et perse (fouilles de 1929 à 1933). Chronique Archéologique. Revue Biblique 96: 224–234. Finkielsztejn G. 1990. Amphores et timbres d’amphores

importées en Palestine à l’époque hellénistique: orientations de recherches et premiers résultats (3 vols.) (I and II, 1–2). Memoir presented to the Académie des Inscriptions et Belles-Lettres. Ecole Biblique et Archéologique Française. Jerusalem (unpublished).

Badalianz Y.S. 1980. On the chronological position of personal names on the Rhodes amphoras. Sovetskaja Archeologἰa 2: 161–166 (Russian, English summary).

Finkielsztejn G. 1993. Amphores et timbres d’amphores importées en Palestine à l’époque hellénistique: études de chronologie et d’histoire.

Badoud N. 2015. Le Temps de Rhodes. Une chronologie des inscriptions de la cité fondée sur l’étude de ses institutions (Vestigia 63). Munich.

Finkielsztejn G. 1994. Review of Ariel 1990 (see above). Palesine Exploration Quarterly 126: 61–62.

Börker Ch. 1974. Griechische Amphorenstempel von Tell Halaf bis zum persischen Golf. Baghdader Mitteilungen 7: 31–49. Börker Ch. 1985. Review of L. Criscuolo, Bolli

d’anfora greci e romani. Gnomon 57: 397–398.

Boussac M.F. 1992. Les sceaux de Délos 1.

Sceaux publics, Apollon, Hélios, Artémis, Hécate. Recherches Franco-Helléniques II. Paris. Brugnone A. 1986. Bolli anforari rodî dalla necropoli di Lilibeo. Κωκαλος XXXII: 1–82. Calvet Y. 1972. Salamine de Chypre III. Les timbres amphoriques (1965–1970). Paris. Calvet Y. 1978. Timbres amphoriques de Salamine (1971–1974). RDAC: 222–234, Pl. LIII–LVI.

Unpublished Ph.D. diss., Sorbonne University. Paris.

Finkielsztejn G. 1995a. Amphoras Imported to EretzIsrael in the Hellenistic Period and New Developments in the Chronology of the Rhodian Stamps. Abstracts of the 21st Archaeological Conference in Israel. (Yad Ben Zvi, May 18–19 1995). Pp. 12–13. Jerusalem. (Hebrew). Finkielsztejn G. 1995b. Chronologie basse des timbres amphoriques rhodiens et évaluation des exportations d’amphores. Acta Hyperborea 6: 279–296. Finkielsztejn G. 2001a. Chronologie détaillée et

révisée des éponymes amphoriques rhodiens de 270 à 108 av. J.-C. environ. Premier bilan (BAR International Series 990). Oxford.

Finkielsztejn G. 2001b. Politique et commerce à Rhodes au IIe s. a.C.: le témoignage des exportations d’amphores. In A. Bresson and R. Descat (eds.).

Les cités d’Asie Mineure occidentale au IIe s. a.C.

CIL Corpus Inscriptionum Latinarum

(Ausonius. Etudes 8). Pp. 181–196. Bordeaux.

Demangel R. and Mamboury E. 1939. Le quartier des

Fisher M. 1989. Hellenistic Pottery, Stamped Handles. In Z. Herzog, G. Rapp, Jr. and O. Nerbi (eds.). Excavations at Tel Michal, Israel. Mineapolis– Tel Aviv. Pp. 185–186, Fig. 13.4, Pl. 68.

manganes et la première région de Constantinople.

(Recherches francaises en Turquie 2). Paris.

Deshayes J. 1963. La nécropole de Ktima. Mission

Jean Bérard, 1953–55. Bibliothèque archéogique et historique 75. Paris.

Galili E. This volume a. The Akko Marina Archaeological Project – Introduction.

Empereur J.Y. 1990. The Chronology of Rhodian Amphoras in the Hellenistic Period. Acts of the

Gentili G.V. 1958. I Timbri Anforari nel Museo Nazionale di Siracusa. Archivio Storico Siracusano IV: 8–95.

Second Scientific Conference on Hellenistic Pottery.

(Rhodes, March 22–25 1989). Pp. 199–209. Athens. (Greek).

Empereur J.Y. and Hesnard A. 1987. Les amphores hellénistiques. Céramiques Hellénistiques et Romaines 2. Pp. 7–71. Besançοn.

Giveon R. 1963. A Ptolemaic Faience Bowl. Israel Exploration Journal 13: 20–29, Pl. 8. Grace V. 1934. Stamped Amphora Handles found in 1931–1932. Hesperia 3: 197–310.

177

Gerald Finkielsztejn

Grace V. 1948. Jars in Florida. Hesperia 17: 44–147.

Masson O. 1986. Les anses d’amphores et l’anthroponymie grecque. In J.-Y. Empereur and Y.Garlan (eds.). Recherches sur les amphores

Grace V. 1965. The Commercial Amphoras from the Antikythera Shipwreck. Transactions of the American Philosophical Society 55: 5–17.

grecques. Bulletin de correspondance hellénique. Supplément XIII: 37–44.

Grace V. 1968. Die gestempelten Amphorenhenkel aus stratigraphisch gesicherten Fundzusammenhangen. In Das Asklepieion. Altertumer von Pergamon XII. Pp. 175–178. Berlin.

Nachtergael G. 1978. La collection Marcel Hombert I. Timbres amphoriques et autres documents écrits acquis en Egypte. Papyrologica Bruxellensia 15.

Grace V. 1974. Revisions in Early Hellenistic Chronology. Athenische Mittelungen 89: 193–200.

Nicolaou I. and Empereur J.Y. 1986. Amphores rhodiennes du Musée de Nicosie. In J.Y. Empereur and Y. Garlan (eds.). Recherches sur les amphores

Bruxelles.

Grace V. 1985. The Middle Stoa Dated by Amphora Stamps. Hesperia 54: 1–54.

grecques. Bulletin de correspondance hellénique.

Grace V. and Pétropoulakou M. 1970. Les timbres amphoriques. In Ph. Bruneau (ed.). L’îlot de la

Nilsson M.P. 1909. Timbres amphoriques de Lindos. Exploration archéologique de Rhodes V.

Supplément XIII: 515–533.

Maison des Comédiens. Exploration Archéologique de Délos XXVII: 277–382.

Copenhague.

Palaczyk M. 1999. Die Zusatzstempel (‘Secondary Stamps’) der rhodischen Amphoren. Jahresheften des Österreichischen Archäologischen Institutes 68: 59–103.

Günsenìn N. 1990. Les amphores byzantines (Xe–

XIIIe siècles): typologie, production, circulation d’après les collections turques. Thèse de Doctorat de l’Université Paris I (Panthéon-Sorbonne). Paris (unpublished).

Palaczyk M. 2000. Secondary stamps in the Rhodian amphora production. In Rei Cretariæ Romanaæ Fautorum Acta 36. Pp. 416–425.

Jalabert L. and Mouterde R. 1959. Les inscriptions grecques et latines de la Syrie V. Bibliothèque archéogique et historique 66. Paris.

Pâris J. 1913. Une nouvelle collection rhodienne de timbres amphoriques. In Mélanges Holleaux. Pp.153–173. Paris.

Kontorini V. 1989. Inscriptions inédites de Rhodes II. Athens (Greek, French summary).

Pâris J. 1914. Timbres amphoriques de Rhodes. Bulletin de correspondance hellénique 38: 300–326.

Landau Y. and Tsaferis V. 1979. Tel Istabah, Beth Shean: The Excavations and Helenistic Jar Handles. Israel Exploration Journal 29:152–159.

Pogwisch P. 1859. Boletino Archeologico Sardo 5, V: 65–73 (non vidi).

Le Roy Ch. 1984. Timbres amphoriques provenant de Tanis: complément. Bulletin de l’institut français d’archéologie orientale 84: 307–315.

Porro G.G. 1916. Bolli di anfore rodie del museo nazionale romano. Annuario della Scuola Archeologica di Atene 2: 103–124.

Levi O. 1964. A Pottery Complex of the IIIrd–IInd Cent. B.C. from the Excavations of the Olbian Agora. In Olbija, Temenos i Agora. Pp. 259–280. (Russian, English summary). Moscow–Leningrad.

Pridik E.M. 1926. Zu den Amphorenstempeln. Klio 20: 303–331.

rhodischen

Regev D. 1991. Maresha. Hadashot Arkheologiyot 97: 81. (Hebrew).

Lund J. 1993. Rhodian Amphorae as Evidence for the Relation Between Late Punic Carthage and Rhodes. Acta Hyperborea 5: 359–375.

Regev D. 1992. Maresha. Excavations and Surveys in Israel 10: 152–153.

Macalister R.A.S. 1901. Amphora Handles with Greek Stamps from Tell Sandahannah. Palestine Exploration Fund Quarterly Statement 33: 25–43, 124–144, 394–397.

Saller S. 1957. Excavations at Bethany (1949–1953). Publications of the Studium Biblicum Franciscanum

Maiuri A. 1924. Una Fabbrica di Anfore Rodie.

Schuchhardt C. 1895. Amphorenstempel. In Die Inschriften von Pergamon II. Pp. 423–499. Berlin.

12. Jerusalem.

Annuario della Scuola Archeologica di Atene 4–5

(1921–1922): 249–269.

178

Amphora Stamps from the Akko Marina

Sztetyłło S. 1976. Les timbres céramiques (1965– 1975). Nea Paphos I. Warsaw.

Ἀριστείδας 2nd ep 24

Van Doorninck F.H. 1989, The cargo amphoras on the 7th-century Yassi Ada and 11th-century Serce Limani shipwrecks: Two examples of reuse of Byzantine amphoras as transport jars. In V. Deroche and J.-M. Spieser (eds.). Recherches sur la céramique byzantine Bulletin de correspondance hellénique 18: 247–257.

Ἀριστογείτου fab 29

Zeitoun N., Christophi C. and Empereur J.-Y. 1998. Les anses d’amphores du Musée gréco-romain d’Alexandrie. L’état d’avancement de la banque de données sur les amphores et le cas exemplaire du fabricant Theumnastos. In J.-Y. Empereur (ed.). Commerce et artisanat dans l’Alexandrie

hellénistique et romaine. Bulletin de correspondance hellénique. Supplément XXXIII: 367–399.

Index Greek inscriptions Ἀγαθοκλῆς (2nd ?) fab 23 Ἀγέστρατος 2nd ep 6, 20, 23 Ἀγορᾶναξ fab 4, 9 Ἀγορᾶναξ ep 46 Ἀγριάνιος month 5, 39, 46 Ἀθανόδοτος ep 27, 32 Ἀθανόδοτος fab 24 Αἰνησίδαμος 2nd ep 1, 9, 12, 13, 21, 25 Αἰνήτωρ ep 12, 13, 23 Αἴσωπος fab 31 Ἀλεξίμαχος ep 27 Ἀμύντας fab 25, 26, 34 Ἀναξίβουλος ep 32 Ἀνδρόνικος ep 42 Ἀντιμάχος fab 27, 28 Ἀντίπατρος ep Ἀπολλοφάνης 1st fab 44, 45 Ἀπολλοφάνης 2nd fab 45 Ἀρατοφάνης 1st ep 19, 32 Ἀρατοφάνης 2nd ep 19 Ἀρίστακος ep 38 Ἀριστᾶναξ 2nd ep 46 Ἀρίσταρχος fab 13 Ἀρίσταρχος 2nd 38

Ἀριστείδας 2nd fab Rh15 Ἀριστοκλῆς 2nd fab 17, 19, 32, 33 Ἀριστομβροτίδας fab Kn 48 Ἀρταμίτιος month 16, 33, 40 Ἀρχοκράτης ep 41 Ἀστυμήδης 2nd ep 37 Αὐτοκράτης 1st ep 41, 43 Β s.s. 18 Βαδρόμιος month 10 Βρόμιος fab 25, 34 Γόργων ep 25, 34 Δάλιος month 17, 19, 47 (?) Δαμοκλῆς 2nd ep 21 Δαμοκράτης 1st fab 16 Δεόσθυος month 13 Διόδοτος 1st fab 43 Διόδοτος 2nd fab 43 Διόσθυος month 13 Δρακοντίδας fab 41 Ἐπίγονος 2nd fab 33 Εὔδαμος ep 35 Εὐφράνωρ 2nd fab 40 Ζμίνθιος month 5 Ἡραγόρας ep 25, 33 Θεαίδητος ep 17, 18 Θεύμναστος fab 34, 35, 36 ἰερεύς 3 Ἱεροκλῆς 1st fab 15 Ἱέρων 1st ep 5, 9 Ἱπποκράτης fab 17, 18 ΙϹ s.s. 17 Καλλικράτης 2nd ep 14 Καλλικράτης 3rd ep 42 Καλλικρατίδας 2nd ep 2 Κλέαρχος ep 1 Κλευκράτης 1st ep 15, 24 Κλεώνυμος 2nd ep 6 Κληνόστρατος ep 40 Κνι(δίον) Kn 48 Κρατίδας ep 4, 5

179

Gerald Finkielsztejn

Devices

Κω( By 49 Κωνσταντῖνος By 49 Λεοντίδας ep 40

Anchor 41, 42

Λυσίων ep 42

Branch 36

Μαρσύας fab 4, 9, 12, 33

Caduceus 5, 6, 27, 28

Νᾶνις fab 30

Cornucopia 24, 40

Νικαγίς fab 2

Grape 43, 45

Νικασαγόρας 1st ep 1, 24

Head of Helios (see Helios)

Νικόμαχος ep 44, 45

Helios 4, 9, 12

Νύσιος fab 31

Incuse stamps 49, 50

Ξενοφάνης ep 6

Lozenge 34, 35, 36

Ὄλυμπος fab 1, 12, 21-22

Monogram 45

Πάναμος month 20, 44

Rhodos (statue of) 31

Παυσανίας 2nd ep 1

Rose 16 (s.s.), 17, 19, 21, 22, 32, 33, 45

Παυσανίας 3rd ep 33

Rouletting 50

Πολυάρατος 2nd ep 40, 41

Secondary stamp 16, 17, 18, 33, 40

Πρατοφάνης ep 28

Star 13, 15

Ῥόδιππος fab 21

Statue 31 (of Rhodos)

Σμίνθιος month 5, 9

Torch 21, 22

Σύμμαχος ep 16

Wreath 5, 6, 11, 25, 26

Σώδαμος ep 4 Σωκράτης ep 47 Σωκράτης 2nd fab 1, 12, 16, 21 Σωσικλῆς ep 21, 27, 33, 47 Σωσίφιλος fab 40 Τεισαγόρας ep 32 Τιμασαγόρας ep 6, 10, 11 Τιμόδικος ep 35 Τιμόθεος ep 39 Τιμόξενος fab 17 Φιλαίνιος fab 5-8, 10, 11 Φιλόδαμος 2nd ep 6, 11 Φρούραρχος title Kn 48 Ὑακίνθιος month 5, 35, 36

180

Chapter 5

Maritime Depictions on Ceramics from the Akko Marina Zaraza Friedman (1), Ehud Galili (2) and Baruch Rosen (3) 1. University of Haifa 2. Israel Antiquities Authority and Zinman Institute of Archaeology, University of Haifa 3. Israel Antiquities Authority Key words: marine archaeology, Acre, Roman ceramics, terra sigillata plate, Roman anchors, bowl

Fig. 1: Location map with the sites mentioned in the text (drawing: Z. Friedman)

Introduction Akko is a classical example of a historic fortified (most of the time) Mediterranean port city that has been functioning for thousands of years as one of the main centers of maritime activity in the Eastern Mediterranean (Fig. 1). The Akko Marina Archaeological Project revealed numerous artifacts from the Roman period (Galili this volume a). The finds include remains of a wooden hull of a medium-sized ship (20–80-ton capacity), a few coins, numerous broken storage

jars and domestic vessels. These include bowls and western terra sigillata ware, pottery that is found only rarely in the Roman ceramic repertoire of inland sites (Silberstein et al. this volume). Two of the ceramic artifacts recovered are decorated with maritime themes. One is a terra sigillata plate bearing the graffito of a ship and a Greek monogram, and the second is a deep bowl decorated with engravings of Roman iron anchors. The description of the artifacts and the maritime motifs depicted on them will be discussed below together with their interpretations. 181

Zaraza Friedman, Ehud Galili and Baruch Rosen

The Plate The plate (Figs. 2, 3, 4; 21cm rim diameter) assigned to the eastern terra sigillata pottery group is completely coated with red slip and has an elevated ring base. A pattern of six palmette leaves bordered by a double rope-like circle is stamped on the inner bottom of the plate (Fig. 2). A ship graffito, a tree and a Greek monogram are incised on the external bottom of the plate within the ring base (Figs. 3, 4; 10cm, inner diameter of the ring base). These types of terra sigillata ware appeared in Paphos, Antioch and the Eastern Mediterranean around 150 BCE and lasted until the 2nd century CE (Silberstein et al. this volume). The Maritime Graffito on the Plate

Fig. 2: Terra Sigillata plate with the maritime scene (drawing: IAA)

The graffito on the plate shows a ship associated with a tree and a Greek monogram (Figs. 3, 4). The scene is depicted in two registers: In the upper register, the monogram, formed by inscribing the letter Α into the letter Π (PA), is found at an angle of ca. 120 degrees at the upper left side of the rim base, and the tree is set almost in the upper center. The ship is found in the center of the lower register. The tree has a tapered trunk narrowing toward the base of the crown which is formed by crescent branches. It is slighgtly tilted to the left. The trunk , slightly tilted to the left, is indicated by three almost vertical lines, with the middle line closer to the left line, probably intended to give some perspective to the trunk. An angled line, with its upper end attached to the left side of the trunk, stretching toward the ship, probably indicates the mooring line (Figs. 3, 4). A second angled line stretching downward from the tip of the stempost may represent the anchor line. The ship is depicted in profile below the tree and the monogram, thus revealing its starboard side. The concave stem is followed by a projecting, pointed forefoot at the water level; this element is not a ram, but rather is the cutwater. The hull narrows slightly from amidships toward the prow. The gunwale is indicated by an almost straight horizontal line. A raised inner curved spar, resembling a ram's horn, forms the stempost. The stern curves at the chine and then rises in a vertical line. The sternpost is formed by a decoration resembling an upraised palm branch with open leaves only on the right side of the spar (symbolic aphlaston ?, fan-like with open branches). The horizontal line marked alongside the middle starboard hull may indicate the wale. The righthand end projecting above the pointed cutwater may indicate a pointed proembolion (upper ram), which was misplaced (Figs. 3, 4). The keel is depicted by double lines on the aft part of the bottom. The ship lacks its rigging (sail, mast, yard) row oars and the steering oars. Row oars were used when the winds were not favorable for sailing or when the vessel entered/left an anchorage or harbor.

Fig. 3: The Akko terra sigillata plate (IAA)

Fig. 4: Drawing of the ship graffito, the monogram and the tree (drawing: Z. Friedman) 182

Maritime Depictions on Ceramics from the Akko Marina

Discussion

The Bowl The deep, small bowl (9cm in diameter at the rim) (Fig. 5) is partly covered by black to gray slip (due to erosion and submersion over time); the fabric is brown with yellow mica. About 25% of the bowl’s side is missing. The side profile is angled and it has a ring-shaped base. Three iron anchors are engraved on the outer sides of the bowl, the fourth anchor is engraved on the bottom, within the ring base (3cm inner diameter). A fifth anchor was probably depicted on the broken missing part of the bowl. This bowl was dated roughly to the Roman period (Silberstein et al. this volume).

The Port of Akko during the Roman Period The ceramic finds recovered from the Akko Marina, along with numismatics and ancient written sources

Fig. 5: The bowl with the iron anchors graffiti (IAA)

The Anchors Graffiti on the Bowl The incised decorations on the bowl represent Roman iron anchors with stocks. One anchor is depicted with a ring on the upper part of the shank, through which the anchor line was passed. The arms of all four anchors are straight and are attached to the lower end of the shank in a sharp angle (ca. 40–60 degrees). The stocks of two anchors are slightly bent. Two stocks are inserted into the shank symmetrically on their central axis (the anchor on the top and the one on the bottom within the inner ring base), while the other two stocks are inserted asymmetrically (the left-hand anchor and the one on the lower side of the bowl). Similar iron anchors were recovered from several shipwrecks off the Israeli coast; e. g., Ashkelon (Fig. 6; Galili et al. 2001), and Neve-Yam (Galili et al. 1998). They were dated to the Early Roman period, from the 2nd century BCE to 1st century CE (Kapitan 1984). Numerous Roman iron anchors of all types were recovered from the Mediterranean, as well as from inland European waters (Bockius 2000). The anchors depicted on the small bowl (Fig. 5), were probably incised by the owner, probably a crew member familiar with ships, the equipment on board and anchors, or by passengers who sailed frequently and were observant of such details.

Fig. 6: A Roman iron anchor from a shipwreck off the Ashkelon coast (drawing: S. Ben-Yehuda) indicate that during the Roman period the harbor enjoyed widespread trade relations with many Mediterranean countries. The rich variety of the sherds attests to the luxury imports brought to this port that were meant to serve a society of mostly wealthy consumers (Zilberstein et al. this volume). These finds are supported by depictions of maritime symbols and harbor buildings on Roman coins minted in Akko, as well as by written references mentioning the importance of the Akko Harbor as a trading center and a military port. The finds recovered from the Akko Marina indicate that during the Roman period the main maritime activities were concentrated in the western basin,which was protected by the southern breakwater and could provide shelter for mediumsize merchantmen with a draft of 2.5m (20–80-ton load capacity). Half the sherds recovered from the seabed were identified as Roman amphorae

183

Zaraza Friedman, Ehud Galili and Baruch Rosen

on Roman ships from the 1st century BCE, and became a distinct feature as a stylized in-turned volute of Roman merchantmen and fishing boats as late as the fourth century CE (Basch 1987: 422, fig. 908; 425, fig. 913; 440, figs. 962–963; 475, fig. 1073; 489, fig. 1117; Friedman 2011: 140, fig. 3.8.7). The Akko ship graffito may represent a small to medium-size merchantman (20–80-ton capacity), or an auxiliary warship carrying supplies and/or troops along the Eastern Mediterranean coasts. The end of the anchor line is tied around the top stempost (misplacement; the line should have been tied to the cathead on the prow, or around the root of the stempost), whereas the lower end of the line was tied to a submerged anchor (not depicted in the scene). The mooring line tied to the tree trunk may illustrate one paragraph of the Roman law concerning sailing and fishing in seas and rivers. Gaius1 in his juridical work Everyday Matters or Golden Words, Book 2 wrote: “Everyone is at liberty to run boats aground the shores, to tie ropes to the trees rooted there…” (Digest, Book 8.5). This paragraph indicates that when a vessel found a shelter close to the shores, the sailors were free to tie the mooring line to the trees growing on the shore, because Roman law stated that the sea, the seashore and the rivers belonged to everyone (Marcian, Institutes, Book 3; Digest, Book I.8.2).

originating in the Aegean Sea and the Black Sea; one fifth were from Italy and the Western Mediterranean; 16 % from the coasts of Syria and Palestine, and 10% from North Africa (Galili et al. 2010; Silberstein et al. this volume). The Ship Graffito Although the Akko ship graffito is quite schematic it shows several distinct elements associated with the hull construction: the stern, the projecting pointed cutwater and the distinct stem and stern posts (Figs. 3, 4). The projecting pointed cutwater is a constructional element that is the forward extension of the keel, and was meant to improve the vessel’s hydrodynamics. The distinct wale alongside the middle section of the starboard hull reinforced the hull longitudinally and also provided protection when the vessel was moored with its long side to the quay. The hull shape of the Akko ship graffito is comparable to a similar graffito from Delos, in Insula II – Maison G, dated to the first century BCE (Basch 1987: 374, fig. 804/14). The prows of the Akko and Delos ships are similar, both ending in a pointed cutwater and concave stem. A sternpost formed by an aphlaston crossed by a stylis is a characteristic element of HellenisticRoman warships dated from the sixth to the second centuries BCE. The sternpost decoration on the Akko ship (four protruding, angled, short lines on the right side of the vertical spar), and a few vertical lines depicted on either side of the Delos ship sternpost, may represent a schematic aphlaston. The sternposts of Roman warships dating from the third century BCE to the fourth century CE are adorned only with an aphlaston and are missing the crossing stylis. The aphlaston became a distinct pattern on Roman merchantmen from the first century CE onward (Friedman 2011: 153, fig. 3.8.25). The sternposts of Roman fishing vessels dating from the third to the fourth centuries CE were also adorned with an aphlaston (Basch 1987: 488, fig. 1115; 489, fig. 489; Blanchard-Lemée et al. 1996: 123, fig. 82; 128, fig. 87). Some aphlastons have five branches (Basch 1987: 429, figs. 928, 930) similar to the Akko ship's aphlaston. The stem posts in the graffiti from both Akko and Delos, resembling a ram's horn, are similar (Figs. 2, 3; Basch 1987: 374, fig. 804/14). Earlier appearances of a ram’s-horn stempost on wooden ships is found on Greek geometric vases beginning in the second half of the ninth to the early eighth centuries BCE (Basch 1987: 166, figs. 334; 177, fig. 372; 188, fig. 394). The ram’s-horn stempost continued to appear

The Tree Graffito A tall, slim tree growing beside the waterline provides the proper device to tie the mooring-line stretching from the stern of the ship (Figs. 3, 4). The earliest examples of a ship associated with a tree or part of it (probably used to tie the mooring-line) are known from Minoan seals dated to the Bronze Age, and are displayed in the Heraklion Museum (Basch 1987: 99, figs. B4a, B5), and the National Museum of Athens (Basch 1987: 102, fig. D6). The ship in these seals is near a palm tree or branches facing the prow and the stern, or only one end of the vessel. The anchoring/mooring type depicted on the Akko plate is not typical of the Israeli coast, which is shallow and sandy. Rather, such anchoring is typical of the southern Anatolian and the Greek shores or their numerous islands, as well as in small anchorages in the northeastern Mediterranean or the Aegean Sea. The topography of these areas is typified by an indented coastline with many bays and deep, navigable waters reaching steep rocky shores on which Mediterranean trees grow down to the water line. Such anchorages provide good shelters, where a

1 Gaius (110–180 CE) was a celebrated Roman jurist. His works were probably composed between 130 and 180 CE, and became most important sources of Roman law. Many quotations from the works of Gaius occur in the Digest, created by Tribonian at the direction of Justinian I. The Digest (or Encyclopaedia) is the most important part of the Corpus Juris Civilis, whereas other works were the Code, the Institutes and the Novellae (Mommsen et al., 1985: xl).

184

Maritime Depictions on Ceramics from the Akko Marina

Fig. 7: Double-line mooring at Pilloried Cove, in Skopia Limani, southern Turkey (modified after Heikell 1997: 217) sailing vessel could moor by tying the bow or the stern to one or two trees on the shore and dropping anchor at the opposite end of the vessel. Such a double-ended mooring system of running lines from the stern and the stempost, one to the anchor, and another to a tree on the shore is practiced today by yachts anchoring in such bays in Turkey and the Aegean islands (Figs. 7, 8; Heikell 1997: 172, 217). Given that sea level and the general characteristics of the southeastern European coastline have not drastically changed in the last 4,000 years, the iconographic representations attest that such a double-mooring method has been used in the eastern Mediterranean since the second millennium BCE. The Akko ship graffito depicting a mooring system unknown in Akko, and in Israel in general, indicates that the ship originated elsewhere in the northern Mediterranean. Thus the plate with the maritime graffito graphically demonstrates the movement of ships and seamen in the Roman period. Probably, the plate belonged to a crew member who sailed in the depicted ship and was accustomed to use the doublemooring method. The Greek Monogram The Greek two-letter monogram may be related to the name of the person who possessed the plate, e. g., the ship’s owner, or to the name of the ship or its geographical origin, such as Paphos or Patara (Fig. 1). ΠΑ (PA) monograms are also found on coins minted by the city of Acre-Ptolemais in the second century BCE (Hill 1910: 128). Another possible interpretation of the monogram may be related to the third journey of Paul the apostle from Ephesus to Jerusalem (53–57 CE), with an overnight stay at Acre-Ptolemais. This journey is described in Acts 21: 1–15, specifying that Paul embarked on a cargo ship at Patara whose destination port was Tyre:

Fig. 8: Recommended double-line mooring at Tarsane, southern Turkey (modified after Heikell 1997: 172) “… we came with a straight course onto Cos, and the day following Rhodes, and from thence unto Patara: And find a ship sailing over unto Phoenicia, we went aboard, and set forth” (Acts 21: 1–2). After a week’s stop at Tyre, Paul embarked on another ship with an overnight stop at Acre-Ptolemais: “And when we had finished our course from Tyre, we came to Ptolemais, and saluted the brethren, and abode with them one day” (Acts 21: 7). It is tempting to speculate that the monogram ΠΑ (PA) may be associated with Paul’s visit to Acre-Ptolemais. This hypothesis can be neither proven nor negated.

Summary The Akko ship graffito represents a small or medium size seagoing vessel mooring where trees were growing near the waterline. Such conditions are not found along the Israeli coast, whereas anchorages with indented coastline bays with deep water and steep rocky shores may be, found in Turkey, Cyprus or Greece and the islands in the eastern Mediterranean or the Aegean Sea. The prow of the Akko ship with the projecting cutwater may be compared to the rare find of the prow of the Pisa Wreck C, dated to the first century CE (Fig. 9; Friedman 2011: 175, fig. 4.17). The projecting pointed end of the wale above the pointed cutwater of the Akko ship may suggest the schematic proembolion, or it may have resulted from the inaccuracy of the artist. The palm-branch decoration of the sternpost may indicate the symbolic aphlaston adornment. The Akko plate with the ship graffito may be dated either in the first century BCE or in the first century CE, considering the comparison to the ship

185

Zaraza Friedman, Ehud Galili and Baruch Rosen

References Basch L. 1987. Le musée imaginaire de la marine antique. Institut Hellenique Pour La Preservation De La Tradition Nautique. Athens. Blanchard-Lemée M., Ennaifer M., Slim H. and Slim L. 1996. Mosaics of North Africa: Floor Mosaics from Tunisia. London. Bockius V.R. 2000. Ein Römischer stockanker aus Trajans Donaukanal bein eiserner tor, Serbien, Ärchaologisches korrspondeenzblat 30: 97–116. Friedman Z. 2011. Ship Iconography in Mosaics: an Aid

to Understanding Ancient ships and their construction.

Fig. 9: The pointed cutwater of Pisa Wreck C (photo: A. Camilli)

(BAR International Series 2202). Oxford.

Galili E., Sharvit J. and Nagar A. 1998. Nevé-Yam – Underwater Survey. Excavations and Surveys in Israel 18: 54–56, 35–36. (Hebrew/English).

graffito from Delos and the typology of the terra sigillata plate. The Greek monogram may indicate the name of the plate's owner or it may suggest the geographical origin of the crew or ship. The geographic origin of the bowl with the iron anchors graffiti cannot be identified with certainty. This type of anchor is supported by many archaeological finds, e. g., Neve-Yam, Ashkelon, or other sites in the Mediterranean. Both ceramic items discussed above attest to the movement of ships and objects to Levant in the Roman period. The anchors depicted on the small bowl were probably incised by a crew member of the ship which carried the bowl. The maritime graffiti on both ceramic items were engraved after manufacture by the people who owned them, probably by crew members familiar with ships, the equipment on board and anchoring practices, or by passengers who sailed frequently and were observant of such details. The Akko plate has a Western origin, and is depicting an anchoring method used in the southern coasts of Turkey or Greece. The plate may have belonged to a crew member or a passenger who sailed from the southern coast of Turkey to the eastern Mediterranean, similar to Paul’s sailing, and anchored at the Akko Port, where he lost the plate, or discarded it when it broke.

Galili E., Sharvit J. and Dahari U. 2001. Ashkelon and the Sea in Light of the Archaeological Coastal and Underwater Finds. In A. Sasson, Z. Safrai and N. Sagiv (eds.). Book of Ashkelon. Pp. 11-38. Jerusalem. (Hebrew, English summary). Galili E., Rosen B., Zviely D., Silberstein N., Finkielsztejn G. 2010. The Evolution of the Akko Harbour and its Trade Links Revealed by Recent Underwater and Coastal Archaeological Research. Journal of Island and Coastal Archaeology 5: 191–211. Galili E. This volume a. The Akko Marina Archaeological Project – Introduction. Heikell R. 1997. Turkish Waters and Cyprus Pilot, St. Ives, Cambridgeshire. Hill G.F. 1910. Catalog of Greek Coins, Phoenicia. P. 128. London. Kapitan G. 1984. Ancient Anchors – Technology and Classification. International Journal of Nautical Archaeology 13: 40–52. Mommsen T., Kreuger P. and Watson A. 1985. The

Digest of Justinian, Vol. I. Philadelphia.

Acknowledgments

Silberstein N., Galili E. and Sharvit J. Hellenistic, Roman and Byzantine Ceramics, this volume.

The authors wish to thank The Israel Antiquities Authority for the institutional support, M. Feinberg Vamosh for the English editing, D. Jacoby for his useful remarks on the manuscript and S. Ben-Yehuda for some of the drawings.

186

Chapter 6

Ship Timbers from the Akko Marina: Early Use of Unpegged Mortise-and-tenon Joints and Mixed Shell-and-frame Construction Yaacov Kahanov* Leon Recanati Institute for Maritime Studies, University of Haifa, Israel Key words: Mortise-and-tenon, frame-based construction, Acre, ship construction, bronze nails.

Abstract Several ship timbers including fasteners were retrieved from the seabed during maintenance dredging works at Akko Marina. The nature of the work resulted in broken, distorted and out-of-context wood pieces. C14 test of the wood, together with some construction features, dates the vessel to which these timbers belonged between the third and the fifth centuries CE. The analysis reveals the use of both relatively closely spaced unpegged mortise-and-tenon joints and some form of a ‘frame-based’ constructed hull. These are therefore among the earliest evidence for the use of this combination of shipbuilding techniques in the Eastern Mediterranean.

Introduction and Methodology About 25 ship timbers containing fasteners in the form of mortise-and-tenon joints, and wood and metal nails, were retrieved from the seabed of the small modern marina of Akko (Acre) in 1992–1993 (Galili et al. 2002; 2010; Galili and Sharvit 2002; Galili and Rosen 2008; Galili and Rosen this volume e).

Akko Old City

Fig. 1: The Akko Marina, the dredging area and the place of the timbers (S. Ben-Yehuda, E. Galili)

They were raised by a dredger that was deepening the marina and discharged with other material into a barge, which unloaded them off-shore. The work was supervised by representatives of the Israel Antiquities Authority (IAA), which collected the material on the barge and on the sea bottom where it was dumped. The first quick field-check of the timbers was conducted on the barge, while several pieces were taken for further analysis at IAA headquarters, where they were studied in their waterlogged state without conservation. The timbers were broken, damaged and out of context. The finds were removed from Areas D and E, which are located in the central part of the marina, at a water depth of 2–3m (Fig. 1). Despite the poor condition of the timbers, their analysis reveals interesting information regarding construction details, which may shed more light on the development of ship construction during the midfirst millennium CE in the Eastern Mediterranean.

Catalogue The following is a concise list of the significant timbers that were selected out of about 25 pieces. Hull Planks Seven heavily damaged pieces were classified in this group: a main timber and six additional small broken pieces. Four of these were easily restored into two pieces that could be positioned in relation to the main piece. The other two were apparently also parts of the main timber. The wood was worked along the grain from an outer section of a cypress trunk.1 Timber A/1, the main timber (Fig. 2): Despite its distortion, the wood itself, identified as cypress, was well preserved. It was 1.38m long, 14.0cm wide (maximum) and 4.0cm thick. Although no framing timbers survived, evidence for plank-frame fasteners

* Prof. Y. Kahanov passed away in December 2016

187

Yaacov Kahanov

projected about 6–7mm above the plank surface. No evidence for the use of plug treenails was found. The remains of unpegged mortise-and-tenon joints were found in this plank (Fig. 5). Three mortises were spaced 26.0 and 23.0cm apart between centers. One of the mortises, 5.7cm wide, was found undamaged. The mortises were 0.6–0.8cm thick and about 4.7cm deep. A section of oak tenon survived in one of the mortises. It was 5.3cm wide and 3.5mm thick, and had a maximum preserved length of 3.5cm. As noted above, it was not locked in its mortise with a tapered peg.

Fig. 2: Timber A/1, the longest plank retrieved (IAA) and traces of five frames altogether were found. Of the latter, two were staining patterns. Remains of a yellow protective material that was spread onto the plank surface were discernible in the spaces between the frames. These remains indicate that the frames were sided approximately 7.0cm. The center-tocenter distance between frames was between 16.0 and 27.5cm, 21.8cm on the average. Frames and planks were fastened together by bronze2 nails and treenails

Fig. 5: Unpegged mortise-and-tenon in a section of timber A/1 (Y. Kahanov) Timbers A/2, two separate sections of the main timber: These two sections were torn away from the main timber. In both, the round holes for treenails, frame staining patterns, and yellow protective material remained. One of the timbers was a maximum of 49.0cm long, 3.2cm wide and 3.2cm thick. The second timber was 48.0cm long, 3.6cm wide and 2.0cm thick. These timbers were refitted to the main timber in their original positions, after which four mortise-and-tenon joints were identified. The center-to-center distances between the mortises were 12.5, 11.8 and 12.0cm. The preserved mortises were 5.9, 4.9 and 5.6cm wide, 0.6–0.7cm thick, and they tapered toward the inner edge. No tapered pegs were evident in these timbers, but in this case their absence could be a result of the breakage of the planks. It is possible that such pegs were originally used to secure tenons closer to the plank edges, but these parts were not evident. Remains of treenail holes 1.7cm in diameter were found in these pieces, but no treenails survived.

Fig. 3: Timber A/1: protective material, frame staining, wood and bronze nails (IAA) (Fig. 3). Their remains survived in the planks as pairs of nail and treenail spaced 10.0, 4.2 and 9.0cm apart. Two treenail holes were 1.7cm in diameter, and a third one was 1.2cm in diameter. In the latter hole the remains of a rectangular cross-section treenail were found. It tapered from 1.2 × 1.2cm on the outside of the plank to 1.1 × 0.9cm on the inside. The bronze nails were found straight and had a square crosssection (Fig. 4). They tapered along their shanks, and in several places the side measured 4mm. The nails had round heads between 1.4 and 1.7cm in diameter (top view), and a trapezoidal cross-section (side view)

Timbers A/3, four additional timbers: Four additional plank pieces were examined; their details (maximum dimensions) are presented in Table 1. Thick Timbers This group comprises six relatively thick timbers. When compared to the pieces of Group A, they were thicker, slightly darker, with larger grain, and in a poorer state of preservation. The wood was very soft and in part was damaged by teredo. Timber B/1 (Fig. 6): This part was 22.0cm long, 18.0cm wide and 5.5cm thick. On one of its sides the remains of four metal nails driven in from the outside were found; they were 0.3cm thick, between 1.2 and

Fig. 4: Bronze nail (IAA) 188

Ship Timbers from the Akko Marina

The largest three pieces, the richest in information, are presented here. Timber C/1 (Figs. 7a, 7b, 7c): This timber was 63.0cm long and 21.0 × 15.4cm in maximum crosssection. It was cut from one side of a soft tree close to the pith and was worked from all sides. The upper side tapered from a maximum of 21.0cm to both edges. It was transversally and longitudinally convex. The opposite side was flat with a longitudinal straight incise. Wide grains followed its longitudinal axis. On one of its sides the remains of 12 nails and treenails were evident. Treenails and nails appeared in clusters. The details of the fasteners on the inside with maximum features are presented in Table 3.

Fig. 6: Timber B/1, a thick timber with bronze nail (IAA) 2.0cm long and spaced 4.0cm apart. The inner edge of a mortise, 6.2cm wide, 0.7cm thick and 1.0cm deep, partially survived. On the opposite side the trapezoidal edge of another mortise was found. It was 6.0cm wide, 0.8cm thick and 5.0cm deep. Assuming the mortises were originally about 6.0cm deep, the width of the plank can be estimated at 24cm. Part of a tenon was found in one of the mortises. It tapered both in width and thickness, and its maximum dimensions were 5.1cm wide and 0.5cm thick, with a preserved length of 4.3cm. No tapered pegs were evident. A bronze nail enveloped with concretion was found in this timber. The cross-section of the inner core of the nail was 4 × 4mm square, and its outer envelope was also square in cross-section, 1.1 × 1.1cm. The diameter of the concreted head was 2.8cm. No plug treenail was used before the metal nail was driven.

Timber C/2: This timber was 55.0cm long, and its cross-section was 15.0 × 16.0cm. The transversal cross-section changed along its length and tapered from the middle toward the ends. Longitudinally it

7a

Timber B/2: This timber was a broken piece of plank from the hull of a vessel, with remains of unpegged mortise-and-tenon joints. It was 18.0cm long, 8.5cm wide and 5.2cm thick at one edge and 4.6cm thick at the other edge. On one of its edges two mortises were visible: One was 6.5cm wide and 0.8cm thick, with remains of a tenon 5.8cm wide and 0.7cm thick. At a distance of 12.5cm between centers a second mortise was evident, but in a poor state of preservation and impossible to measure. The two mortises were staggered in the plank’s thickness. On the other side of this timber another mortise survived, 6.0cm wide and 0.7cm thick, with part of a tenon inside that was 5.5cm wide and 0.5cm thick. A hole for a treenail 1.6cm in diameter was also found in this plank.

7b

Timbers B/3: Four additional pieces were found. Their characteristics are detailed in Table 2.

7c

Thick, Irregular Timbers About ten irregular pieces of timbers were found. They are difficult to classify and their use is unclear.

Figs. 7a,7b,7c: Timber C/1, thick, irregular timber with bronze and wood nails (drawing S. Ben Yehuda; photo Y. Kahanov and IAA) 189

Yaacov Kahanov

Pomey et al. 2012) was extensively preserved and contributes significant information for comparison with the timbers from the Akko Marina. However, the evaluation will refer to the hull only, and not to the inner components. It was dated to the end of the second century CE (and see Pomey, et al. 2012: 242). The ship was constructed ‘shell first’ with mortiseand-tenon joints. Planks were generally 2.5cm thick, except for the garboards, which were 4.5cm thick at their keel face, and 14.0–30.0cm wide. Tenons were 6.0cm wide, 1.0cm thick, 12.0–13.0cm deep, and spaced at 12.0cm. Tapered pegs 0.9–1.1cm diameter locked tenons within their mortises. Frames were connected to the existing hull by treenails 1.5cm in diameter. Additional reinforcement for the garboard (to keel and frames), second and up to the third strakes, with copper or bronze nails, was also applied.

followed the grains of the wood, but was worked asymmetrically. On one of its sides (side 1) the remains of nine nails and treenails were recorded (Table 4). Saw marks and adze traces were visible on this surface. On the opposite surface (side 2) the remains of three nails were found (Table 5). On the third side one treenail 1.6cm in diameter was found. A yellow material, apparently for protecting the wood or for sealing it, remained on the surface of the fourth side. Timber C/3: This was a tapered piece of soft wood with large grains and some teredo and barnacle remains. It was 29.0cm long, 16.0cm wide and 12.0cm thick. On one of its sides four treenails and three bronze nails were found. The treenails were 1.8, 2.1, 1.7 and 1.6cm in diameter. Of the last, only the hole was found without a treenail. These treenails were actually multifaceted, almost cylindrical. Two of them completely penetrated the timber. One of the bronze nails was concreted, and was thus unavailable for measuring. The second was broken, with a square inner core, 0.7 × 0.7cm. The third was 12.7cm long, with a square cross-section 0.7 × 0.7cm. Its head was 1.7cm in diameter. Several more small pieces of timbers were found, but they presented no significant details.

La Bourse (Lacydon), Marseille (Gassend 1982; Gassend and Cuomo 1982, 1985; Pomey 1988: 407–408, 1998: 68, 2004: 31–32; Gassend 1989; Rival 1991: 245–265): These ship remains were dated to 190–220 CE. The planks were 18.0–23.0cm wide and 6.0cm thick, and were mortise-and-tenon joined. The garboards were attached to the keel by pegged mortise-and-tenon joints. The mortises were 6.0cm wide, 1.0cm thick and 10.0cm deep. The space between edges of adjacent mortises was 12.0–15.0cm. Tenons were 6.0cm wide, 1.0cm thick and 12.0–13.0cm long, and were spaced 20.0cm center-to-center. They were locked in their mortises by tapered pegs of 0.8 to 1.5cm diameter. Frames and planks were connected by treenails 1.5–2.0cm in diameter. Copper bolts 2.0cm in diameter connected every eighth to tenth frame to the keel.

Summary of the Fastener Elements The main significant fasteners features mentioned above include at least 39 treenails, with an average diameter of 1.68cm and fitted within their holes. The other fasteners features are summarized in Tables 6 and 7.

Discussion The timbers from the Akko Marina were probably parts of vessels that were torn away during the dredging operations. C14 tests of wood samples gave dates from 147 to 540 CE.3 The preferred assumption here is that these were ship’s timbers and probably not, for example, harbor installation components. Evidence for this view is to be found in the fact that several wood types were identified – including soft and hard woods, some curved, and others shaped to different thicknesses – and the connection techniques were those used in shipbuilding. At least three connection methods were used: mortise-andtenon joints, treenails and bronze nails. In addition, another timber that was retrieved during the dredging from the same location (not reported here) can be attributed to a later, Ottoman-period ship’s frame. Below is a concise summary of relevant construction components from shipwrecks for comparison.

Torre Sgarrata (Casson 1995: 216; Fitzgerald 1994: 182; Parker 1992: 429; Throckmorton 1969, 1985): This wreck is dated to the end of the second century/beginning of the third century CE. Hull planks were on the average 22.0cm wide and 7.2cm thick. They were fastened to each other by mortise-andtenon joints locked by tapered pegs. Mortises were 10.4–14.5cm wide, 1.0–1.2cm thick and 8.0–10.0 (averaging 8.3)cm deep. Mortises were spaced at 2.8–8.7cm edge-to-edge, averaging 3.5cm, giving a center-to-center spacing of about 16.0cm. Tenons were 12.0cm wide, 1.1cm thick and 16.7cm long. The pegs were 1.1–1.4cm in diameter. Frames were fastened to planks with treenails 2.0cm in diameter. Copper nails were also used, but in scarfs and edges only. Several iron nails were found in repairs of the hull.

Laurons 2: Laurons 2 (Gassend et al. 1984: 75–105; Pomey 1987–1988: 23–25, 1988: 406–407, 1992: 24–25, 2004: 32; Ximénès and Moerman 1991;

Monaco A (Benoit 1961: 146; Mouchot 1968– 1969; Basch 1972: 48–50; Parker 1992: 279): The shipwreck was dated to the end of the second century

190

Ship Timbers from the Akko Marina

or first half of the third century CE. It was probably small, not more than 15 m long. Garboards were tenoned to the keel, their mortises were 4.0–6.0cm wide, 0.6cm thick, and spaced 6.0–7.0cm between the edges of adjacent mortises. Based on a sketch of the wreck (Mouchot 1968–1969: 201, pl. XV) the center-to-center distance between tenons can be estimated at 12.5cm. Tenons, which were 0.5–0.6cm thick and 7.0–8.0cm long, were locked in their mortises by tapered pegs. Planks 18.0–27.0cm wide, 3.0–4.0cm thick were mortise and tenon joined. Frames were connected to the hull by treenails. Bronze nails fastened ceiling planking, and possibly also reinforced the garboards-keel connection. Pommègues (Gassend 1979): This shipwreck was dated to the second half of the third century CE. It was constructed ‘shell first’ with mortise-and-tenon joints. The mortises were 5.0–7.0cm wide, 0.7cm thick and 5.0cm deep. Tapered pegs 1.1–1.2cm in diameter secured the tenons in their mortises. They were spaced 14.0–15.0cm center-to-center. Frames were connected to planks by treenails 1.5–1.8cm in diameter, about 15.0cm apart. County Hall (Riley and Gomme 1912; Marsden 1972, 1974, 1994): This shipwreck was dated to ca. 300 CE. Although built in Britain, it was constructed in the Mediterranean tradition. Planks were 26.7–38.1cm wide and 5.1–7.6cm thick. They were fastened to the keel and to each other by mortise-andtenon joints locked by tapered pegs. Tenons were 6.4cm wide, 0.7cm thick and 12.7cm long, and the pegs had a diameter of 1.6cm. The distances between the edges of adjacent mortises varied from 15.2 to 97.0cm; one mortise was 11.0cm wide, 0.7cm thick and 6.5cm deep. Frames were attached to the hull by treenails 3.2cm in diameter. Iron nails were also used, but for limited purposes. Point de la Luque B (Joncheray 1972; Clerc and Negrel 1973; Liou 1973, 1975; Negrel 1973; Guibal and Pomey 2009): This wreck was dated to the fourth century CE. Planks were 15.0–23.0cm wide and 3.0cm thick. They were connected to each other and to the keel by mortise-and-tenon joints locked by tapered pegs. The mortises were 6.0cm wide, 0.7cm thick and 4.0cm deep. Tapered pegs 0.8–1.1cm in diameter spaced 10.0–12.0cm apart locked the tenons at the keel-garboard attachments. Frames were connected to planks by treenails 1.0cm in diameter. Yassi Ada 2 (Bass and Van Doorninck 1971; Van Doorninck 1976; Pomey 1988: 409; steffy 1994: 79– 80): This shipwreck was dated to the fourth century CE. The hull was built ‘shell first’ with hull planking varying in width from 11.0 to 25.0cm and 4.2cm thick, with the garboards 5.3mm thick. Planks were connected by mortise-and-tenon joints. Their center-

to-center spacing varied between 15.0 and 32.0cm. The mortises were 7.0–9.0cm wide, 0.7cm thick and 5.0–5.5cm deep. The tenons were 4.5cm wide and their full length from the two sides of the seams were 8.5cm. Tapered pegs 0.7–1.1cm in diameter locked the tenons in their mortises. Mortises and tenons were tapered, and the tenons occupied about 50–64% of the mortise width. Treenails 1.6cm in diameter connected frames to planks and iron nails were used to secure planks in sensitive places. Dramont F (Joncheray 1975, 1977): This small shipwreck was dated to the second half of the fourth century CE. It was constructed ‘shell first’ with mortise-and-tenon joints. Mortises in the keel were 5.0–7.0cm wide, 0.7cm thick and 4.0cm deep. They were 5.8–8.2cm wide in the planking. In the keelgarboard attachments they were spaced 15.0–23.0cm (between edges of adjacent mortises, thus about 25.0cm center-to-center), and in the planking the center-to-center distance was 19.5–25.0cm. Tenons were 3.5–5.5cm wide, 0.5–0.6cm thick and their full length was 6.0–7.0cm. They were locked in their mortises with tapered pegs, 0.7–0.8cm in diameter. The tenons, which tapered both in their width and thickness, were fairly loose in their mortises and occupied about 70% of the mortise. Frames were connected to the hull with tapered treenails 1.1– 1.4cm in diameter. Several metal nails were also found, probably for fatening internal components. Port Vendres 1 (Chevalier and Santamaria 1972; Chevalier and Liou 1974; Liou 1974; Pomey 1988: 408–409, 2004: 32; Rival 1991: 267–296): This wreck is dated to about 400 CE. Its length is estimated at 18–20 m. Some of the frames were iron-bolted to the keel; the bolts possibly extended farther inside, also connecting the keelson. The garboards were 6cm thick and fastened to the keel with mortise-and-tenon joints, and to the endpost with copper nails. Planks were 4cm thick and connected to each other by mortise-and-tenon joints locked by tapered pegs. Mortises were spaced 6–15cm apart. Average mortises were estimated at 7cm wide. Tapered tenons fitted well within their mortises although no dimensions were specified. Frames were attached to the planks by treenails. The significance of this wreck to this study is the combination of closely spaced pegged mortise-and-tenon joints, the presence of copper nails and the use of iron bolts all together in a 400 CE ship. Parco di Teodorico, Ravenna (Medas 1999, 2001, 2003): This is a Late Roman fifth-century CE wreck, found in Parco di Teodorico, Ravenna, Italy. The hull remains included keel, keelson, frames and planking. Unpegged mortise-and-tenon joints for the planking were widely spaced (ca. 80cm), with tenons occupying only part of their mortises. The garboard was simply fixed to the keel with iron nails and inside the stem rabbet. Some of the floor timbers were fixed to the keel

191

Yaacov Kahanov

with iron nails. Planks were connected to frames by iron nails and treenails. This wreck is a good example of the combination of a mixed transitional concept and technique that still employed plank-oriented principles and the ‘shell-first’ construction method, with frames providing the hull integrity, but not yet full skeleton construction.

st. Gervais 2 (Jézégou 1983; Carre and Jézégou1984; Jézégou 1985, 1989, 1998; Pomey 2004: 33, 1988: 410–411): Dated to the beginning of the seventh century CE, this ship is one of the earliest skeleton-constructed vessels in the Mediterranean, although some relics of mortise-and-tenon joints were present. The mortises were very few, in the extremities; some did not have a corresponding mortise in the adjacent strake and did not have tenons. In the others, tenons were unpegged. Planks were about 2.5–3.0cm thick. Neither pattern nor system could be identified in positioning the tenon locations; no doubt they were of secondary importance in the plank joints. Iron nails and treenails were used to fasten planks to frames and other hull components.

Tantura A (Kahanov and Breitstein 1995a; 1995b; Kahanov and Royal 1996; Wachsmann and Kahanov 1997; Kahanov 2001, 2011a, 2011b; Kahanov et al. 2004): This wreck, which was dated to the late fifth–early sixth centuries CE, had a frame-based hull. Frames were nailed to the keel with iron nails. The garboards were not connected to the keel, but just touched it. Planks were connected to frames by iron nails, and were joined to strakes by butt-joints at frame stations. No planking edge-fasteners of any kind were found anywhere in the hull remains, and seam caulking was identified.

Yassi Ada 1 (Van Doorninck 1972; steffy 1982; Van Doorninck 1982; steffy 1994: 80–85; Pomey 2004: 32–33, 1988: 409–10): This vessel is dated to about 625 CE. The strakes varied in width from 13.0 to 25.0cm and in thickness from 3.5 to 4.2cm. Unpegged mortise-and-tenon joints aligned the lower 16 strakes. The mortises were 5.0cm wide, 0.5cm thick and 3.5cm deep, tenon length was equal to the combined mortise depth, but they were only 3.0cm wide, thus occupying only 60% of the mortise width. Center-to-center mortise spacing was about 225cm along the garboards, 35–50cm in the stern area and about 90cm in the middle of the hull. Iron nails, spikes and bolts of different sizes fastened the planks and the frames as well as the other hull components. This vessel shows that the transition in construction from ‘shell first’ to ‘frame based’ was well underway by the beginning of the seventh century. The main strengthening of the hull was based on frames and longitudinal members, although the shipwright still employed the mortise-and-tenon joint for a secondary purpose.

Dor 2001/1 (Mor and Kahanov 2006; Kahanov and Mor 2014): This is the wreck of a coaster, whose original dimensions are estimated at 16.9 m long and 5.4 m wide. It was dated to the beginning of the sixth century CE. The framing pattern was of alternating floor timbers and pairs of half-frames. Almost all frames were nailed to the keel by iron nails 6 × 6mm square. The planks were 5.0–20.6cm wide, and 2.0– 3.3cm thick. The garboards were not connected to the keel, but were connected to the endpost. Planks were connected to frames by iron nails, and were joined to strakes by butt-joints at frame stations. Neither mortiseand-tenon joints, nor any other type of planking-edge joints, were found anywhere in the hull. Caulking was found in planking seams. This concept of the hull and its method of construction were based on frames.

Pantano Longarini (Throckmorton and Kapitän 1968; Throckmorton and Throckmorton 1973; Kampbell 2007): The wreck was dated to the early seventh century CE. Unpegged, widely spaced (up to 1m) mortise-and-tenon joints served to align the planks. The ship was probably built on mixed strakeand-frame-principle and methods. It is suggested that this hull construction demonstrates a stage in the transition from and abandonment of the ‘shell-first’ technique.

Dor 2006 (Navri et. al. 2013; Barkan et al. 2013): The shipwreck was discovered in 2006, 800 m south of the Dor (Tantura) lagoon, Israel. The hull remains included sections of large frames, stringers and ceiling planks, a large number of strakes, and five wales. The shipwreck was dated to between the second half of the sixth and the first quarter of the seventh centuries CE. Unpegged mortise-and-tenon joints were identified. Mortises were 3.8–6.3cm wide, 0.5–0.6cm thick, and 2.4–3.5cm deep. Hardwood tenons tapering in width and thickness were found in all the mortises. They were 2.6–3.2cm wide in the seams, occupying about 60% of the mortise width, 0.4–0.5cm thick, and 2.6–3.0cm long (half length). The unpegged tenons served to align the planks, which were later secured by frames and so the concept and construction sequence of the hull was shell-based. The upper part of the hull was based on frames. Thus as a whole this hull was built in a mixed shell-and-frame construction.

Dor D (Kingsley and Raveh 1996; Kahanov and Royal 2001; Kingsley 2002; Kahanov 2003; Royal and Kahanov 2005): The dating of this shipwreck is based on C14 results, which give a range of 350–621 CE. The maximum width of planks was estimated at about 25cm, the majority were 3.0cm thick. Unpegged mortise-and-tenon joints served to keep the planks aligned during the building of the hull. The mortises were, on the average, 6.3cm wide, 0.5cm thick, slightly over 3.0cm deep, and

192

Ship Timbers from the Akko Marina

spaced at an average of 28.6cm between centers, with a maximum of 62.0cm. Tenons were tapered: They were on the average 3.5cm wide and 0.3–0.5cm thick, thus occupying about 60% of the mortise width. Tenon lengths were equal to their combined mortise depths. All but one of the mortise-and-tenon joints were unpegged. A single 1.0cm-diameter peg passed through the center of a tenon. Although no framing timber survived, evidence for plank-frame attachments was found. Treenails 1.5cm diameter and iron nails 0.5 × 0.5cm in cross-section were both employed for frame-plank attachments. Dor D thus could indicate a shell-longitudinal concept for the shape, but a structural concept with mixed shellskeleton construction methods. Fiumicino vessels (Scrinari 1979; Parker 1992; Boetto 2000, 2003, 2006, 2008; Boetto pers. comm.): The main difficulty with the analysis of these shipwrecks is the uncertainty in their dating. Although 14C samples range from 92 BCE (Barca del pescatore) to 230 CE (Oneraria minore 1), associate finds led to a later dating, e.g. fourth–fifth centuries CE (Oneraria maggiore 1); second–fourth centuries CE (Oneraria minore 2); and second century CE (Barca del pescatore). However, the researchers of the vessels tend to ascribe them to the later dates based on the construction details. Significant details of several vessels from Fiumicino are given in Table 8. The information on the shipwrecks above, excluding the Fiumicino vessels, focusing on the mortiseand-tenon joints and nail and treenail fasteners, is summarized in Table 9. The vessel from the Akko Marina had frames about 7cm sided, spaced at about 22cm center-to-center. Despite the sparse evidence, the width of a plank of this vessel was estimated at about 24cm. Planks were of different thicknesses: Some were 2.0cm and others 5.5cm. This suggests that this was a medium-size vessel. McGrail (1995, 1997), Kahanov (2010) and Pomey et al. (2012) presented meticulous discussions on ship construction methods, and the transition in ship construction. Because of their fragmentary condition the wood remains from the Akko Marina are objectively difficult to analyze. However, these timbers revealed the use of relatively closely spaced (average 16.3cm) unpegged mortise-and-tenon joints. The planks were connected to the frames by treenails and a few bronze nails. These features point to an early stage of the transition and a hull of mixed construction. The metal used, as shown in Table 9, conforms to the dates of use of copper or bronze nails in early vessels (Laurons 2, Bourse, Torre Sgaretta, Monaco), and iron in later wrecks (County Hall, Yassi Ada 2, Parco di Teodorico, Tantura A, Dor 2001/1, Dor 2006, St

Gervais B, Yassi Ada 1, Pantano Longarini, Dor D). An early date for the Akko Marina timbers matches this classification. The change in the metal used for the nails is evident also in the Fiumicino vessels, despite the uncertainty in their dating. Copper nails were found in the (probably) earliest vessel (Fiumicino 5), while iron nails were used in the later vessels. The mortise spacing is somewhat confusing, since no exact correlation can be identified between it and the dates of the shipwrecks in Table 9. However, despite several exceptions, it can be seen that mortise spacing tends to increase over the years. The mortises of the Akko Marina timbers were relatively close to each other. Their center-to-center distance varied between 12.0 and 26.0cm, but they were consistent on each plank fragment: 23.0–26.0cm on the piece that was designated as A.1 above, and 11.8–12.5cm on A.2 and B.2. This suggests that mortises were cut in variable patterns along the hull. However, the Akko Marina mortises are well fitted with the closely spaced tenons, as in the earlier vessels represented in Table 9. The fraction of a mortise occupied by a tenon correlates well with the transition stage in shipbuilding technique from shell-first to some form of frame-based construction. The ratio (in percentages) of tenon width and thickness relative to the mortise corresponds well with the chronology. The vessels with a high ratio (above 90%) are earlier (La Bourse, Torre Sgaretta), while those with a low ratio (about 60%–70%) are later (County Hall, Yassi Ada 2, Dramont F, Dor 2006, Yassi Ada 1, Dor D). The tenons from the Akko Marina occupied a high fraction of their mortises – 93% of the mortise width, and 72% of the mortise thickness. The latter examinations date the Akko Marina ship among the early vessels to the third century CE. Unpegged mortise-and-tenon joints made their appearance at the end of the fourth or the beginning of the fifth century CE, as demonstrated by Fiumicino 1 and Parco di Teodorico. This technique later appeared systematically in hulls (Dor D, Dor 2006, St-Gervais 2, Yassi Ada 1 and Pantano Longarini). If a date at the end of fourth or the beginning of the fifth century CE is allowed for the Fiumicino 1 vessel, or a date between the second and the fourth centuries CE is allowed for the Fiumicino 4 vessel, this technique can be related to a slightly earlier date. Instead of giving a later date to the Akko Marina ship timbers, it is suggested that they be counted as evidence for the early use of unpegged mortise-and-tenon joints. Treenails for plank-frame attachments cannot be used for dating analysis as has already been shown (Kahanov and Royal 2001).

193

Yaacov Kahanov

References

Analysis by C14 dating of wood samples gave dates from 147 to 540 CE. There is a slight chance that all the timbers from the Akko Marina were broken along their edges, farther in than the tenon pegs, yet leaving the mortises and tenons. If this hypothetical breakage happened, it also broke the tenon that survived along 4.3cm of its length after the peg (see timber B.1 above). This length is longer than the pegged (half) tenons of the end of fourth century, as evident in the Yassi Ada 2 and the Dramont F shipwrecks, as well as in the earlier Monaco and in the later Yassi Ada 1 shipwrecks. If indeed this happened along all twelve mortises analyzed, it would mean that all tenon pegs have disappeared from the archaeological finds. Thus the retrieved timbers would not represent the original hull construction. If this is so, the above analysis cannot hold. However, this consistent line of breakage and the exceptional tenon length is very unlikely, although a parallel for relatively late use of closely spaced tenons and copper nails can be found in the Port Vendres 1 shipwreck.

Barkan D., Yasur-Landau A., Mommsen H. Ben-Shlomo D. and Kahanov Y. 2013. The ‘Dor 2006’ Shipwreck: the Ceramic Material. Tel Aviv 40: 117–143. Basch L. 1972. Ancient wrecks and the archaeology of ships. International Journal of Nautical Archaeology 1: 1–58. Bass G.F. and Van Doorninck F.H. Jr. 1971. A Fourthcentury Shipwreck at Yassi Ada, AJA 75: 27–37. Benoit F. 1961. L’épave du Grand Congloué a Marseille, XIV supplément a Gallia, Paris. Boaretto E. 2004. Unpublished C-14 analysis report performed at the Weizmann Institute of Science, Rehovot, Israel. Boetto G. 2000. New technological and historical observations on the Fiumicino 1 wreck from Portus Claudius (Fiumicino, Rome). In J. Litwin (ed.).

Although unlikely, the possibility that these timbers were parts of harbor installations cannot be absolutely discounted. Since the subject is maritime construction technology, either ships or harbors, this question does not significantly influence the conclusions.

Down the River to the Sea, Eighth International Symposium on Boat and Ship Archaeology, Gdańsk 1997. Pp. 99–102. Gdańsk.

Conclusions

the Ninth International Symposium on Boat and Ship Archaeology, Venice 2000. Pp. 66–70. Oxford.

Boetto G. 2003. The Late Roman Fiumicino 1 Wreck: Reconstructing the Hull. In C. Beltrame (ed.). Boats, Ships and Shipyards. Proceedings of

From the above evidence, analysis and the crossdating methods, the timbers from the Akko Marina can be dated between the third and the fifth centuries CE. Of course, if an earlier date (second century CE) or even if the latest C14 date (mid-sixth century) is considered, the timbers would be among the earliest evidence of the use of this combination of shipbuilding techniques in the Eastern Mediterranean. These features point to an early stage of the transition and a hull of mixed construction. Signifying one of the earliest uses of unpegged mortise-and-tenon joints in that region, they represent a relatively early application of the frame-based technique in ship construction in the eastern Mediterranean.

Carre M.-B. and Jézégou, M.-P. 1984. Pompes à chapelet sur des navires de l’Antiquité et du début du Moyen-Âge. Archaeonautica 4: 115–43.

Acknowledgment

Casson L. 1995. Ships and Seamanship in the Ancient World. Baltimore and London.

Boetto G. 2006. Les navires de Fiumicino (Italie): architecture, matériaux, types et fonctions. Contribution à l’étude du système portuaire de Rome à l’époque impériale. Ph.D. diss. Université de Provence, Aix-Marseille 1. Aix-en-Provence. Boetto G. 2008. L’épave de l’Antiquité tardive Fiumicino 1: Analyse de la structure et étude fonctionnelle. Archaeonautica 15: 29–62.

Chevalier Y. and Santamaria C. 1972. L’épave de l’Anse Gerbal à Port-Vendres (Pyrénées-Orientales). Rivista di Studi Liguri 37: 7–32.

My thanks to Mr. John Tresman for the English editing.

Chevalier Y. and Liou B. 1974. Sauver l’épave antique de Port-Vendres. Archeologia 70: 49–55.

194

Ship Timbers from the Akko Marina

Clerc J.-P. and Negrel J.-C. 1973. Premiers résultats de la campagne de fouilles 1971 sur l’épave B de la Pointe de la Luque. Cahiers d’archéologie subaquatique 2: 61–71. Fitzgerald M.A. 1994. Chapter VI: The Ship. In J.P. Oleson (ed.). The Harbours of Caesarea Maritima, Vol. II: The Finds and the Ship. Pp. 163–223. Oxford.

Gassend J.-M. and Cuomo J.-P. 1985. Un acquis récent des recherches d’architecture navale: la ‘construction alternée’ des navires antiques. In VI

congreso internacional de arqueologia submarina Cartagena 1982. Pp. 343–350. Madrid. Guibal F. and Pomey P. 2009. Ancient Shipwrecks, Naval Architecture and Dendrochronology in the Western Mediterranean. In R. Bockius (ed.). Between the seas: Transfer and Exchange in Nautical Technology, Proceedings of the 11th International

Galili E., Sharvit J., Dahari U., Bahat-Zilberstein N., Finkielsztejn G., Stern E.J., Kool R., Kahanov Y. and Rosen B. 2002. Akko Harbor, Underwater Surveys. Hadashot Arkheologiyot 114: 13–16, 12*–15* (Hebrew).

Symposium on Boat and Ship Archaeology, Mainz 2006. Pp. 219–226. Mainz.

Galili E. and Sharvit J. 2002. Acre Port—New Finds Revealed from Underwater Archaeological Surveys. In B. Galil and Y. Mart (eds.). The 14th Symposium

Jézégou M.-P. 1983. L’épave II de l’Anse SaintGervais à Fos-sur-Mer. Ph.D. diss. Université de Provence, Aix-Marseille 1. Aix-en-Provence.

Pp. 12–13, 41–42. Haifa. (Hebrew/English).

Jézégou M.-P. 1985. Eléments de construction sur couples observés sur une épave du haut Moyen-Âge découverte à Fos-sur-Mer (Bouches-du-Rhône). In VI Congreso internacional de arqueologia submarina, Cartagena 1982. Pp. 351–356. Madrid.

on the Mediterranean Continental Margin of Israel. Galili E. and Rosen B. 2008. Acco Harbor: New Finds Revealed while Deepening the Port. In E. Stern, H. Geva and A. Paris (eds.). The New Encyclopedia of Archaeological Excavations in the Holy Land 5. Pp. 1558–1561. Jerusalem. Galili E., Rosen B., Zviely D., Silberstein N. and Finkielsztejn J. 2010. The Evolution of the Akko Harbour and its Trade links Revealed by Recent Underwater and Coastal Archaeological Research. Journal of Island and Coastal Archaeology 5: 191–211. Galili E. This volume a. The Akko Marina Archaeological Project–Introduction. Galili E. and Rosen B. this volume e. The Akko Marina Archaeological Project, Summary. Gassend J.-M. 1979. Vestiges d’une épave antique dans le port de Pommègues (Marseille). Archaeonautica 2 (1978): 101–107. Gassend J.-M. 1982. Le navire antique du Lacydon, Musée d’histoire de Marseille. Marseille. Gassend J.-M. 1989. La construction navale antique de type alterné; un exemple de construction. In H. Tzalas (ed.). Tropis I, Proceedings of the 1st

International Symposium on Ship Construction in Antiquity, Piraeus 1985. Pp. 115–127. Athens.

Gassend J.-M. and Cuomo J-P. 1982. La construction alternée des navires antiques et l’épave de la Bourse à Marseille, Revue Archéologique de Narbonnaise XV: 263–72. Gassend J.-M., Liou B. and Ximénès S. 1984. L’épave 2 de l’Anse des Laurons (Martigues, Bouches-duRhône), Archaeonautica 4: 75–105.

Jézégou M.-P. 1989. L’épave II de l’anse SaintGervais à Fos-sur-Mer (Bouches-du-Rhône): un navire du haut Moyen-Âge construit sur squelette. In H. Tzalas (ed.). Tropis I, Proceedings of the

1stInternational Symposium on Ship Construction in Antiquity, Piraeus Greece 1985. Pp. 139–146.

Athens.

Jézégou M.-P. 1998. Le mobilier de l’épave SaintGervais 2 (VIIe s.) à Fos-sur-Mer (B.-du-Rh.). In M. Bonifay, M.-B. Carre and Y. Rigoir (eds.). Fouilles à

Marseille. Les mobiliers (Ier–VIIe siècles ap. J-C.), 5. Pp. 343–351, 418. Marseilles.

Joncheray J.-P. 1972. La recherche archéologique subaquatique en France: perspectives et realization. Archéologia 48: 8–15. Joncheray J.-P. 1975. Une épave du Bas empire: Dramont F. Cahiers d’archéologie subaquatique IV: 91–140. Joncheray J.-P. 1977. Mediterranean Hull Types Compared 2. Wreck F from Cape Dramont (Var), France. International Journal of Nautical Archaeology 6.1: 3–7. Kahanov Y. 2001. The Byzantine Shipwreck (‘Tantura A’) in the Tantura Lagoon, Israel. Hull Construction Report. In H. Tzalas (ed.). Tropis VI, Proceedings of the 6th International Symposium on Ship Construction in Antiquity, Lamia Greece 1996. Pp. 265–271. Athens.

195

Yaacov Kahanov

Kingsley S. and Raveh K. 1996. The Ancient Harbour and Anchorage at Dor, Israel: Results of the Underwater Surveys 1976–1991 (BAR International

Kahanov Y. 2003. Dor D Wreck, Tantura Lagoon Israel. In C. Beltrame (ed.). Boats, Ships and

Shipyards, Proceedings of the Ninth International Symposium on Boat and Ship Archaeology, Venice 2000. Pp. 49–56. Oxford.

Series 626). Oxford.

Liou B. 1973. Recherches archéologiques sousmarines, Gallia 31.2: 571–608.

Kahanov Y. 2010. From shell to skeleton construction: introduction. In P. Pomey (ed.).

Transferts technologiques en architecture navale méditerranéenne de l’Antiquité aux temps modernes: identité technique et identité culturelle. Actes de la Table Ronde Internationale d’Istanbul, 19–21 mai 2007 (Varia Anatolica XX). Pp. 95–87. Istanbul.

Liou B. 1974. L’épave romaine de l’Anse Gerbal à Port-Vendres, Comptes rendus de l’Académie des inscriptions & Belles-Lettres: 414–433. Liou B. 1975. Recherches archéologiques sousmarines. Gallia 33.2: 571–605.

Kahanov Y. 2011a. Ship reconstruction, documentation, and in situ recording. In A. Catambis, B. Ford and D. Hamilton (eds.). Oxford Handbook of Maritime Archaeology. Pp. 161–81. Oxford.

Marsden P. 1972. Ships of the Roman period and after in Britain. In G.F. Bass (ed.). A History of Seafaring. Pp. 114–132. London.

Kahanov Y. 2011b. The Dor/Tantura Shipwrecks: Clues to their Construction Tradition. Archaeologia Maritima Mediterranea 8: 137–151.

Marsden P. 1974. The County Hall Ship, London. International Journal of Nautical Archaeology 3.1: 55–65.

Kahanov Y. and Breitstein S. 1995a. Tantura excavation 1994: a preliminary report on the wood. CMS News 22: 10–12.

Marsden P. 1994. Ships of the Port of London, Archaeological Report 3. Pp. 109–129. London. McGrail S. 1995. Romano-Celtic Boats and Ships: Characteristic Features. International Journal of Nautical Archaeology 24.2: 139–145.

Kahanov Y. and Breitstein S., 1995b. A preliminary study of the hull remain. INA Quarterly 22.2: 9–13.

McGrail S. 1997. Early frame-first methods of building wooden boats and ships. The Mariner’s

Kahanov Y. and Mor H. 2014. The Dor 2001/ 1 Byzantine Shipwreck, Israel: Final Report. International Journal of Nautical Archaeology 43.1: 41–65.

Mirror 83.4: 76–80.

Medas S. 1999. Il relitto tardo-romano del Parco di Teodorico a Ravenna. Nota preliminare. Archeologia delle Acque 2: 135–138.

Kahanov Y. and Royal J.G. 1996. The 1995 INA/ CMS Tantura, a Byzantine Shipwreck excavation— Hull Construction Report. CMS News 23: 21–23.

Medas S. 2001. Il relitto tardo-romano del Parco di Teodorico a Ravenna, NAVIS rassegna di studi di archeologia, etnologia e storia navale 2: 104–118.

Kahanov Y. and Royal J.G. 2001. Analysis of Hull Remains of the Dor D Vessel, Tantura Lagoon, Israel. International Journal of Nautical Archaeology. 30.2: 257–265.

Medas S. 2003. The Late-Roman ‘Parco di Teodorico’ Wreck, Ravenna, Italy: Preliminary Remarks on the Hull and the Shipbuilding. In C. Beltrame (ed.).

Kahanov Y., Royal J.G. and Hall J.L. 2004. The Tantura Wrecks and Ancient Mediterranean Shipbuilding. In F.M. Hocker and C.A. Ward (eds.). The Philosophy of Shipbuilding. Pp. 113–127. College Station, Texas.

Boats, Ships and Shipyards. Proceedings of the Ninth International Symposium on Boat and Ship Archaeology, Venice 2000. Pp. 42–48. Oxford. Mor H. and Kahanov Y. 2006. The Dor 2001/1 Shipwreck, Israel. A Summary of the Excavation. International Journal of Nautical Archaeology 35.2: 274–289.

Kampbell S.M. 2007. The Pantano Longarini Shipwreck: A Reanalysis, unpublished MA thesis, Texas A&M University. College Station, Texas. (http: //nautarch.tamu.edu/Theses/pdf-files/KampbellMA2007.pdf).

Mouchot D. 1968–1969. Épave romaine ‘A’ du port de Monaco. Bulletin du Musée d’Anthropologie Préhistorique de Monaco 15: 159–201.

Kingsley S. 2002. A Sixth Century AD Shipwreck off the Carmel Coast, Israel. Oxford.

Navri R., Kahanov Y. and Cvikel D. 2013. The Byzantine-Period Dor 2006 Shipwreck, Israel:

196

Ship Timbers from the Akko Marina

Preliminary Hull Construction Report. International Journal of Nautical Archaeology 42.2: 305–325.

Steffy J.R. 1994. Wooden Ship Building and the Interpretation of Shipwrecks. College Station, Texas.

Negrel J.-C. 1973. Une coque du bas-empire dans la rade de Marseille. Archéologia 55: 59–65.

Steffy J.R. 2004. Construction and Analysis of the Vessel. In G.F. Bass, S.D. Matthews, J.R. Steffy and F.H. Van Doorninck Jr. Serçe Limanı, an Eleventh-Century Shipwreck, Vol. 1: The Ship and Its Anchorage, Crew and Passengers. Pp. 153–169. College Station, Texas.

Parker A.J. 1992. Ancient shipwrecks of the Mediterranean and the Roman provinces. Oxford.

Pomey P. (ed.). 1987–1988. Recherches sous-marines, Gallia Informations. Préhistoire et Histoire 1.

Pomey P. 1988. Principes et méthodes de construction en architecture navale antique. In Navires et

commerces de la Méditerranée antique, Hommage à Jean Rougé, Cahiers d’Histoire XXXIII.3–4:

Throckmorton P. 1969. Ancient shipwreck yields new facts and a strange cargo, National Geographic, 135.2: 282–300.

Pomey P. (ed.). 1992. Recherches sous-marines,

Throckmorton P. 1985. The Torre Sgaratta ship. In H. Tzalas (ed.). Tropis I, Proceedings of the 1st International Symposium on Ship Construction in Antiquity, Piraeus, Greece, 1985. Pp. 263–274. Athens.

Pomey P. 1998, Conception et réalisation des navires dans l’Antiquité méditerranéenne. In E. Rieth (ed.).

Throckmorton P. and Kapitän, G. 1968. An Ancient Shipwreck at Pantano Longarini. Archaeology 21.1: 182–187.

49–72. Ramonville Saint-Agne.

Throckmorton P. and Throckmorton J. 1973. The Roman wreck at Pantano Longarini. International Journal of Nautical Archaeology 2.2: 243–266.

397–412.

Gallia Informations. Préhistoire et Histoire 1: 1–85.

Concevoir et construire les navires. De la trière au picoteux (Technologie, Idéologies, Pratique, Revue d’anthropologie des connaissances XIII.1). Pp.

Pomey P. 2004. Principles and methods of construction in ancient naval architecture. In F.M. Hocker and C.A. Ward (eds.). The Philosophy of Shipbuilding. Pp. 25–36. College Station, Texas. Pomey P., Kahanov Y. and Rieth E. 2012. Transition from Shell to Skeleton in Ancient Mediterranean Ship-construction: Analysis, Problems, and Future Research. International Journal of Nautical Archaeology 41.2: 235–314. Riley W.E. and Gomme L. 1912. Ship of the Roman

period discovered on the site of the new County Hall, London. London. Rival M. 1991. La Charpenterie navale romaine. Paris.

Royal J.G. and Kahanov Y. 2005. New Dating and Contextual Evidence for the Fragmentary Timber Remains Located in the Dor D Site, Israel. International Journal of Nautical Archaeology 34.2: 308–313.

Van Doorninck F.H. Jr. 1972. Byzantium, mistress of the sea. In G.F. Bass (ed.). A History of Seafaring Based on Underwater Archaeology. Pp. 133–158. London. Van Doorninck F.H. Jr. 1976. The 4th century wreck at Yassi Ada. An interim report on the hull. Iinternational Journal of Nautical Archaeology 5.2: 115–31. Van Doorninck F.H. Jr. 1982. The hull remains. In G.F. Bass and F.H. Van Doorninck Jr. (eds.). Yassi Ada, Vol. 1: A Seventh-century Byzantine Shipwreck. Pp. 32–64. College Station, Texas. Wachsmann S. and Kahanov Y. 1997. The INA/ CMS Joint Expedition to Tantura Lagoon, Israel. International Nautical Archaeology Quarterly 24.1: 3–18. Ximénès S. and Moerman M. 1991. Le materiel archéologique de l’épave Laurons II. (Martigues, Bouches-du-Rhône). Cahiers d’archéologie subaquatique X: 209–222.

Scrinari V.S.M. 1979. Le navi del Porto di Claudio. Roma. Segal D. and Carmi I. 2004. Rehovot Radiocarbon Date List VI. ‘Atiqot 48: 127. Steffy J.R. 1982. Reconstructing the Hull. In G.F. Bass and F.H. van Doorninck Jr., Yassi Ada, Vol. 1: A Seventh-Century Byzantine Shipwreck. Pp. 65–86. College Station, Texas.

1 The tree species were identified by Prof. N. Liphschitz from Tel

Aviv University. 2 The metal nails were analyzed by Prof. Sariel Shalev of the Leon

Recanati Institute for Maritime Studies, University of Haifa. 3 The tests were: RT-2517: 147–250 CE; RT-2519: 249–372 CE

(Segal and Carmi 2004: 127); RT-4386: 420–470 (26.3%), 480– 540 CE (39.9%) (Boaretto 2004).

197

Yaacov Kahanov

Table 1: Details of plank pieces (Timbers A/3) TimberNo.

Length (cm)

Width (cm)

Thickness (cm)

Notes

1a

80

40

0.5

Teredo traces. Close to the bark, a few small tree knots. Pieces 1a and 1b were broken from one timber and can easily be reassembled. The wood was identified as cypress. It was C14 tested.

1b

63

40

2.5

Heavily damaged. Large splits. Teredo-eaten.

2

45

5.8

3.8

3

23

4.5

3.5

Heavily damaged. Teredo-eaten. On one edge, remains of 1.6cm hole for treenail. Square bronze nail, 0.5cm in cross-section, 1.8cm in diameter, cap-head projecting 0.7cm above the wood surface.

Table 2: Details of thick timber pieces (Timbers B/3) Timber No. 1 2 3 4

Length (cm) 16.0 14.0 13.5 7.0

Width (cm) 10.5 14.0 12.0 10.0

Thickness (cm) 4.8 4.0 3.5 3.5

Notes Beveled on one side. Treenail hole 1.6cm in diameter. Remains of a circular hole. On one side, signs of treenail hole 2.1cm in diameter.

198

Ship Timbers from the Akko Marina

Table 3: Details of bronze nail and treenails (Timber C/1) Nail No.

1

2

3

4

5

6

Description

Round treenail

Round treenail

Bronze nail

Round treenail

Round treenail

Round treenail

Diameter (or cross-section) (cm)

1.7

1.5

Square section 0.8

1.9

1.6

1.8

Distance to previous nail (cm)



2.3

4.8

1.8

5.3

2.1

Notes

Treenail Treenail within hole, within hole close to no. 1

14.5cm long, head concretion 2.2cm, protruding 0.6cm

Treenail within hole, fully penetrated

Treenail within Treenail hole, fully within hole penetrated

Nail No.

7

8

9

10

11

12

Average

Description

Round treenail

Round treenail

Round treenail

Round treenail

Round treenail

Round treenail

Diameter (or cross-section) (cm)

1.8

1.6

1.7

2.0

1.8

NA

1.74

Distance to previous nail (cm)

2.5

1.4

7.4

7.2

14.0

2.5

4.66

Notes

Treenail within hole

Treenail within hole, fully penetrated

Treenail within hole, fully penetrated

Treenail within hole, fully penetrated

Hole remains without nail

Hole remains without nail

Table 4: Details of metal nails and treenails (timber C/2 side 1) Nail No.

1 2 3 Square Round Round Description treenail bronze nail treenail 5mm Diameter 1.6 1.5 section (cm) Distance to previous nail – 5.5 9.0 (cm) Hole Black without material treenail, – Notes 5.5 cm 4.0 cm deep deep

4 5 6 7 8 9 Round Round Round Round Round Round treenail treenail treenail treenail treenail treenail

Average

1.8

1.6

1.8

1.7

1.9

1.7

1.73

9.5

3.9

4.3

3.7

7.3

6.6

6.23

Treenail Treenail Treenail Treenail Treenail Treenail within within within within within within hole hole hole hole hole hole

199

Yaacov Kahanov

Table 5: Details of nails and treenails (timber C/2 side 2) Nail No.

1

2

3

Description

Round treenail

Round treenail

Round treenail

1.7

1.5

1.5

1.57



7.5

11.0

9.25

Treenail within hole, fully penetrated

Hole without treenail

Treenail within hole

Diameter (cm) Distance to previous nail (cm) Notes

Average

Table 6: Mortise-and-tenon (unpegged) details of Akko Marina timbers

Timber no. A.1

A.2 B.1 B.2 Average Tenon width / mortise width Tenon thickness / mortise thickness

Mortises-and-tenon joints Mortises (cm) Tenons (cm) Center Tenon Mortise Width Thickness Width Thickness Depth to center no. no. distance 1 5.7 0.6 4.7 – 1 5.3 0.35 2 – 0.7 – 23.0 – – – 3 – 0.8 4.7 26.0 – – – 4 5.9 0.6 – – – – – 5 4.9 0.7 – 12.5 – – – 6 5.6 0.7 – 11.8 – – – 7 – – – 12.0 – – – 8 6.2 0.7 – – 2 5.1 0.5 9 6.0 0.8 – – – – – 10 6.5 0.8 – – 3 5.8 0.7 11 – – – 12.5 – – – 12 6.0 0.7 – – 4 5.5 0.5 5.85 0.71 4.7 16.3 5.43 0.51

Maximum preserved length 3.5

93% 72%

200

4.3

3.9

Ship Timbers from the Akko Marina

Table 7: Bronze nail details Bronze nails Nail no.

Length (cm)

A/3. 3

1 2 3 4

– – – –

B/1

5



C/1

6

14.5

C/2.1

7 8 9 10

– – 12.7

Timber no.

A/1

C/3

Head Transversal diameter cross-section (cm) (square) (cm)* 1.7 0.4 1.7 0.4 1.4 0.4 1.8 0.5 2.8 0.4 concretion 2.2 0.8 concretion 0.5 – – – 0.7 1.7 0.7 1.66 0.53

Average * Nails tapered along their shanks

Table 8: Attachment details of Fiumicino vessels Distance Tenon width Planking between / mortise Frame-plank joints tenon pegs width attachments (cm) (average, %)

Plank maximum width and thickness (cm)

M&t (the majority unpegged) + iron nails

M&t (many unpegged)

36

58

Iron nails through plug treenails

40.0 × 4.4

Pegged m&t + iron nails

Pegged m&t

42.3

79

Iron nails through plug treenails

27.0 × 5.5

Pegged m&t Iron nails (to sterngripe)

Pegged m&t

28.9

61

Treenails

21.5 × 3.0

Unpegged m&t + iron nails

Pegged m&t

27.2



Treenails

17.0 × 2.5 (average)

Pegged tenons (keel elements were connected by copper nails and treenails)

Pegged m&t

27.7



Treenails

22.0 × 2.3

Keel-garboard attachments Fiumicino 1 (Oneraria maggiore 1; Parker F) Fiumicino 2 (Oneraria maggiore 2; Parker G) Fiumicino 3 (Oneraria minore 1; Parker B) Fiumicino 4 (Oneraria minore 2; Parker C) Fiumicino 5 (Barca del pescatore; Parker A)

201

Yaacov Kahanov

Table 9: Summary of fastener details

Vessel

Date (CE)

Mortise Centerdimensions to- center (width, mortise thickness, distance depth, cm) (cm)

Tenon dimensions (width, thickness, full length, cm)

Tapered pegs diameter (cm)

Tenon width / mortise width (%)

Treenail diameter (mm)

Metal nails Metal type

Dimension (cm)

Laurons 2

End of 2nd c.



18.01

6.0 × 1.0 × 12.0–13.0

0.9–1.1

Similar to mortise

1.5

Copper or bronze



Bourse (Lacydon)

190–220

6.0 × 1.0 ×10.0

20.0

6.0 × 1.0 × 12.0–13.0

0.8–1.5

Well fitted

1.5–2.0

Copper

Long bolts, 2.0 cm in diameter

Torre Sgaratta 2

End of 2nd c.– start of 3rd c.



Monaco

End of 2nd c.– first half of 3rd c.

4.0–6.0 × 0.6 × ?

Akko Harbor

3rd c.

Pommègues

2nd half of 3rd c.

County Hall

300

Pt. de la Luque B

4th c.

Yassi Ada 2

4th c.

2nd half of Dramont F 4th c.

2.8–8.7, spaced 3.5 10.4–14.5 × (average), 12.0 × 1.1 × 1.0–1.2 × center-to16.7 8.0–10.0 center 16 (average)

1.1–1.4

96

2.0

Copper, bronze (and iron in repairs)

12.5 ? × 0.5–0.6 × (average) 7.0–8.0

Used



Used

Bronze



5.9 × 0.7 × 4.7

11.8–26.0, 16.3 5.4 × 0.5 × ? (average)

Unpegged

93

16.8

Bronze

Square, 0.53

5.0–7.0 × 0.7 × 5.0

14.0–15.0

1.1–1.2



1.5–1.8





1.6

58 (one example)

3.2

Iron (limited)



0.8–1.1



1.0





0.7–1.1

50–-64

1.6

Iron



1.1–1.4

Metal

Square, 0.8

Metal bolt

1.0–1.5 in diameter



15.2–97.0 11.0 × 0.7 (spaced, not 6.4 × 0.7 × × 6.5 center-to12.7 center) 6.0 × 0.7 × 10.0–12.0 – 4.0 7.0–9.0 × 0.7 × 15.0–32.0 4.5 × ? × 8.5 5.0–5.5 5.0–8.2 × 0.7 × 4.0–5.5

19.5–25.0, 22.4 (average)

3.5–5.5 × 0.5–0.6 × 6.0–7.0

0.7–0.8

70

Port Vendres 1

ca. 400

Est. 7.0 wide

Estimated average. 15.0





Well fitted

Used

Iron and copper (bronze)

Long iron bolts, copper nails

Parco di Teodorico

5th c.



ca. 80



Unpegged

Part of the mortise

Used

Iron



Tantura A

End of 5th– beginning of 6th c.

Beginning Dor 2001/1 of 6th c.

No mortise or tenon, frame-based construction, seam caulking

No mortise or tenon, frame-based construction, seam caulking

202

Iron

Iron

Square, tapered, near the head 0.73

Ship Timbers from the Akko Marina

Vessel

Date (CE)

Mortise dimensions (width, thickness, depth, cm)

Centerto- center mortise distance (cm)

Tenon dimensions (width, thickness, full length, cm)

Tapered pegs diameter (cm)

Treenail Tenon diameter width / (mm) mortise width (%)

Metal nails Metal type

Dimension (cm)

Dor 2006

Middle of 6th–first quarter of 7th c.

3.8–6.3 × 0.5–0.6 × 2.4–3.5

151 (one case)

Tapered: 2.6–3.2 × 0.4–0.5 × 2.6–3.0

Unpegged

60

A few were used

Iron

Various square 0.4 and 10 long–17 × 19. 50 long, and possible round bolts 4.0 in diameter, 73 long

St. Gervais B

7th c.



100.0



Unpegged



Used

Iron



Yassi Ada 1

625

5.0 × 0.5 × 3.5

Unpegged

60

_

Iron

Several sizes

Pantano Longarini

Early 7th c.



Unpegged





Iron



60

1.5

Iron

0.5 × 0.5

Dor D

Fiumicino

5.0–9.0, 350–621 average 6.3 × 0.5 × 3.0

35.0-225.0 3.0 × ? × 7.0 100.0

28.6 (average)



Unpegged except to 3.5 × 0.3–0.5 one peg, × 6.0 1.0 cm in diameter See Table 8 above

1 Center-to-center distance is based on the reported 12.0cm edge-to-edge spaces between adjacent mortises, with additional 6.0cm average mortise width, as reported. 2 There are several sources for this wreck’s details, as mentioned above. The dimensions in the table are of the largest margins.

203

Chapter 7

The Fatimid, Crusader, Mamluk and Ottoman Ceramics from the Akko Marina Ehud Galili (1), Baruch Rosen (2) and Jacob Sharvit (2) 1. Israel Antiquities Authority and Zinman Institute of Archaeology, University of Haifa 2. Israel Antiquities Authority Key words: Acre, marine archaeology, underwater archaeology, tobacco pipes, Greek fire, porcelain, Islam

Introduction

The Fatimid Pottery

After the Muslim conquest of the Levantine coast Akko often served as a base for Islamic naval expeditions. In the ensuing Fatimid period Akko was a fortified city with a regionally important port that was rebuilt and improved ca. 960–970 CE. Akko was captured by the Crusaders in 1104 and soon became the major port of the Kingdom of Jerusalem. During the heyday of that kingdom the city and its port served water-borne commerce from the entire Mediterranean and from Atlantic Europe. Merchants from Genoa’s, Pisa’s and Venice traded in Akko and their ships visited its port. In 1291 Akko was conquered by the Mamluks after a lengthy siege and the city fortifications were destroyed. It is commonly assumed that the city was in ruins for several hundred years. However, it is now known that the port continued to function during the fourteenth and fifteenth centuries and later, albeit at varying intensity (Galili this volume g). Numerous pottery artifacts dated to the Fatimid, Crusader, Mamluke and Ottoman periods were recovered during the deepening of the Akko Marina (Galili this volume a). The detailed report on the Fatimid, Crusaders and Mamluk–Early Ottoman ceramic finds from the Akko Marina (Stern 2013) is summarized briefly here. In addition, the late Ottoman ceramics, the tobacco pipes and the so called Greek fire handgrenades recovered from the marina are described briefly as they are awaiting a detailed study.

Only two sherds of the Fatimid period, from a cooking bowl and a glazed bowl, were recovered from the Akko Marina (Fig. 1). Vessels similar to the first type were recovered from several eleventhcentury sites in Beirut, Caesarea and the Arnon. The second type, having an elaborate champlevé design, like those found in the Serçe Limani shipwreck, was popular in the Fatimid and early Crusader period. It has been suggested that it was produced in Beirut (Stern 2013: 141–142).

The Crusader Pottery Most of the pre-late Ottoman pottery recovered from the Akko Marina belongs to the Crusader period. The types of 68 rims, 50 bases and 105 vessels from the Crusader period are listed (Figs. 2–6) (Stern 2013: 142-157). Vessels from the coastal Levant included: 3 table amphorae; 13 cooking pots; three cooking bowls and two Levantine glazed bowls. There were six examples of Port St. Symeon ware from the North Syrian coast. There were 14 examples from Cyprus: six cooking pots; six pieces of sgraffito ware and two of slip-painted ware. Ceramics from Greece and the Aegean included three Byzantine sgraffito vessels, seven Aegean ware vessels and one champlevé vessel.

Fig. 1: Fatimid vessels: (1) cooking pot, (2) glazed bowl (after Stern 2013) 204

The Fatimid, Crusader, Mamluk and Ottoman Ceramics from the Akko Marina

Fig. 2: Crusader ceramic vessels: (3) amphora; (4–8) cooking pots; (9–11) cooking bowls (after Stern 2013)

Fig. 3: Crusader ceramic vessels: (12–18) glazed bowls; (19–20) cooking pots; (21–24) glazed bowls (after Stern 2013) 205

Ehud Galili, Baruch Rosen and Jacob Sharvit

Fig. 4: Crusader glazed bowls (after Stern 2013)

Fig. 5: Crusader amphorae (after Stern 2013)

206

The Fatimid, Crusader, Mamluk and Ottoman Ceramics from the Akko Marina

Fig. 6: (39–49) glazed bowls; (50) lid; (51) lamp spout; (52) lamp (after Stern 2013)

Western Turkey and/or Greece yielded most of the ceramics; 48 specimens came from that region, including 28 amphorae of three types and 20 examples of Zeuxippus and Zeuxipus-derivative ware. One proto-maiolica ware item came from southern Italy or Sicily. The rest were undefined according to the geographical typology used here. In summary, the majority of the Crusader ceramics recovered were imported (Fig. 7). The ware came from various production centers in the Mediterranean and was obviously imported by sea. The main production centers were in present-day Lebanon and Syria, Cyprus, southern Turkey, Greece and Italy (Stern 2013: 157).

Fig. 7: Quantity distribution of the main Crusader pottery vessels (after Stern 2013) 207

Ehud Galili, Baruch Rosen and Jacob Sharvit

Sphero-conical Vessels (Greek Fire Hand Grenades?)

sort. However the large variety of types and the labor invested in forming and decorating artifacts that were meant to be expendable has engendered debate among scholars as to their true nature (Tushingham 1985; Ettinghausen 1965). Were they vehicles for delivering fire or were they containers of valuable substances? Presently there is no consensus on their intended use, and their use as containers of high-value liquids is often assumed (Avissar and Stern 2005: 119–121). Similar objects were discovered south of the Pisan quarter in Akko (Flinder 1985: 71), in Banias and in Tiberias excavations. Recent reviews about these artifacts recovered from Safed (Barbé 2014), Tiberias (de Vincenz 2008: 147–148) and Banias (Avissar and Stern 2005) include basic classification, discussions as to their possible uses, parallels and relevant references. Scores of such items were discovered in the Neve-Yam Bay, south of Atlit (Galili and Sharvit 1997: 143–144, fig. 206). The closest parallel to the Neve-Yam items and to the one recovered from the Akko Marina (Fig. 8: 5) is a vessel from Banias belonging to Type II.5 (Avissar and Stern 2005: 119–120, fig. 50: 2)

Eleven ceramic artifacts commonly identifiable as “Greek fire hand grenades” were recovered from the marina. These are small containers shaped like pomegranates, pears or abstract spheres. They may be divided into two classes: ball-like objects, of which nine were recovered, and pear-shaped objects, two of which were recovered (Fig. 8). The nine items identified as ball-like containers were made by conventional ceramic techniques. They are globular, made of pink clay and were well fired. The cylindrical neck terminates in a thickened ring-lip; the opening has a diameter of 1.6cm. The conical base, marked by two circles, is a continuation of the globular body. The two pear-shaped artifacts were recovered from the marina entrance. Their basic form is half a globe extended by an inverted cone in the base. The opening at the top is at the apex of an inverted nipple, attached at its base, to the top of the half-globe. The bottom of the half globe, from the mid-point downward, is cone-shaped and terminates in a truncated point. Like other spherical objects of this type, the pear-shaped artifacts are heavily decorated. Such spheroid pottery vessels have been recovered during archaeological excavations of sites in the Levant and farther east, dating from the eighth to the fifteenth centuries CE. They usually have the approximate volume of a cupped hand and are decorated with a variety of patterns formed by several techniques – engraving, printing, attaching and glazing. It was once assumed that they were fire grenades of some

The Mamluk–Early Ottoman Pottery Thirteen ceramic sherds from this period were recovered (Fig. 9), three of which were Italian monochrome graffito ware (Stern 2013: 157–161). These ceramics are also known from Jerusalem and Ramla. Outside of Israel they are known from Alexandria, Cyprus, Greece and Venice. Two samples of polychrome sgraffito ware bowls were also examined.

Fig. 8: Greek fire hand grenades (IAA) 208

The Fatimid, Crusader, Mamluk and Ottoman Ceramics from the Akko Marina

Fig. 9: Mamluk–early Ottoman glazed bowls (after Stern 2013) 209

Ehud Galili, Baruch Rosen and Jacob Sharvit

These are known from urban centers in Israel, e.g. Jerusalem and Ramla and additional such sites. It is believed that these bowls were the most common type of imported glazed bowls in the region during the Mamluk period (Stern 2013: 160). Samples of polychrome sgraffito ware have been found in several Mediterranean countries, e.g. Egypt, northern Syria and Albania. Ceramics from two other Italian types, including incised decoration with thin to large points (a punta e a stecca) and champlevé, with or without green-yellow glaze, were also present in the assemblage from this period. Byzantine-style glazed ware, probably from Greece, was also identified among the sherds. There is a connection between the presence of these ceramic types and the small-scale and limited, but continuous, presence of northern Italian traders in the Akko area starting soon after the fall of Crusader Akko.

a

b

The Late Ottoman Pottery About 700 indicative pottery fragments from the late Ottoman period, possibly with some unrecognized

1

2

c

Fig. 10: Late Ottoman local ware from Rashaya al Fouchar (IAA)

1

2

3

4

Fig. 11: Late Ottoman glazed bowls and casseroles (IAA)

Fig. 12: Varia: (a) lamp; (b) basket handle jar; (c) churn 210

The Fatimid, Crusader, Mamluk and Ottoman Ceramics from the Akko Marina

earlier intrusions, were recovered during the Akko Marina dredging project. The finds have been preliminarily grouped and classified and only samples are presented here. The vast majority of ceramic vessels were the locally manufactured coarse ware typical of sites throughout southern Lebanon and Palestine known as Gaza ware. Among the local ware were some typical Rashaya al-Fouchar items (Fig. 10) from the Lebanese Hermon. These were examples of typical inland (terrestrial) ware, thus finding them in Akko, in a possible marine assemblage, is of interest. Ceramics recovered from this period also include coarse jars/amphorae, hand-made basins, open and closed, glazed, cooking pots with handles, wheelmade lids of cooking pots, some having handles in the center, flat-base, glazed, pans with handles, necks and rims of jugs, some bearing stamps, small jars and juglets of various types, one wheel-made flask and the typically Late Ottoman glazed deep casseroles (e.g. Fig. 11). Among the non-local Ottoman wares, were Turkish Kutaya ware types. Also noted were plates glazed by yellow, green and light blue glaze; plates with marbled glaze; and plates glazed by yellow or brown glaze and decorated in brown. The varia category includes a partly glazed, open wheel made, lamp with a punched nuzzle (Fig. 12:a), a wheel made, glazed, jar having an upper basket handle and a short, thick

1

2

a 3

b

Fig. 13: (a) European imitations of Chinese-style porcelain; (b) producers’ markings (IAA)

nuzzle (Fig. 12:b) and a wheel made churn with an elongated, conical neck (Fig. 12:c). Many of these vessels represent production centers in the eastern Mediterranean. European imitations of Chinese-style porcelain: These include white porcelain plates decorated with blue motifs (Fig. 13a). On the bases of some pottery vessels there are producers or owners’ markings. Most items are dated to the second half of the nineteenth century, and were produced by European manufacturers. Some of them can be identified e.g., E. CHALLINOR & CO (active in England 1853-1860), or less likely, another CHALLINOR manufacturer (all active in England during 1842-1891) (Godden 1964 repr. 1968: 137–138) (Fig. 13b:1).

Ottoman Tobacco Pipes The custom of tobacco smoking, and therefore tobacco pipes, appeared in the Levant during the early seventeenth century (Robinson 1985). Visiting seamen as well as the inhabitants of Akko – a port city in touch with ports all over the Mediterranean – probably adopted the habit at that time. During the seventeenth century pipes were made of white to pale gray clay, and before the end of the century, red or buff clay replaced the previous type. Pipes covered with burnished red slip were also used (Robinson 1983, 1985: 153; Hayes 1980: 4). Bowls were usually ornamented in the mold or after removal with engraving on the body or wreath (Wightman 1989: 73). Some 160 clay tobacco pipes or their remnants were recovered from the Akko Marina. They were preliminary studied according to their fabric, form and surface treatment and grouped according to the shape of the bowls and the stem. When possible the groups were given tentative dates and relevant parallels are cited. The technical terms employed by Robinson (1985) are used when appropriate. Most pipe bowls in this assemblage are of three types: Type A: round bowl (it was divided into three sub groups); Type B: round bowl with flat or disc-shaped base and Type C: lilyshaped stem (Robinson 1983: 270–272; 1985: 163). A few items did not fit any of these types. The types are described below and selected samples from each group are presented. Type A/I: Round bowl, late seventeenth–early eighteenth century (Fig. 14). Twenty exemplars were of this type. It is an early type, made of fine pale clay, burnished on a mold-made body. The bowl is sack-like, stamped with geometric, floral design or rouletted bands. The shank flared into a simple, stepped-ring termination. The shank end was decorated by wheel roulette, the bore is narrow (ca. 0.8cm). Parallels were identified in seventeenth– eighteenth centuries Istanbul and in Mytilene (Hayes 1992: 392–393). In Israel, related items were excavated in Tel Yoqne‘am (Avissar 1996: 198–199, no. 4).

211

Ehud Galili, Baruch Rosen and Jacob Sharvit

1

2

1

2

3 3

4

4

Fig. 16: Ottoman tobacco pipes – Type A/III Round bowl with funnel-like upper lip folded out (IAA) Fig. 14: Ottoman tobacco pipes – Type A/I: Round bowl (IAA) found. They are burnished variants of the dark, reddish-brown type, their fabric is occasionally coarse and they are often lightly polished. The stems are short and circular. The cylindrical upper portion of these bowls has a tendency to break. They are typically heavily decorated with a combination of stamping, notch rouletting and incising. In Israel they have been reported from Dor (Stern 1994: 324, 328, fig. 231); Jerusalem (Wightman 1989; 257, Pl. 63.15, 21); Safed (Barbé 2014; Cohn 2014); and Tel Yoqne`am (Avissar 1996: 198 nos. 1, 2). This type may represent a local tradition.

Type A/II: Round bowl and cylindrical upper portions (Fig. 15). Their color is light brown to reddish-brown. Twenty exemplars of this type were

1

2

4

5

3

6

Type A/III: Round bowl with funnel-like upper lip folded out; these types date from the nineteenth century, with 14 examples in this assemblage (Fig. 16).

Fig. 15: Ottoman tobacco pipes – Type A/II: Round bowl and cylindrical upper portions (IAA)

1

4

2

3

5

6

Fig. 17: Ottoman tobacco pipes – Type B: Red-slipped burnished pipes with disc-based bowl (IAA) 212

The Fatimid, Crusader, Mamluk and Ottoman Ceramics from the Akko Marina

1

2

3

4

5

6

7

8

9

10

11

13

Fig. 18: Ottoman tobacco pipes – Type C: Red-slipped, burnished, lily-shaped (IAA) 213

12

Ehud Galili, Baruch Rosen and Jacob Sharvit

Type B: Red-slipped, burnished pipes with disc-based bowl; common in the eighteenth–nineteenth centuries (Robinson 1985: 198, A23). Eight exemplars belonging to this group were identified (Fig. 17). The bowl has a slightly flared upper portion and it is joint to a thickwalled cylindrical or conical stem. It features a short shank (ca. 5cm) and circular bore ca. 1.5cm at shank end. The lower bowl is decorated with up to four paralleled horizontal ribs and features a series of straight or diagonal lines scored above the disc base. Hayes (1992) suggested that this type developed from earlier disc-based bowls (Robinson 1985, no. C11– C13, A8; 1986: 198, A23).

The scarcity of Fatimid pottery recovered from the Akko Marina is somewhat unexpected in view of the known activities in the port in that period. One of the explanations for that paucity could be that the port at that period was in the eastern basin, away from the surveyed zone. The Crusader pottery assemblage from the Akko Marina as a whole is also very similar to such assemblages in Frankish-Crusader sites in Israel, Cyprus and Syria, attesting to a relatively intense coastal trade interacting with some overland trade. The 700 years of Fatimid, Crusader, Mamluk and Early Ottoman rule in Akko have most probably left numerous archaeological relics on the sea bottom. However, finds recovered from these periods as a whole are scarce relative to those of the earlier (Hellenistic, Roman and Byzantine) and the later (late Ottoman) periods. This anomaly may be associated with the location of the harbor in the different periods or to modern post deposition processes such as dredging activity (Galili and Rosen this volume e).

Type C: Red-slipped, burnished, lily-shaped pipes; most dated to the nineteenth–twentieth centuries (Barbé 2014; Cohn 2014); 39 exemplars were found in the assemblage (Fig. 18). Tobacco Pipes, Summary: Of the three major identified types of smoking pipes recovered from the Akko Marina, rounded bowl pipes constituted 53%; Type B disc-based pipes, 8%; and Type C, lily-shaped pipes 39%. More than half of the smoking pipes were recovered from Area C, 14% from Area B; 15% from Area A; and 7% from Area F. Manufacturer’s marks were identified in pipes from the seventeenth and eighteenth centuries. The assemblage as a whole is similar to such assemblages recovered in Greece, Turkey (Robinson 1985), Cyprus (Baram 1995) and, closer to Akko, in Ti`innik (Ziadeh 1995: 210), Yoqne`am (Avissar 1996) and elsewhere in Israel (Pringle 1986: 142; Wightman 1989: 73–74; Barbé 2014; Cohen 2014; Simpson 2000). A comprehensive reexamination of that assemblage, in view of new discoveries may facilitate a finer resolution of the typology and dating suggested here.

Acknowledgments The authors wish to thank the Israel Antiquities Authority for the institutional support, M. Feinberg Vamosh for the English editing, D. Jacoby and E. Stern for their useful remarks on the manuscript, T. Sagiv and C. Amit for the photos and H. TahanRosen who drew the pottery vessels.

References Avissar M. and Stern E. 2005. Pottery of The Crudsader, Ayyubid, and Mamluk Periods in Israel. (IAA reports 26). Jerusalem. Avissar M. 1996. The Clay Tobacco Pipes. In A. BenTor, M. Avissar and Y. Portugali (eds.). Yoqne`am I: The Late Periods. (Qedem Reports 3). Pp. 198–201. Jerusalem.

Summary and Conclusions It is unlikely that refuse from the city was dumped in the active harbor, which was one of the most important facilities in Akko. Thus it is assumed that most of the pottery recovered from the Akko Marina originated in ships that anchored or were wrecked in the harbor. Thus the material retrieved from the marina probably represents the true maritime activity in Akko port during the discussed periods. Most of the post Byzantine pottery recovered from the Akko Marina can be dated to the Late Ottoman period, some to the Crusader period and few sherds may be dated to the Mamluk and the early Ottoman periods. This diachronic distribution of ceramic remains reflects the presently known distribution of the ceramic materials recovered from the city itself. This confirms the assumption that the underwater collection may represent a true picture of the activities in the port during these periods.

Baram U. 1995. Notes on the Preliminary Typologies of Production and Chronology for the Clay Tobacco Pipes of Cyprus. Report of the Department of Antiquities Cyprus 1995: 299–305. Barbé H. 2014. Mamluk and Ottoman remains in the al-Wata Quarter, Safed (Zefat). ‘Atiqot 78: 113–137. Cohen M. 2014. Remains from the Mamluk and Ottoman periods at the Al-Wata Quarter, Safed (Zefat). `Atiqot 78: 73*–85*, 165–166. Ettinghausen R. 1965. The use of Sphero-Conical Vessels in the Muslim East. Journal of Near Eastern Studies 24: 218–229.

214

The Fatimid, Crusader, Mamluk and Ottoman Ceramics from the Akko Marina

Flinder A. 1985. Secrets of the Bible Seas – an Underwater Archaeologist in the Holy Land. London.

Ziadeh G. 1995. Ottoman Ceramics from Ti`innik, Palestine. Levant XXXVI: 29–245.

Galili E. and Sharvit J. 1997. Underwater Surveys in the Mediterranean 1992–1996. Hadashot Arkheologiot 107 (Hebrew). Galili E. This volume a. The Akko Marina archaeological Project, Introduction. Galili E. and Rosen B. This volume e. The Akko Marina archaeological Project, Summary. Godden A.G. 1964. Encyclopedia of British Pottery and Porcelain Marks. London. Hayes J.W. 1980. Turkish Clay Pipes: A Provisional Typology, In P.J. Davey (ed.). The Archaeology of the Clay Tobacco Pipe IV. Europe I, (BAR International Series 92). Pp. 3–10. Oxford. Hayes J.W. 1992. Excavations at Sarachane in Istanbul II: The Pottery. Princeton. Pringle D. 1986. Finds from the Red Tower 1: The Pottery. Pp. 135–158. London. Robinson R.C.W. 1983. Clay Tobacco Pipes from the Keramicos. Athenische Mittelungen 98: 265–285. Robinson R.C.W. 1985. Tobacco Pipes of Corinth and of the Athenian Agora. Hesperia 54: 149–203. Simpson J. 2000. The Clay Pipes. In R.P. Harper and D. Pringle (eds.). Belmont Castle: The Excavation

of the Crusader Stronghold in the Kingdom of Jerusalem. Pp. 147–171. Oxford.

Stern E. 1994. Dor, Ruler of the Seas: Twelve Years of Excavations at the Israelite-Phoenician Harbor Town on the Carmel Coast. Jerusalem. Stern E. J. 2013. The Fatimid, Crusader and Early Ottoman Ceramic Finds From the `Akko Marina: Some Insights into Medieval Maritime Activity. `Atiqot 76: 139–168. Tushingham A.D. 1985. Excavations in Jerusalem 1961–1971. Toronto. Vincenz de A. 2008. The Pottery. In Y. Hirschfeld and O. Gutfeld (eds.). Tiberias, Excavations in the House of the Bronzes, Final Report, Vol. 1. (Qedem 48). Pp. 105–164. Jerusalem. Wightman G.J. 1989. The Damascus Gate. (BAR International Series 519). Pp. 73–74. Oxford.

215

Chapter 8

Coin Finds from the Akko Marina: Isolated Coins and a Hoard of Thirteenth-Century Gold Florins Robert Kool (1) 1. Israel Antiquities Authority Key words: marine archaeology, Acre, harbor, gold coins, bronze coins, Crusader, Byzantine coins, Umayyad coins

Introduction

three coins were all minted at the local mint of Akko between 198 and 249 CE under the emperors Caracalla, Severus Alexander and Philip I. These finds join the numerous Roman-period ceramic vessels found in the western basin of the marina, both local and imported ware (Galili et al. 2010:197–198; Silberstein et al. this volume).

During two consecutive seasons (1993, 1994) of deepening the Akko Marina (Galili this volume a) 51 gold and bronze coins were found. Unlike in regular surveys of this kind, the coins were discovered under conditions of controlled dredging (Galili et al. 2002: 13–15; Kool 2006).1 The coins were found to belong to four main periods, the Roman (first–third centuries CE), Byzantine (fifth–seventh centuries CE); Umayyad (seventh–eighth centuries CE); and Medieval (eleventh–thirteenth centuries CE).

The Byzantine-Period Coins Five bronzes date to the Byzantine period. All are regular issues of a variety of denominations (Cat. Nos. 5–9). The earliest coin is a small bronze of the fifth century from before the Anastasian reform of 491 (Fig. 2). There are two small 12 nummi coins from the Byzantine mint of Alexandria, both from the reign of Justinian I (527–565). Finally, two badly preserved Byzantine folles are dated approximately to the end of the sixth–beginning of the seventh

The Roman-Period Coins Four Roman Provincial bronzes were found in the dredging of the marina floor (Fig. 1: 1–3) (Cat Nos. 1–4). The earliest specimen was a bronze of Nero (54–68 CE), minted in Antioch. The remaining

1

2

3

Fig. 1: The Roman-period coins 1 All the specimens were cleaned in the IAA metals laboratory, headed by E. Altmark, and photographed by C. Amit of the IAA photography department.

216

Coin Finds from the Akko Marina

10

6

14

Fig. 2: The Byzantine-period coins

Fig. 3: Umayyad-period coins

centuries. These finds join the considerable quantity of Byzantine-period artifacts found near the marina entrance on the seabed, among the remains of a vessel and large quantities of Byzantine pottery (Galili et al. 2010: 198–199).

Medieval Coins

Umayyad-Period Coins Ten coins dating to the period of the Umayyad rule were discovered on the seabed (Fig. 3: 10, 14) (Cat. Nos. 10–19). Two are Byzantine coins, a bronze dodecanumium and a gold solidus – dating to the beginning of Constans II’s reign (641–646 CE) and which arrived in Akko after the Arab conquest (640– 641 CE). Isolated finds of Byzantine solidi are a rare occurrence in excavations, in contrast to their relative abundance in hoard finds.2 Excavation material shows that Byzantine gold continued to arrive even after the Arab conquest of Byzantine Palestine in the 640s albeit in much-reduced numbers compared with the previous period (Bijovsky 2012). The remaining eight coins are anonymous post-reform fals minted by the Umayyad caliphs. Among these regular issues only one copper (No.13) contained an identifiable mint name (al-Urdun). These coins pre-date the alleged construction of the harbor in the days of Ahmad Ibn-Tulun (835–884 CE). Like the seventh-century Syrian amphorae discovered in a Byzantine shipwreck of Yasi Ada, they show that goods and money continued to flow freely despite hostilities between the Byzantine Empire and the new Umayyad state (Bass 1982).

Three exceptional categories of foreign gold finds date from the medieval period (Fig. 4) (Cat. Nos. 20, 21; H.1–H.30). The concentrated find of such diverse gold denominations would be rare in a controlled excavation. Their find in a dredging operation of this kind demonstrates Akko’s importance as a transnational port in the Mediterranean. Akko and its hinterland were part of a triangular trade system that linked Egypt, the Byzantine Empire and the West. This, and its expansion into a major transit center for thousands of pilgrims and immigrants during the Crusader era, transformed Akko into an important international money market during the medieval period (eleventh–thirteenth centuries), the entry point for a multitude of foreign coins in various denominations.3 Among them the following were retrieved in the dredging operation.

Byzantine Histamenon The earliest coin is a thin nomisma histamenon of the Byzantine emperors Basil II and Constantine VIII, dated to the early eleventh century – (Cat. No. 20). Presumably it reached Akko when the port was under Fatimid rule before its capture by the Crusaders in 1104. The histamenon or helioselenaton was the last of six issues introduced by Basil II and Constantine VIII between 1005 and 1025 (DOC 3: 57–58).

2 A few specimens were excavated from seventh-century settlements like Korazin (IAA 24755) (Ariel 2000: 37), Shoham (IAA 63333-4; 66596) (Bijovsky forthcoming) and more recently at Apollonia/Tel Arshaf (IAA 117033; my thanks to O. Tal of Tel Aviv University for permission to mention this find). Hoards that contained substantial quantities of solidi of Constans II in the territory of former Byzantine Syria/Palestine are more numerous: see Bijovsky 2002: 180–183. 3 For the Fatimid period see Gil (1992). For the Crusader period see Jacoby (1998: 106–107; 114–119).

217

Robert Kool

20

21

H1

H2

H3

H4

H5

H6

H7

H8

H9

H10

H11

H12

H13

H14

Fig. 4: Isolated gold coins and hoard dredged from the medieval Akko Marina 218

Coin Finds from the Akko Marina

H15

H16

H17

H18

H19

H20

H21

H22

H23

H24

H25

H26

H27

H28

H29

H30

Fig. 4: Isolated gold coins and hoard dredged from the medieval Akko Marina 219

Robert Kool

This coin supposedly continued the weight standard of the old solidus (DOC 3: 7) but in fact had rapidly debased in the eleventh century and contained barely 90% gold (Morrisson et al. 1985: 129). The histamenon was the Byzantine Empire’s main gold currency struck en par with a smaller, light weight gold coin, the tetarteron, until the monetary reform of Alexius I in 1092 when both were replaced by the heavier 21k Hyperperon. The helioselenaton seems to have circulated widely throughout the Byzantine Empire (DOC 3:57–58). But to my knowledge, finds of these gold coins are almost unknown from Muslim Syria (Heidemann 1998: 383–386) and Fatimid-ruled Palestine.4 Grierson (1954: 75–84) suggested that the lighter weight gold tetarteron circulated in Syria because it resembled in weight the Muslim dinar but he never gave any empirical excavation data to prove his point. More likely, however, the infrequent appearance of Byzantine gold during this period is connected to two monetary developments. First, the extremely limited flow of gold beyond the political borders of the Byzantine Empire – the Byzantine imperial tax administration tended to hoard gold coins rather than allow their widespread circulation.5 After the Byzantine reconquista of Northern Syria (969–1084) These borders now extended eastward along a narrow coastal strip of Southern Anatolia ending at Antioch (lost in 1084), and southward to Laodikeia (Latakia) (Hendy 1985: 278). Secondly, the dominant position of the Fatimid dinar in the region from the late tenth to the second half of the thirteenth centuries (Tal, Kool and Baidoun 2013: 272–273). Interestingly, contemporary anonymous bronze folles did circulate in larger quantities as excavations in Hama (Plough and Oldenburg 1969: 169–170) and Tiberias in northern

Israel show (Bijovsky and Berman 2008: 63–102). Possibly, finds of these fractional bronze are connected either to the above-mentioned Byzantine military activity or increasing contacts between the Byzantine and Fatimid empires.

Muwahhid Dinar The finds in the marina also included a North African dinar of the Muwahhid ruler Abu Muhammad ‘AbdAl-Mu’min Ibn-’Ali (AH 524–528/1130–1163 CE), minted at the port of Bijayah/Bougie in Eastern Algeria (Fig. 4: Cat. No. 21). Gold from North African mints circulated abundantly in this area during Fatimid rule in the tenth–eleventh centuries. This is attested by documents from the Cairo Geniza as well as from the surviving hoard evidence. Numerous documents from the Ben Ezra synagogue at Fustat illustrate the free flow of gold and silver currencies in economic exchanges. Jewish bankers at the money exchange of Fustat exchanged coins from Tunisia and Syria on a daily basis (Goitein 1967: 238–239).6 A substantial number of Fatimid gold hoards deposited in the eleventh century, prior to or on the eve of the Crusader conquest, contain numerous dinar denominations minted in North African mints showing that such North African gold was an integral part of the money pool of the Fatimid state during this period (Tal et al. 2013: 268–270).7 The flow of gold from African mints diminished with the Norman conquests of Sicily (1071) and parts of the North African coast (al-Mahdiyya, 1148) which led to the de-activation of these Ifriqiya mints and the concentration of Fatimid rule to Egypt proper during

4 The above coin is currently the only provenanced specimen known to me. There are two additional gold histamenon of this period in the IAA collection one of Basil II and Constantine VIII (IAA 52795) dated to 1025 CE, and one of Constantine VIII (IAA 52799) dated to 1025–1028. Both coins were bought from local dealers by the Mandatory Coin Department in 1929/30 without provenance cited. 5 For example Basil II’s treasury accumulated an unprecedented amount of 200,000 pounds of gold according to the Byzantine statesman and historian Psellus (1018–1078) (DOC 3: 42). Likewise in the eleventh-century Byzantine province of Calabria, hoard evidence shows that few Byzantine gold coins circulated despite the fact that it was an imperial province: an early eleventh-century hoard from Ordona (Foggia, South Italy) contained only a single Histamenon of Basil II deposited ca.1025 whereas the majority are imitative taris of pre-Norman Amalfi and Salerno (Grierson and Travaini 1998: 40). Two additional gold histamenon were found in two hoards in Rome and Pisa but with subtantial later deposition dates (post-1185 and post-1252) possibly showing that these single specimens were used as religious amulets rather than as coins (Travaini 1995: 11, n. 4; Travaini 2001: 179–196). 6 Goitein (1965: 1–46; 1967). See especially the chapter on money, banking and finance, pp. 229–266. One document written in Jerusalem, ca.1050– 1070 detailed the exchange of ‘reddish’ Egyptian ruba’iyya or quarter dinars (Goitein 1967: 378, n. 35) Another document from Jerusalem, dated 1060, specified the content of a money purse sent from Egypt containing Sicilian quarters and full-weight dinars (Goitein 1967: 459, n. 42). 7 See in particular the three coin hoards that were discovered during the Temple Mount excavations in Jerusalem between 1968 and 1976, and were presumably interred just before or during the conquest of the city by the Crusaders in 1099 (Berman 1976: 76–78). They contained a substantial number of Fatimid dinars and gold fractions minted in North African mints such as Palermo/Sicily, Mahdiyya, Sijilmasa (southern Morocco). One of the hoards also contained a dinar dated 1091 CE, minted under the North African dynasty of the Almoravids. Two additional Almoravid dinars were excavated at Caesarea (Lampinen 1987: 143). Wasserstein (1998: 11) published a dinar minted in Mahdiyya from a hoard excavated in Tiberias. A hoard of 376 dinars discovered in Ramla and deposited some time at the beginning of the eleventh century (?) (Levy and Mitchell 1966: 37–66, No.145) contained coins minted in Mahdiyya and al-Muhammadiya M’silah/Algeria). As yet unpublished hoards excavated at Caesarea and Ramla by the IAA contain numerous quarter dinars and dinars from Mahdiya Mansuriyya and neighboring Muslim-ruled Sicily. 14% of the coins from a gold Fatimid hoard from the Crusader castle of Arsur-Apollonia originated from Ifriqiya mints – al-Mansòūriyya, al-Mahdiyya, Madīnat Izzi al-Islām al-Qayrawān, and Siqilliyya and southern Bilad al-Shams (Tal et al. 2013: 270). Finally a North African hoard found buried in Ayla/Aqaba contained 32 gold dinars, of which 29 came from Sijilmassa in southern Morocco (Whitcomb 1994: 18). 8 Fatimid rule in North Africa, and consequently control over a number of important mints, was reduced considerably with the Norman conquests of Sicily (1071), Malta (1098) and Al-Mahdiyya (1148). A similar process took place on the Palestinian seaboard where the Fatimids lost their mints with the establishment of the Frankish Kingdom and northern principalities: Filastin (Ramla 1099), Tabariya (Tiberias 1099), Akka (Akko 1104), Tarablus (Tripoli 1109), Sur (Tyre 1124), ‘Askalan (Ascalon 1153).

220

Coin Finds from the Akko Marina

the first half of the twelfth century.8 Presumably North African gold continued to enter the Latin Kingdom possibly on a smaller scale with the trickling of Muslim pilgrims from the Maghreb transiting its territory to Mecca.9 With the activities in the Frankish Kingdom of Pisan and Genoese traders, who served as middlemen and carried goods between the Maghreb and the Middle Eastern seaboard, plentiful North African gold must have flowed to the Middle East (Jacoby 1998; Abulafia 1982: 227–243; Messier 1974: 32–33; Spufford 1988: 169–171).

Hoard of Fiorini The third and most significant find is a group of 30 gold fiorini minted in the thirteenth century by the medieval city-republic of Florence (Fig. 4: Nos. H1H30, Table 1).10 The fiorini were recovered in two groups at the same point (Area A) on the seabed, in two consecutive seasons of underwater surveys in 1993 and 1994 preceding clearing work of the medieval harbor of Akko. In the dredged mud and sediments taken from the entrance of the ‘inner’ Western section of the Crusader harbor, 21 coins were discovered in 1993 and an additional nine were excavated during 1994. According to the excavators an additional 20 to 30 coins still lay dispersed on the seabed. Unconfirmed reports speak of yet another 20 or more coins found

by local fishermen.11 It is likely then that this group of florins belonged to a single hoard of at least 70 or 80 coins that were lost on the harbor seabed, as such large quantities of gold coins from a single location are almost never merely chance losses (Table 1). Finds of gold coins, especially single ones, are very rare. It was a precious commodity, frequently hoarded in times of crisis (war, conquest) and debasements, or melted down to be incorporated into the next generation of coins (Casey 1986: 51–57; Harl 1996: 11–14). Finds of Florentine gold in this region are extremely rare. There is only one other recorded find of Florentine gold from the Middle East, and it is a hoard. This find was reported to have come from Aleppo. It contained c. 600 florins, part of a hoard of approximately 630 European gold coins. The hoard, found in the mid-1950s, was briefly described by Philip Grierson as being part of the plunder of Akko which was captured and destroyed by the Mamluks in 1291 (Grierson 1965: 6, n. 2; see below). In 1252 the Italian republics of Genoa and Florence virtually simultaneously re-introduced the minting of gold coinage in Western Europe after a gap of more than five centuries.12 Without doubt the most influential of the two was the gold florin (fiorino d’oro), a coin of almost pure gold (98–99%) weighing 3.53gr, minted by the Florentine Commune. The obverse showed a fleur-de-lis or ‘fiorino,’ symbol of the city and the legend +FLOR – ENTIA. The

Table 1: Estimated contents of the Akko Marina hoard (for season 1993–1994 ‘H’ refers to the catalogue number below) Source/provenance

Number of coins

Salvage excavation 1993 season

H.1– 2; 4–11; 14; 17–18; 21; 23–24; 27–30.

Salvage excavation 1994 season

H.3; 12–13; 15–16; 19–20; 22; 26

Estimation of the excavators

Possible another 20–30 coins that were not recovered (?)

Found by local fishermen (unavailable)

22 coins

Estimated total

70–80(?) coins

9 The Arab noble Usamah who lived in the first half of the twelfth century related how he ransomed several Magribi pilgrims caught by pirating Franks (Usamah ibn Munqidh 1929: 110). 10 The text below follows, with some small changes, my article on the fiorini hoard previously published in the Numismatic Chronicle (Kool 2006: 301–320). 11 E. Galili and Y. Sharvit, pers. comm. May 29, 1997. 12 Gold coinage had disappeared from much of the Christian West between 700 and 1252, although gold in the form of bullion, jewelry, and sometimes Muslim coins were used in payment (Grierson and Blackburn 1988: 90–97; Spufford 1988: 14–26). The exception was southern Italy, which was economically more integrated with the central and eastern Mediterranean, and never really ceased to use gold coinage (Day 1997: 25–45). There exists an extensive body of literature on the return of an effective European gold unit and its repercussions for the medieval economy (Spufford 1988; Cipolla 1956; Bloch 1967: 186–229; Watson 1967: 1–34; Ashtor 1971: 15–29; Richards 1983).There has been some argument over whether the first gold coins were struck by Genoa or Florence (Lopez 1953: 19–55 and Watson 1956: 161–198) argued that Genoa was first with genovino d’oro, a gold coin weighing 3.52gr, while Florence followed several months later minting its first fiorino d’oro in November. Travaini (1991:187–194) proposed that Genoa began to strike gold coins much earlier in the 1180s for trade with the Norman kingdom of Sicily. Recently new information, to be published in the forthcoming volume in the MEC that covers northern Italy, has come to light indicating that Genoa’s large genovino was introduced two decades after the gold florin, c. 1270.

221

Robert Kool

Middle East during the fourteenth century.16 There is much less information about the second half of the thirteenth century; however, a recent article on the subject (Spufford 2006: 421–436), which appeared simultaneously with the publication of the hoard below, has gone a long way to rectify this situation. Documents record the use of fiorini throughout Italy by the 1270s and northward along the Via Francigena, the principal trade route to the Champagne fairs and Paris, and westward as far as Barcelona (Spufford 2006: 429–429). There are a few hoards. One, buried at Pisa in the last decade of the thirteenth century, contained Sicilian taris and Florentine florins (Castellani 1937: 476–84 and Lenzi 1978). A contemporary account of a treasure recovered from French ships returning from Tunis after the 1270 Crusade, which were wrecked in a storm off Trapani in Sicily and included a high proportion of florins, is discussed below (Carolus-Barré 1976: 115–118).17 Peter Spufford has argued that north of the Alps gold coins in general functioned as a commodity rather than a currency. On the other hand there are commercial documents from France that mention the growing presence of gold florins at trade fairs (Dubois 1976: 21; Paoli and Piccolomini 1871) and consumption centers in France from 1265 onward (Spufford 2006: 60–139). A reference to them in England dates from 1285, but such references are few and far between.

reverse depicted the patron saint of the city, John the Baptist in a coat of animal hair, his left hand holding a staff over his shoulder and his right hand raised in benediction with the legend S. IOHA – NNES.B. The introduction of the new currency coincided with the economic and political ascendancy of Florence, in particular the period of rapid growth between the late twelfth and the middle of the thirteenth century. During this period Florence witnessed an unprecedented demographic and economic expansion.13 The resulting accumulation of landed wealth coupled with an expanding trade in cloth and grain generated a growing class of rich Florentine burghers during the first half of the thirteenth century (Brucker 1998; Hunt 1994). Already under Hohenstaufen rule, an increasingly independent Florence had opened its own mint in c. 1237, striking the silver fiorino grosso, a multiple penny. With the death of its nominal overlord Frederick II, in 1250 Florence openly asserted its independence and became a sovereign entity. Research on the economic development of Florence has repeatedly stressed that the new gold currency introduced two years later formed the foundation upon which the Florentine economy expanded internationally.14 Gold florins minted on a large scale allowed Florentine merchants to accumulate vast wealth as papal tax collectors and bankers in Tuscany and Italy from the 1290s onward (Day 1968). More important, Florentine gold financed an extensive wool and cloth trade with the Champagne fairs in eastern France, Flanders and England, and the establishment of a large cloth manufacturing industry on the banks of the river Arno (Goldthwaite 1980). As a result, between 1250 and 1320 Florence developed into one of Europe’s principal commercial and financial centers (Cipolla 1956; 1983; Abulafia 1981). Curiously, in the first decades after its introduction, the gold florin remained unpopular in the city itself. The chronicler Paolo Pieri writing in 1305 stated that “there was virtually nobody who wanted it” (Lopez 1956: 236).15 There is extensive literature on the circulation and imitation of the Florentine florin in Europe and the

The Use of Gold Coins in the Crusader States With the arrival of the First Crusade at Antioch in 1097 the debased gold Histamenon nomisma of Michael VII was still readily available in large quantities in northern Syria (Heidemann 2002: 381– 383). Traveling southward toward Jerusalem, the Crusaders encountered the almost pure gold Fatimid dinar. Presumably vast quantities of these dinar misri, minted in Egypt, remained in circulation for most of the twelfth and the beginning of the thirteenth centuries in the Latin Kingdom of Jerusalem (Kool

13 Florence’s population grew four times its original size in this period, whereas between 1252 and 1325 it merely doubled. For the importance of the early economic development of Florence see Day 2002: 93–129 and Day forthcoming. 14 This is evident both in general historical works like the Cambridge Medieval History and researches related to the economic and monetary history of Florence like Sapori (1955–1967), de La Roncière (1973), de La Roncière (1976), Walker (1983: 29–52), Goldthwaite and Mandich (1994). 15 The same applied to its circulation in Tuscany in the 1270s, where the gold florin was little used for the payment of the papal tithe. 16 The older literature is listed in Nohejlova-Prátová (1971–1972) which covers Italy, Germany and Luxemburg as well as Central Europe, and Spufford (1988: 267–288). There are the following later publications: For the circulation of florins in Italy from the 1320s onward see La Roncière (1976: 505–547) and for southern Italy and Sicily, Grierson and Travaini (1988). The most recent survey is Spufford (2006). For France the presence of florins in hoards is recorded Duplessy (1995); for England see Cook (1999: 255). For the circulation of florins in fourteenth-century Syria and Palestine see Schlumberger (1878: 492) and Ashtor (1992: 178–179). 17 A good example of the problems facing researchers in this area is the hoard list published by Duplessy (1995) for medieval France. Twenty-one of the 440 hoards found in the historical territory of the kingdom of France contain gold florins of Florence. The deposit dates of the hoards are virtually all fourteenth century but no doubt many also contain earlier florins. Unfortunately, details on individual coins are difficult to find, significantly hampering research in this area. Grierson and Travaini (1998: 176) note the same problem for the territory of southern Italy and Sicily.

222

Coin Finds from the Akko Marina

2007). In addition, Frankish mints at Akko, Tyre and Tripoli struck large numbers of gold coins imitating Arabic types from the 1140s onward; however, it is unclear exactly when and why the issue ceased in the mid-thirteenth century.18 In 1251, following papal intervention, a small issue of gold imitation ‘bezants’ with correct Arabic script and Christian legends were struck at Akko but this issue seems not to have lasted beyond 1258 (Balog and Yvon 1958: 158–159). There was also a supply of Ayyubid gold dinars from Egypt, although these apparently did not attain large-scale circulation in the remaining territories of the Crusader states and Muslim-controlled Syria.19 According to current thinking very few such coins can be attributed to the years after 1258. After the Fatimid mints closed there was no Islamic gold coined in Syria until the Mamluks under Baybars I (1260–1277) struck dinars, but only on a limited scale, at Damascus. However, Arab sources note that Crusader imitations continued to circulate well into the 1280s with the last mention being 1302–1303 (Irwin 1980: 92). It is possible that the Crusader gold stopped being issued because it was replaced by the new Italian gold coins. One difficulty with this explanation is that the Syrian Arabic sources do not mention genovini or fiorini at least in the thirteenth century, and so far only one unpublished hoard from Syria has been noted, discussed below.

Numismatic and Documentary Evidence of Gold Florins in Akko Evidence for the use of florins in Outremer can be found in a number of Tuscan and Venetian trading manuals which record their use in the second half of the thirteenth century. It would seem that florins were little used before the late 1270s. A private manual written c.1270 by a Venetian merchant or notary residing in Akko, which contains detailed

information on the workings of the mints of Akko and Alexandria, the transfer of gold bullion and coins and the cost of minting, omits any allusion to florins.20 The earliest reference appears in the Memorie de tucte le mercantie, a manual compiled by a Pisan merchant in 1278 that notes the exchange rate of the florin in Akko and Alexandria (Lopez and Airaldi 1983: 123–124):

Florini d’oro tornano inn Acchari biz. 3… ‘One florin of gold is changed in Acre for three bezants …’

Li pesi C di biz. torna in Alexandria entra fiorini d’oro CXXII… ‘The weight of 100

bezants is changed in Alexandria for a 122 florins of gold …’ They are also referred to in later fourteenth-century manuals such as Pegolotti’s Pratica della mercatura and by the Venetian Zibaldone da Canal (Pegolotti 1936; Grierson 1979: 487; Zibaldone da Canal 1967).21 The Pratica – some of which reflects conditions in the last decade of Frankish rule – mentions the rate of exchange of florins against gold dinars and silver dirhams in Alexandria (Pegolotti 1936: 72). Neither of these mentions their occurrence in Akko and as noted, there is nothing in the Arab sources for thirteenth-century Syria and Palestine. Contemporary commercial contracts and notarial records from the Levant also hardly mention the Florentine currency.22 Among the more than 1000 contracts drawn up by the Genoese notary Lamberto di Sambuceto in Famagusta several years after the fall of Akko, only two documents, dated 1296 and 1302, mention a transaction in florins (Desimoni 1893: 104, no. 202).23 Their near omission from Latin documents originating in the Latin East does not prove that they did not circulate locally. Some Florentine gold must have come eastward with the growing presence of Florentine merchants there. Indeed, in the contracts of Sambuceto one can read of a large presence of Florentine bankers and

18 Against my views on the continuous importance of the misri dinar en par with the Frankish imitations see Heidemann (2009) who cited the scarcity of genuine Fatimid gold as the reason for the predominance of the imitation dinar suri till the late 1250s. 19 I know of the existence of two hoards. One (1150–1250 CE), allegedly found near Aleppo (Metcalf 1995: 316, No. 31), contained a single Ayyubid dinar together with 23 imitation bezants. The other was an allegedly large hoard of dozens, possibly hundreds, of gold dinars found near the medieval port of Acre during a previous dredging operation in 1987, containing late Fatimid, Zandjid and Ayyubid dinars minted by Saladin (1174–1193) and which were dispersed on the market. Otherwise there are no Egyptian-Ayyubid recorded hoards from the Crusader territory or Muslim Syria (Heidemann 2009). 20 D. Jacoby, pers. comm. See also Jacoby (1998: 105–119). 21 The most recent discussion of Pegolotti and the Memorie de tucte le mercantie cited above is in Travaini (2003: 118–130). 22 For example, in neighboring Tyre surviving documents dated around 1265 mention no gold florins, only ‘bissancios sarracinales…ad iustum pondes Syrie’ (Balletto1986: 178, 195–204; 255–267, docs. 5–17). 23 This document, drawn up on September 21, 1300 mentioned the sum of 1,839 gold florins. The merchants involved in the transaction came from Florence and Pisa. The second document is in Polonio pp. 158–159, no.142 and is a contract dated December 3, 1300 from Famagusta involving a loan of 1,450 gold florins to the Florentine firm of the Mozzis. Desimoni published 369 contracts from the archives of the notary Lamberto di Sambuceto for the first time in 1884 and 1893. Since the 1980s a large collection of additional contracts of Sambuceto discovered in the Archivio di Stato in Genoa has been published in the series ‘Collana Storica di fonti e studi’: Lamberto di Sambuceto, Notai Genovesi in Oltremare: Atti Rogati a Cipro da Lamberto di Sambuceto. See Polonio 1982; Pavoni 1982, 1987; Balard 1983; Balletto 1989: 359–364, nos. 121–123; 367–368, no. 128 also contains three documents from Beirut and Laodicea but none mentions florins, only the use of local gold and silver currencies bissancios…sarracinales ad iustum pondus and daremos…novos de Armenia; Balard 1983.

223

Robert Kool

Numismatic Evidence for a Single Hoard: Types, Mint Symbols and Die-links

merchants in Famagusta by the end of the thirteenth century.24 Also, earlier in Akko, merchants from the Tuscan hinterland, including Florentines, traded with the Crusader states under the Pisan flag (Abulafia 1982: 234–243). By posing as pseudo-Pisans these merchants had the advantage of being exempted from the taxes levied on Latin merchants. Sources mention the presence of Florentine spice dealers, cloth merchants and representatives of merchant companies in Akko, who used the city as a base for trade with the Syrian hinterland, in particular Aleppo (Abulafia 1982: 227–233). One of these was Guido dell’ Antella, scion of a family that belonged to the guilds of money changers and wool traders and was intimately involved in the management of the gold mint of Florence from its inception (Spufford 2006: 429).25 It was Florentine capital that financed much of the Sicilian trade in grain and other foodstuffs to the Kingdom of Jerusalem in the second half of the thirteenth century. The eviction of the Genoese from Akko after the War of St. Sabas (1256–1258) meant that Florence may well have dramatically augmented its share of trade with that city, under Pisan guise (Abulafia 1988: 183–202). Presumably this involved transporting its new gold currency eastward, sealed in leather ounce bags (Spufford 1988: 176). Until now the numismatic evidence for the circulation of early florins in the eastern Mediterranean has been exclusively based on the discovery of a single hoard of some 600 early florins and a small number of other gold coins in Aleppo in 1953–1955.26 This unique find, which was never published, has been used by successive historians and numismatists to argue that florins circulated extensively in the Levant. The fact that the Aleppo hoard is but a single hoard and perhaps an anomaly tended to be overlooked. Instead it was taken as an indication of the high position the florin held in the Eastern Mediterranean in the later thirteenth century.27 It also helped fill the vacuum created by the cessation of Crusader gold coins. The discovery of another group of early-period florins is therefore important because it seems to confirm that florins indeed circulated in the Levant in the closing decades of the thirteenth century, a belief until now based solely on the Aleppo hoard and the aforementioned scattered documentary references.

Additional evidence for the view that the two groups of florins found in the Akko Marina belong to one hoard is that they show a considerable degree of homogeneity. Three factors in particular point this way: 1. Same type of fiorini. The fiorini in both groups belong to the same early “small fiorini” type (see below). 2. Occurrence of similar mint symbols in both groups. Five of the 20 coins of the first season (nos. 14, 17, 18, 22, 28) had segni that occurred on four of the nine coins discovered in the second season (nos. 13, 16, 21, 27). This establishes a significant correlation between the two groups. 3. Die-linking. Also significant is the fact that two pairs of coins discovered in different seasons – nos. 14 and 17 in 1993 and nos. 13 and 16 in 1994 – were struck from the same obverse dies. The reappearance of similar mint symbols and die-links is summarized below in Fig. 1.

Dating the Akko Marina Hoard: Identification of Types and Segni Types The 30 florins in the hoard can be dated to the earlier period of issue (up to 1380) by their style and typology: 1. All carry the inscription S. IOHA-NNES. B, which is characteristic of coins struck between 1252 and 1450.28 2. All belong to the ‘small gold florin’ type with a diameter of 20mm. These coins were struck before 1422. 3. All depict St John wearing a coat of animal hair fringed at the sides and the bottom. These coin types were struck before 1411 (Bernocchi 1978: 124 ff).29 4. The fleur-de-lis on the obverse of all the coins portrays two W-shaped (unopened) buds with globules attached at each of the three points. These appear on types struck before 1380 (Ives 1952: 107– 108; Bernocchi 1978: 124 ff).

24 Many of the documents issued between 1296 and 1302 include names of individual merchants (‘de Florencie’) as well as the presence of representatives of large merchant companies (‘societas’) such as the Bardi and the Peruzzi (Polonio 1982a and Polonio 1982b; Pavoni 1982; Balard 1983; Pavoni 1987). 25 Lambertus de Antilla had been the first official charged with managing the new gold mint established in 1252. His sons Donato and Giovanni occupied the position of mint masters in the first decades of the fourteenth century (Spufford 2006: 422). 26 The only published account of the Aleppo hoard consisted of incidental remarks by Philip Grierson in papers devoted to other subjects (Grierson 1957a: 488–489; Grierson 1957b: 82–83; Grierson 1965: vi). 27 Notably, Grierson (1957a); Walker (1994: 43); Lane and Mueller (1985: 179, 276–277); Spufford (1988: 177). Other scholars have been more circumspect, e.g., Yvon 1966: 91); Jacoby (1986: 411). 28 Between 1450 and 1533 the inscriptions read S. IOAN-NES. B (Ives 1952: 108). 29 Bernocchi further distinguished between 1252 and 1421, the period of the small florin, 19 different types of fringed coats made from a coarser textile, with ornamental stitching at bottom. (Bernocchi 1978: Tables I–XIX).

224

Coin Finds from the Akko Marina

Table 2. Occurrence of segni and die links between florins from the 1993 and 1994 seasons

1993

1993

Season

Season

225

Robert Kool

to our knowledge of the earliest florin issues minted between 1252 and 1303. Accordingly, the Akko Marina hoard below warrants a closer inspection of the both relevant mint documents, and the segni displayed.

Herbert Ives’ analysis of the design of the Florentine florins, mentioned also the ‘unopened leaves’ motif on the pendants of the obverse fleur-de-lis. According to him these only appeared on coins minted before 1300. This particular stylistic element was not included by Bernocchi in his elaborate classification of 14 different obverse types appearing between 1252 and 1422 (Bernocchi 1985: 124 ff). Consequently, this motif has been omitted as a dating device.

The Earliest Gold Florin Segni (1252–1303), Problems of Identifications and Dating: In 1317 the Commune of Florence established the Liber

Monetae Communis Florentiae or Libro di Zecca.32

Segni

The purpose of this “mint book” was to collect the different records related to the activity of the mint into one registrum or mint register. The account was organized chronologically, containing the names of the principal mint officials serving every semester.33 The register not only specified names of the two mint masters for gold and petty coinage, it also reproduced in the margins of the manuscript drawings of the segni they had used.34 By 1533, the ultimate date of its codification, the register consisted of 29 sections or cahiers – assemblages of folded sheets – which were composed by 90 different scribes. This collective work contained the names of hundreds of mint masters together with nearly 1000 segni compiled over a period of 217 years.35 In his famous chronicle of Florence Giovanni Villani (1276–1348)36 – who during his term as mint master in 1316–1317 initiated the compilation of the liber monetae – stated that the Republic of Florence had begun minting gold 65 years earlier, in 1252.

Stylistic analysis only helps to restrict the dating of the hoard to before 1380. Fortunately the segni provide a more precise dating instrument. In the case of the Florentine gold coinage this takes the form of a small symbol appearing on the coin’s reverse before the inscription (early florins), but mostly between the end of the inscription and the halo enclosing St John’s head.30 This segno was the emblem of the officiating mint master. According to the Florentine Commune’s constitution the office of the mint master rotated every semester (six months) among members of the city’s leading economic institution, the Arte di Calamala, the guild of the foreign cloth-dressers (Simpson 1952: 121; Bernocchi 1976: 1–2).31 At the inception of gold minting in 1252 the emblem or badge (segno) was a simple mark, namely an annulet. It rapidly developed into a pictorial design alluding to the mint master’s family (type parlant). During the fifteenth century it evolved into a fullfledged coat of arms. Bernocchi’s extensive research on the Florentine mint produced 668 segni for the 282-year period (1252–1533) in which the florins were produced. In theory, we therefore possess an extremely accurate dating device which allows us to identify a coin to within six months. In practice, several crucial hiatuses exist, in particular with regard

...i mercatanti di Firenze per onore del comune, ordinaro col popolo e comune, che ssi battesse moneta d’oro in Firenze; e eglino promisono di fornire la moneta d’oro, che in prima battea moneta d’ariento da danari XII l’uno. E allora si cominciò la buona moneta d’oro fine di XXIIII carati, che si chiamano fiorini d’oro e contavasi l’uno soldi XX; e

30 The earliest florins had no mint master’s mark according to the Libro della zecca: ‘Reperitur florenos auri coniatos fuisse per commune Florentie sine aliquo puncto’, (Bernocchi 1974: 2). 31 For the importance of the wool guilds in Florence see Hoshino (1980). 32 The original manuscript Zecca No.79-Fiorinaio (Inventario 89) can be consulted in the Archivio di Stato, in Florence. A new edition was published by Bernocchi-Fantappie (1974). 33 Among those mentioned systematically are the: domini et officiales monete auri et legarum (mint masters for gold and petty coinage), sententiatores (wardens of the mint), sagiatores (assayers) and the notarius (notary) of the mint. For a more detailed description of the Florentine mint personnel see “La composizione del personale” (Bernocchi-Fantappie1976: 5–14). For the division of labor in Italian mints between the twelfth and fourteenth centuries see Travaini (1977: 47–50). For a comparison with the contemporary mint of Venice see Stahl (2000) and Lane and Mueller 1985). 34 The images of coins appearing in the register consisted of two types: impressions of the obverse coin die on the parchment using gold or silver leaf or even wax, some of them dated to the fourteenth century; and drawings of the segni appearing on the reverse, which were added in the fifteenth and sixteenth centuries. 35 Between 1317 and 1533 797 segni can be securely identified within a six-month period. Another 153 segni exist but their date can only be fixed approximately (from several years to several decades). For the complete listing of mint masters and symbols see Bernocchi 1975). 36 Giovanni Villani was the scion of an important Florentine mercantile family. He became a shareholder in the Perruzi (who also were active in Acre) and Buonaccorsi Compagnies, two of the leading Florentine trading and banking firms. As a merchant in the service of the Peruzzi, he travelled around Europe for several years. Upon his return in 1307 he became involved in government of his native city. He died, in 1348 presumably of the plague, which killed half the population of Florence. Villani’s Cronica, divided into 12 books, described not only Florentine political history but also contained a wealth of economic information (statistical data, cost of provisions, details of the finances of the state). For Villani see Luzzati (1971) and Kleinhenz 2004: 1143–1147).

226

Coin Finds from the Akko Marina

ciò fu al tempo del detto messere Filipo degli Ugoni di Brescia, del mese di novembre gli anni di Christo MCCLII... (Villani, Cronica Liber VII, cap. 53 =

Villani 1990–1991: A 346). From Villani’s words it would appear that we lack even the most basic record on more than 100 mint masters and their segni during the first 65 years of the gold minting. In fact, in the original Liber Monetae compiled on March 15, 1317 (and preserved in the first two sheets of the manuscript) its author, the notary Salvi Dini de Florentia, recounted how he had succeeded in fully reconstructing the records as far back as 1303 (Bernocchi 1974: 2). Dini noted that for the first 50 years of the mint’s existence (1252–1303) he had been unable to find any books or documents that named the mint masters.37 He was, however, able to compile a list of 72 segni belonging to the early period. These segni were based on a previous inquiry made earlier in 1317. This inquiry had been based on three sources that were no longer available to Dini. The first was the missing liber sive quaternus, the semesterial records of the mint written by the notaries of the mint up to 1303. They were later incorporated in the Memoria or deeds of the mint. Secondly, there were oral accounts of several functionaries who had worked in the mint during this early period. Thirdly, the list of 72 segni was supposedly based upon examination of the coins themselves but errors in their description by Dini showed he was using an earlier list (Bernocchi, 1974: 20–23). Consequently, the Dini-‘Zecca’ list, as it may be called, became the authoritative reference for the early segni (1252–1303). It was published for the first time in 1761 when Ignazio Orsini reproduced the entire text of the Liber Monetae in his history of the Florentine mint (Orsini 1760). Orsini noted errors in the original ‘Zecca’ list but reproduced the original drawings of the segni with their inaccuracies as they had been incorporated in the Liber Monetae by Salvi Dini in 1317. Orsini’s work on the Florentine mint was barely superseded in 1930 by the catalogue of the ‘Toscana-Firenze’ region appearing in Volume 12 of the Corpus Nummorum Italicorum. It contained no new information on the earlier segni; indeed only nine of the 48 segni drawn by Orsini were inserted. The work provided little new information on the earlier period. It noted some varieties but presented no systematic analysis of the material. In 1952 Carson Simpson published a seminal article which for the first time outlined the problems of researching the early-period segni of 1252–1303 (Simpson 1952: 113–155). The article questioned the status of the Liber monetae as an accurate source, and suggested possible serious shortcomings in the

original list. It noted that a more comprehensive study of the segni and mint masters, which consulted external source material, both archival and coins themselves, was needed. This was largely accomplished in Mario Bernocchi’s five-volume study of the Florentine mint published between 1974 and 1985. Bernocchi made an extensive study of the mint’s archive and related documents preserved in the Archivio di Stato in Florence. He also studied a large number of florins in major collections, in Italy and elsewhere. The result was a detailed catalogue with more than 800 gold and silver coin types struck by this city between 1252 and 1533. With regard to the earliest period of the mint’s activity (1252–1303), Bernocchi presented a comprehensive list with a large number of new types. These he divided into two major groups. The first comprised 101 segni dated to 1252–1303 (as against the original 72 mentioned in the Liber Monetae). To these he added a second group consisting of another 54 segni appearing on the ‘small florin’ fiorino stretto type which was struck until 1421.38 These coins he dated to the period 1252–1421, uncertain if they were the earlier 1252–1303 or post 1303–1421 types. The Segni of the Akko Hoard: The 30 fiorini of the Akko Marina hoard bear 20 different segni, one of which seems to be a previously unpublished type. The coins divide into several groups according to periodisation of the segni established by Bernocchi and the Logge dei Banchi hoard (Bernocchi 1976: 107–108; Lenzi 1978): a) One coin belongs to the earliest group of florins without segni struck in the first year of the mint’s establishment in 1252. Series I without Segni

No. 1

b) Two coins bear one segno used in the period

ca.1252/3–60.

Series III with cluster of three pellets

Nos. 2 and 3

c) One coin bearing a segno used in the period 1260–7. Series III with 7-petal flower

No. 4

37 ‘videntibus et audientibus quod, de dominis et officialibus et signis ditte monete, qui a die ipsius initiate, fatte et ordinate monete usque in diem presentem fuerunt et fatti sunt, nulla in communi Florientie supraditto reperitur memoria, que singulis ipsos volentibus pateat’ (Bernocchi 1974: 1–2). The names of the mint masters for two terms 1280–1281 (Guido Cambi and Tedicio Manelli) and 1286–1287 (Coppo Guiseppi and Ticio Manovelli) were added to the Liber Monetae at the beginning of the fifteenth century, but without the mint symbols (Bernocchi 1974: 10). 38 These coins had a diameter of 20mm. From 1422 to 1533 Florence struck a larger gold florin type with a diameter of 22mm.

227

Robert Kool

d) Twenty-one coins bear 15 segni that were used between 1252 and 1303. Pellet between legs of saint

No. 5

Ring with pellet

Nos. 6 and 7

Six-pointed star

No. 8

Crescent with pellet

Nos. 9 and 10

Five-petal flower

No. 11

Key

No. 12

Voided cross with four pellets

Nos. 13 and 14

Four-petal flower with stem

No. 15

Vine leaf

Nos. 16,17 and 18

Two oak leaves with stem

No. 19

Hammer

No. 20

Shell

No. 21

Pine cone

Nos. 22 and 23

Trefoil with stem

No. 25

Hook

No. 26

Wheat sheaf

g) Lastly, two coins carry an identical but possibly unpublished segno resembling a ‘poppy’ which for purposes of analysis can be dated to 1252–1421. This design resembles the papavero dated by Bernocchi to 1251–1421 but differs substantially from it with regard to the size of the upper cone, body and stem. In addition, Bernocchi lists other fiorini stretti bearing a poppy as the emblem used by Nerone del Nero in the first semester of 1347. Bernocchi also lists a fifteenth-century grosso and several soldini with a poppy as the mint master emblem but also this mark hardly resembles the above ‘poppy’ mark.40

‘Poppy’

Nos. 29 and 30

The above data show that the majority of the coins are dated exclusively to within the early period 1252–1303. Four of these can be firmly dated to the early years of the florin’s existence (1252 to the late 1260s). The remaining coins bear segni which cannot be exactly dated bearing in mind the typology and fabric of the coins as outlined above but all of them can be dated to the early period (1252–1380). More important, none of them can be dated with certainty to the period after 1303. Only one coin bears a segno (wheat sheaf), which was struck during the early period but was then re-used by a mint master in 1344. Another two coins have an identical emblem (segno 7) which Bernocchi was unable to date precisely and classified it to the general period 1252–1421. Lastly, two coins carry a design (‘poppy’) which does appears neither among the group of segni that Bernocchi was able to describe and date accurately (1303–1380) nor among those in the general category 1252–1421. We presume that this unknown segno belonged to the earlier period. In sum, the segni clearly indicate that the hoard dates to the second half of the thirteenth century. More important, the hoard’s provenance in Frankish Akko contributes to a refinement of the dating. Frankish Akko with its large Italian merchant quarters was captured and destroyed by the Mamluks in May 1291. The terminus ante quem of these early segni can thus be taken back from 1303, a date based on Dini’s findings in 1317, to at least a dozen years earlier. The presence of three other Byzantine and North African gold coins (seventh to twelfth centuries) found with the florins, all considerably pre-dating the fall of Akko (1291), and the absence of any European or Muslim

e) Two coins bear a segno which Bernocchi dates to 1252–1421.

Segno 7

No. 24

Nos. 27 and 28

f) One coin carries a segno that may belong to the early period (1252–1303) but also resembles a type appearing on a coin struck in 1344. A gold florin with the same privy mark appeared also in the Aleppo hoard dated to the 1290s (see below).39

39 For further examples of two pre-1303 coins and one from 1344 bearing this privy mark (Toderi and Vannel 2005: Tables 48–49, nos. 108–109 and Table 74, no. 493). 40 For gold florins dated 1252–1421 see Bernocchi (1975: 46, nos. 519–520); for the gold florins dated to 1347 see Bernocchi (1975: 186, nos. 1543– 1544) and Toderi and Vannel (2005: 46, no. 500); for fifteenth-century grossi and solidini see Bernocchi (1975: 396, nos. 2849–2855).

228

Coin Finds from the Akko Marina

gold post-dating the 1290s, further indicate an early dating of these Florentine gold coins. It therefore seems extremely unlikely that the coins could have been lost after 1291, following the capture of Akko, and the systematic destruction of the town as a commercial center and port by the Mamluks.

Comparison with the Aleppo Hoard At the start of my research on the Akko Marina hoard the only relevant numismatic references were the brief comments by Philip Grierson on a large hoard of similar coins which had been found in Aleppo in Syria, in the mid-1950s. The hoard was subsequently dispersed in Europe and the United States. Since then to the best of my knowledge, no other hoard of this kind has ever been noted or published. During a sojourn at Oxford University in 1996 I contacted the late Professor Grierson at Cambridge University and asked him if he had received any additional information on these enigmatic coins. In reply I received his unpublished notes and a draft for an article on the Aleppo hoard. These substantially enlarge our knowledge of this hoard and form the basis for the comparisons below.41 The Aleppo hoard contained a sizeable quantity of gold ca. 2.2kg or 629 coins. The majority of these

were florins (ca. 600). Apparently the coins were found during repairs on a wall of an old house in the city of Aleppo around 1953–1955.42 Grierson noted that the hoard contained the following coins. He tentatively dated this hoard to the plunder of Akko (1291) by the Mamluk Sultan al Ashraf Khalil (AH 689–893/1290–1293 CE), a conclusion that was subsequently adopted by historians and numismatists dealing with the period.43 Re-examination of the surviving unpublished documentation of the hoard seems to confirm its dating. In a letter to Philip Grierson, the French archaeologist and numismatist Henri Seyrig discussed in detail the mint symbols and dates of a large part of the ca. 600 florins in the hoard that he examined in Beirut in April 1955, before its dispersal around July 1955.44 He studied 110 coins and recorded 40 different segni, all of them dating to the early period (1252–1303). From this it can be inferred with a high degree of certainty, assuming Seyrig’s lot was chosen at random, that the entire florin complement belonged to the earliest group. Another 35 coins of the dispersed Aleppo hoard bought by Grierson and the American Numismatic Society between 1955 and 1958 yielded an additional eight to ten segni all belonging to the early florin types.45 The 110 to 120 fully documented coins (assumed to be a random sample) from the total of about 600

Table 3. The Aleppo hoard State

Ruler

Denomination

Quantities of Coins

France

Philip IV, the Fair (1285–1314)

Petit royal assis

8

Venice

Giovanni Dandolo (1280–1289)

Ducat

4

Venice

Pietro Grandenigo (1289–1311)

Ducat

17

Florence

Commune (1252–1303)

Florin

ca. 600

41 Letter to the author from Grierson, dated September 5, 1996. 42 Details on the provenance of the hoard were relayed to Mr. Leopoldo Cancio of the International Bank for Reconstruction and Development by Mrs. Dorsey Stephens, the wife of the Middle East Bank representative in Beirut in three letters between May 8 and August 16, 1955. The information was subsequently sent by Cancio to Henry Grunthal of the American Numismatic Society in a letter dated September 29, 1955. My thanks to Alan Stahl who kindly provided me with this unpublished material from the ANS archives in January 2000. 43 For the ‘plunder of Acre’ explanation see Lane and Mueller (1985: 276–277; Jacoby (1986: 411–412); Spufford (1988: 177); Metcalf (1995: 323). 44 Letter from Seyrig to Grierson, April 13, 1955. The following December, the New York firm Coin Galleries offered “Gold Coin from the Aleppo Hoard’ in both Fine and Very Fine condition” (Coin Galleries FLP 6 [December 1955: 19]). The number available was not specified but there were about 200 coins of which the ANS bought 20 (Miles to Grierson 2 February 1956). The following year the same firm offered gold ducats of Pietro Gradenigo (1289–1310) in EF condition, (Coin Galleries FPL 8 (Summer 1956 after item 245). These were presumably also from the Aleppo Hoard. Another florin, described as EF, was in an earlier Coin Galleries auction (July 11, 1955 lot 226). It was catalogued as “CNI 62, 392 var., first semester 1384” but this was incorrect and the coin could have come from Aleppo. I am grateful to William R. Day for this data and also to Alan Stahl for information from the ANS archives. 45 A large number of the coins were subsequently put up for sale in coin markets in Western Europe and the United States. Grierson bought a number of Aleppo gold florins in Rome in early 1955. He bought 11 specimens in Brussels in January 1956 and three (?) other coins over the next two years in London. In addition to the 20 Aleppo florins acquired in 1955 (see above note), the ANS received a gift of another specimen in 1956. All of these coins have segni dating exclusively to the early period. This further strengthens the impression that the Aleppo hoard contained only early-period florins (1252–1303).

229

Robert Kool

constitute a statistically significant group (c. 20%). All of the documented coins contain only earlyperiod segni which make them extremely suitable for comparison with the Akko Marina hoard. Such comparison shows that 14 of the 21 segni of the Akko Marina hoard reappear among the 48 recorded marks of the Aleppo hoard. Does this relatively strong correlation between segni populations possibly point to a connection between the hoards? The question is difficult to answer unequivocally. The rarity of the find, the chronological similarity and the geographical proximity of the coins point in this direction.

event. There is an interesting parallel in the account of the treasure salvaged at Trapani, mentioned briefly above, preserved in the reconstructed registers of the Angevin chancery. During a violent storm in November 1270 parts of the French fleet, returning from the unsuccessful Crusade to Tunis led by Louis IX, was destroyed in the Sicilian port of Trapani. Witnessing the disaster, Charles of Anjou, King of Sicily (1266–1285) nominated a royal commission to draw up an inventory of all the unclaimed wreckage lying in the harbor. The inventory itself was lost but a list detailing the finds of coins and objects of precious metals, super naufragio Trapani, has survived (Carolus-Barré 1976). According to the list more than 200 gold coins were recovered. Of these the majority, 137 coins, were florins. In sum, the evidence of the Akko Marina hoard together with the above documentary information imply that by the late 1270s imports of high-quality gold florins began to reach Akko, supplementing the imitation dinars (bezants) struck in the Latin Kingdom until 1260 (Metcalf 1995: 44–45). This abundant flow of florins was presumably instrumental in undermining the last attempts to mint gold in the Frankish east after the 1260s (for example the rare Agnus Dei bezant). Florins continued to circulate locally till the last days of Crusader settlement, and also beyond in the fourteenth century, as historical sources attest.46 The other high-quality gold coin minted in the West, the Venetian ducat, only became available in 1285, too late to make a serious impact on the economy of what remained of the kingdom of Jerusalem. The discovery of this small hoard thus considerably enlarges our understanding of the last chapter of the kingdom’s monetary history prior to its fall in 1291.

The Akko Marina Hoard and the End of Crusader Akko On Friday May 18 1291 after a siege of several weeks Frankish Akko fell to the Mamluk army. In the succeeding mêlée soldiers and citizens, desperate to escape the enemy, crowded into the harbor. Eyewitnesses like the anonymous Templar of Tyre and accounts written several decades later related that a few, mostly noble ladies and merchants, succeeded in escaping by bribing owners of small rowing boats with jewelry and gold to be ferried off-shore to ships sailing for Cyprus and Tyre (Templar of Tyre: 111–117; Ludolph von Suchem 1897: 58–59). Many, however, drowned with their precious possessions in the stormy sea. Whether the Akko Marina hoard is related to this event remains in the realm of conjecture. The numismatic evidence presented above clearly suggests that this group of coins constituted a hoard or part of a hoard of coins which reached Akko in the last decades of Frankish rule, and presumably were dispersed on the seabed due to a catastrophic

46 Schlumberger (1878: 492) noted that a fifteenth-century Arab chronicle recorded the circulation of florins in Galilee until the end of the fourteenth century; Ashtor (1992: 178–179), observed that European travelers in the fourteenth century exchanged gold florins for Mamluk dirhams to pay for their expenses. The Muslim ruler Omar Beg (1341–1348), of the emirate of Aidin on the Ionian coast, even struck an imitation gold florin at Ephesus. See Auction Sale Catalogue of the John J. Slocum Collection of Coins of the Crusades (1997: 100–101, lot 928) now in the Fitzwilliam Museum (Grierson bequest).

230

Coin Finds from the Akko Marina

Table 4. Comparison between the Akko Marina hoard and the Aleppo hoard segni

Mint mark

Symbol

Akko Marina Hoard

Aleppo Hoard

-

1

-

Series III with cluster of three pellets

2

-

Pellet between legs of saint

2

1

Seven-leaf

1

2

Ring with pellet

2

1

Six-pointed star

1

1

Crescent with pellet

2

1

Five-petal flower

1

-

Key

1

9

Voided cross with four pellets

2

3

Four-petal flower with stem

1

1

Vine leaf

3

1

Two oak leaves with stem

1

-

Hammer

1

7

Shell

1

7

Pine cone

2

-

Trefoil with stem

1

2

Hook

1

3

Segno 7

2

-

Poppy

2

?

Wheat sheaf

1

Series I

231

1

Robert Kool

References

Bijovsky G. and Berman A. 2008. The Coins. In Y. Hirschfeld and O. Gutfeld (eds.). Tiberias.

Abulafia D. 1981. Southern Italy and the Florentine Economy:1265–1370. Economic History Review Series 2, 34: 377–388.

Excavations in the House of the Bronzes. Final Report, Vol. 1: Architecture, Stratigraphy and Small Finds (Qedem 48). Pp. 63–102. Jerusalem.

Abulafia D. 1982. Crocuses and Crusaders: San Gimignano, Pisa and the Kingdom of Jerusalem. In B.Z. In B.Z. Kedar, H.E. Mayer, and C. Smail (eds.).

Bijovsky G. 2002. A Hoard of Byzantine Solidi from Bet She’an in the Umayyad Period. Revue Numismatique 158: 161–227.

Outremer: Studies in the History of the Crusading Kingdom of Jerusalem. Pp. 227–243. Jerusalem.

Bijovsky G. Forthcoming. Coins and Two Byzantine Gold Hoards from Excavations at Shoham ‘Atiqot. Jerusalem.

Abulafia D. 1988. The Levant Trade of the Minor Cities in the Thirteenth and Fourteenth Centuries: Strengths and Weaknesses. In B.Z. Kedar and A.L. Udovitch (eds.). The Medieval Levant. Studies in Memory of Eliyahu Ashtor (1914–1984). (Asian and African Studies 22). Pp. 183–202. Jerusalem.

Bloch M. 1967. The Problem of Gold in the Middle Ages. In Land and Work in Medieval Europe (M. Bloch transl.). Pp. 186–229. London. Brucker G. 1998. The Golden Age, 1138-1737. Berkeley.

Ariel D.T. 2000. The Coins from the Synagogue at Korazim. In Z. Yeivin, The Synagogue at Korazim: The 1962–1964, 1980–1987 Excavations (IAA Reports 10). Pp. 33–49. Jerusalem.

Carolus-Barré L. 1976. Objet précieux et monnaies retrouvés dans le port de Trapani en 1270, dont 21 écus d’or de Saint Louis. Revue Numismatique 18: 115–118.

Ashtor E. 1971. Les métaux précieux et la balance des payments du Proche-Orient à la Basse Epoque. Paris.

Casey P.J. 1986. Understanding Ancient Coins. An

Ashtor E. 1992. The Cost of Living in Eretz-Israel during the Middle Ages. In Y. Drori (ed.). Palestine in the Mamluk Period. Pp. 172–188. Jerusalem.

London.

Introduction for Archeaeologists and Historians. Castellani G. 1937. Il ripostiglio di Pisa Bolletino

d’Arte 30: 476–484.

Balard M. (ed.). 1983. Notai Genovesi in Oltremare:

Atti Rogati a Cipro da Lamberto di Sambuceto (11 ottobre 1296–guigno 1299). Genoa.

Cipolla C.M. 1956. Money, Prices and Civilization in the Mediterranean World. Princeton.

Balog P. and Yvon J. 1958. Monnaies à légendes arabes de l’orient latin Revue Numismatique 6e serie 1:133–168.

Cipolla C.M. 1983. The Monetary Policy of Fourteenth-Century Florence. Berkeley. Cook B. 1999. Foreign Coins in Medieval England. In L. Travaini (ed.). Local Coins, Foreign Coins: Italy and Europe, 11th– 15th Centuries. Pp. 231–284. Milan.

Balletto L. 1986. Fonti notarili genovesi del secondo Duecento per la storia del Regno latino di Gerusalemme. In G. Airaldi and B.Z. Kedar (eds.). I

(Collana storica di fonti e studi, diretta da G. Pistarino 48). Pp. 197, 255–274. Genoa.

comuni italiani nel Regno crociato di Gerusalemme

Day J. 1968. La circulation monétaire en Toscane en 1296. Annales, Economies, Societes, Civilisations 23:1054–1066.

Berman A. 1976. Medieval Hoards from the Southern Wall Excavations and their Lessons. Israel Numismatic Society Newsletter 5:76–78. (Hebrew).

Day W.R. 1997. The Monetary Reforms of Charlemagne and the Circulation of Money in Early Medieval Campania. Early Medieval Europe 6: 25–45.

Bernocchi M. 1974–1985. Le monete della repubblica fiorentina, Vols. I–IV. Florence.

Day W.R. 2002. The population of Florence before the Black Death: Survey and Synthesis. Journal of Medieval History 28: 93–129.

Bijovsky G. 2012. Gold Coin and Small Change: Monetary Circulation in Fifth–Seventh Century Byzantine Palestine (Polymnia Numismatica antica e

Day W.R. Forthcoming. Florence before Dante: the Early Development of the Florentine Economy, c. 1115–1265. Leiden.

medievale. Studi 2). Trieste.

232

Coin Finds from the Akko Marina

Desimoni C. 1884. Chartes: Actes passés à Famagouste de 1299 à 1301 par devant le notaire génois Lamberto di Sambuceto. Archives de l’Orient Latin II. Pp. 3–120. Paris.

Desimoni C. 1893. Chartes: Actes passés à Famagouste (1300) par devant le notaire génois Lamberto di Sambuceto. Revue de l’Orient Latin I. Pp. 275–353. Paris.

DOC 1: A.R. Bellinger. Catalogue of the Byzantine Coins in the Dumbarton Oaks Collection and in the Whittemore Collection 1: Anastasius I to Maurice. 491–602. Washington, D.C. 1966. DOC 2: P. Grierson. Catalogue of the Byzantine Coins in the Dumbarton Oaks Collection and in the Whittemore Collection 2: Phocas to Theodosius III. 602–717. Washington, D.C. 1968. DOC 3: P. Grierson. Catalogue of the Byzantine Coins in the Dumbarton Oaks Collection and in the Whittemore Collection 3: Leo III to Nicephorus III. 717–1081. Washington, D.C. 1973. Dubois H. 1976. Les foires de Chalon-sur-Saône et le commerce dans la vallée de la Saône à la fin du Moyen Age. Paris. Duplessy J. 1995. Les Trésors monetaires médiévaux et modernes découverts en France, Vol. 2. Paris. Galili E., Sharvit J., Dahari U., Silberstein N., Finkielsztejn G., Stern E., Kool R. and Kahanov Y. 2002. Akko Port, Underwater Surveys. Hadashot Arkheologiyot 114: 12–15. Galili E., Rosen B., Zviely D., Silberstein N. and Finkielsztejn G. 2010. The Evolution of Akko Harbor and its Mediterranean Maritime Trade Links. Journal of Island and Coastal Archaeology 5: 192–211. Galili E. This volume a. The Akko Marina Archaeological Project. Introduction. Goitein S.D. 1965. The Exchange-rate of Gold and Silver Money in Fâtimid and Ayyûbid Times. A Preliminary Study of the Relevant Geniza Material.

Journal of the Economic and Social History of the Orient 8: 1–46.

Goitein S.D. 1967. A Mediterranean Society. The

Jewish Communities of the Arab World as portrayed in the Documents of the Cairo Geniza, Vol. I. Economic Foundations. Berkeley. Goldthwaite R. and Mandich G. 1994. Studi sulla moneta fiorentina (secoli XIII-XVI). Florence.

Goldthwaite R. 1980. The Building of Renaissance Florence: An Economic and Social History. Baltimore. Grierson P. 1954. The Debasement of the Bezant in the Eleventh Century. Byzantinische Zeitschrift 47(2):379–394. Grierson P. 1957a. La moneta veneziana nell’ economia mediterranea del Trecento e Quattrocento in La civiltà veneziana del Quattrocento. Pp. 77–97. Venice. Grierson P. 1957b. The Coin List of Pegolloti in Studi in onore di Armando Sapori, Vol. I. Pp. 483–492. Milano. Grierson P. 1965. The Interpretation of Coin Finds (I): Presidential Address Delivered at the Annual General Meeting, the Royal Numismatic Society, June 16, 1965. Numismatic Chronicle 5: I–XII. Grierson P. and Blackburn M. 1986. Medieval

European Coinage with a Catalogue of the Coins in the Fitzwilliam Museum, Cambridge, Vol. I: The Early Middle Ages (5th–10th centuries). Cambridge.

Grierson P. and L. Travaini 1998. Medieval European

Coinage 14. With a Catalogue of the Coins in the Fitzwilliam Museum, Cambridge. Italy (III) South Italy, Sicily, Sardinia. Cambridge.

Hazard H.W. 1952. The Numismatic History of Late Medieval North Africa (Numismatic Studies 8). New York. Heidemann S. 1998. The Merger of Two Currency Zones in Early Islam. The Byzantine and Sasanian Impact on the Circulation in Former Byzantine Syria and Northern Mesopotamia. Iran 36: 95–112. Heidemann S. 2002. Die Renaissance Der Stadte in Nordsyrien Und Nordmesopotamien: Stadtische Entwicklung Und Wirtschaftliche Bedingungen in Ar-Raqqa Und Harran Von Der Zeit Der Beduinischen Vorherrschaft Bis Zu Den Seldschuken. Leiden. Heidemann S. 2009. Economic Growth and Currency in Ayyubid Palestine. In S. Auld and R. Hillenbrand (eds.). Ayyubid Jerusalem: The Holy City in Context. Pp. 276–300. London. Hoshino H. 1980. L’arte della lana in Firenze nel Basso Medioevo: il commercio della lana e il mercato dei panni fiorentini nei secoli XIII-XV. Florence. Hunt E.S. 1994. The Medieval Super-Companies:

A Study of the Peruzzi Company of Florence. Cambridge.

233

Robert Kool

Ilisch L. 1993. Sylloge Nummorum Arabicorum Tübingen. Palästina IV a Bilād aš Šām I. Tübingen.

Levy S. and Mitchell H.W. 1965–1966. A Hoard of Gold Dinars from Ramlah. Israel Numismatic Journal 3: 37–66.

Irwin R. 1980. The Supply of Money and the Direction of Trade in Thirteenth Century Syria. In P.W. Edbury and D.M. Metcalf, Coinage of the Latin

Lopez R. 1953. Settecento anni fa: il ritorno all’ oro nell’ occidente duecentesco. Rivista Storica Italiana 65: 19–55, 161–198.

East. The Fourth Oxford Symposium on Coinage and Monetary History. (BAR International Series 77). Pp.

Lopez R. 1956. Back to Gold, 1252. Economic History Review Second Series 9: 219–240.

67–76. Oxford.

Ives H. 1952. The Design of Florentine Florins as an Aid to their Dating. American Numismatic Society

Lopez R. and Airaldi G. 1983. Il più antico manuale italiano di pratica della mercatura. Miscellanea di studi storici 2, Fonti e Studi 38. Pp. 99–117. Genoa.

Museum Notes 5:103–112.

Jacoby D. 1986. A Venetian Manual of Commercial Practice from Crusader Acre. In G. Airaldi and B.Z. Kedar (eds.). I comuni italiani nel regno crociato di Gerusalemme. Atti del colloquio “The Italian Communes in the Crusading Kingdom of Jerusalem (Jerusalem, May 24-May 28, 1984). Pp. 403–428.

Meshorer Y., Bijovsky G. and Fischer-Bossert W. 2013. Coins of the Holy Land: The Abraham

and Marian Sofaer Collection at the American Numismatic Society and the Israel Museum. D.

Genoa.

Hendin and A. Meadows (eds.). New York.

Jacoby D. 1998. Migration Trade and Banking in Crusader Acre. BALKANIA KAI ANATOLIKH

Messier R.A. 1974. The Almoravids: West African Gold and the Gold Currency of the Mediterranean Basin. Journal of the Economic and Social History of the Orient 17: 31–47.

Kleinhenz C. (ed.). 2004. Medieval Italy. An Encyclopedia, Vol. II. New York.

Ludolph von Suchem. 1897. Ludolph von Suchem’s Description of the Holy Land. In Paletine Pilgrims’ Text Society Library, Vol. 12. London.

Kool R. 2006. A Hoard of 13th Century Gold Florins from the Medieval Harbor of Acre. Numismatic Chronicle 166:301–320 and Pls. 57–61.

Luzzati M. 1971. Giovanni Villani e la Compagnia dei Buonaccorsi. Rome.

MEΒΑΛΚΑΝΙΑ ΚΑΙ ΑΝΑΤΟΛΙΚΗ ΜΕCOGEIOSΟΓΕΙΟΣ (12oc-17oc AIWNEΑΙΩΝΕC). Pp. 105–119. Athens.

Metcalf D.M. 1995. Coinage of the Latin East. London.

Kool R. 2007. Coin circulation in the villeneuves of the Latin Kingdom of Jerusalem: the Cases of Parva Mahumeria and Bethgibelin. In P. Edbury and S. Kalopissi-Verti (eds.). Archaeology and the Crusades. Pp. 91–112. Athens.

Morrisson C., Brenot B., Callu J.P., Barrandon J.N. Poirier J. and Halleux R. 1985. L’or monnayé I. Purification et altérations. de Rome à Byzance. Paris.

La Roncière C.M. de. 1973. Un changeur florentin

Nohejlova-Prátová E. (ed.). 1971–1972. ‘Les Commencements de la grosse monnaie et de la monnaie d’or en europe centrale 1250–1350’ = Numismatický Sborník, 12 Prague.

de Trecento, Lippo di Fede del Sega 1285 env.- 1363 env. Paris.

La Roncière C.M. de. 1976. Florence, centre économique régional au XIVe siècle. 5 vols. Aix-enProvence.

Orsini I. 1760. Storia delle monete della Repubblica Fiorentina. Florence.

Lampinen P.1987. Numismatic Evidence. In J.A. Blakely, Caesarea Maritima: the Pottery and Dating of Vault I, Vol. IV. Pp. 143–145. Lewiston N.Y.

Paoli C. and Piccolomini E. (eds.). 1871 (repr. 1968).

Lane F.C. and Mueller R.C. 1985. Money and Banking in Medieval and Renaissance Venice.

Pavoni R. (ed.). 1982. Notai Genovesi in Oltremare:

lettere volgari del secolo XIII scritte da Senesi. Bologna.

Atti Rogati a Cipro da Lamberto di Sambuceto (6 luglio 1301–27 ottobre 1301) Genoa.

Baltimore and London.

Lenzi L. 1978. Il ripotiglio di monete auree scoperto in Pisa sotto le logge dei Banchi. Pisa.

Pavoni R. (ed.). 1987. Notai Genovesi in Oltremare: Atti Rogati a Cipro da Lamberto di Sambuceto (gennaio-agosto 1302). Genoa.

234

Coin Finds from the Akko Marina

Pegolotti Francesco Balducci 1936. La pratica della mercatura, A. Evans (ed.). Cambridge, Mass. Plough G., Oldenburg E., Hammershaimb E., Thomsen R., Løkkegaard F. and Riis P. 1969. Les monnaies. Hama: Fouilles et Recherches 1931–1938, Vol. IV, 3: Les petits objets médiévaux sauf les verreries et poteries. Pp. 143–71. Copenhagen. Polonio V. (ed.). 1982a. Notai Genovesi in Oltremare:

Travaini L. 1977. Mint Organization in Italy between the Twelfth and Fourteenth Centuries: a Survey. In N.J. Mayhew and P. Spufford (eds.). Later Medieval

Mints: Organisation, Administration, Techniques. 8th Oxford Symposium on Coinage and Monetary History (BAR International Series 389). Oxford. Travaini L. 1991. Genova e i tarì di Sicilia. Rivista

italiana di numismatica 93:187–194.

Atti Rogati a Cipro da Lamberto di Sambuceto (3 Luglio 1300- 3 Agosto 1301). Genoa.

Travaini L. 1995. La monetazione nell’ Italia normanna. Rome.

Polonio V. (ed.). 1982b. Notai Genovesi in Oltremare: Atti Rogati a Cipro da Lamberto di Sambuceto (6 luglio 1301- 27 ottobre1301). Genoa.

Travaini L. 2001. The Normans between Byzantium and the Islamic World. Dumbarton Oaks Papers 55: 179–196.

Templar of Tyre: The Templar of Tyre: Part III of the ‘Deeds of the Cypriots’. 2003. P. Crawford

Travaini L. 2003. Monete, Mercantie e Matematica. Rome.

(ed.). Crusade Texts in Translation 6. Aldershot.

Richards J.F. (ed.). 1983. Precious Metals in the Later Medieval and Early Modern Worlds. Durham, United Kingdom. Sapori A. 1955–1967. Studi di storia economica medievale-secoli XIII, XIV, XV. Firenze. Schlumberger G. 1878. Numismatique de l’Orient latin. Paris. Simpson C. 1952. Mint Officials of the Florentine Florin. American Numismatic Society Museum Notes 5: 113–155. Spufford W. 1988. Money and its Use in Medieval Europe, Cambridge. Spufford W. 2006. The First Century of the Florentine Florin. Rivista Italiana di Numismatica 107: 421–436. Stahl A. M. 2000. Zecca: The Mint of Venice in the Middle Ages. Baltimore and London. Tal O., Kool R. and Baidoun I. 2013. A Hoard Twice Buried? Fatimid Gold from Thirteenth-Century Crusader Arsur (Apollonia-Arsuf). Numismatic Chronicle 173:261–292; Pls. 43–49.

Usamah ibn Munqidh (1095–1188) 1929. Memoirs

of an Arab-Syrian gentleman; or, An Arab knight in the Crusades: memoirs of Usamah ibn-Munqidh (Kitab al-‘itibar). (P.K. Hitti transl.). New York.

Walker T. 1983. The Italian Gold Revolution of 1252: Shifting currents in the pan-Mediterranean Flow of Gold. In Precious Metals in the Later Medieval and Early Modern Worlds, F. Richards (ed.). Pp. 29–52. Durham, United Kingdom. Wasserstein D. 1998. The coins in the Golden Hoard from Tiberias. ‘Atiqot 36:10–14. Watson A. 1967. Back to Gold – and Silver. Economic History Review Second Series 20: 1–34. Villani Giovanni1990–1991. Nuova Cronica, a cura di Giuseppe Porta, 3 vols. (I. Libri I–VIII; II. Libri IX-XI; III. Libri XII-XIII). Parma. Yvon J. 1966. Monnaies et sceaux de l’Orien Latin. Revue Numismatique 6: 89–107.

Zibaldone da Canal, manoscritto mercantile del secolo XIV. A. Stussi (ed.). Fonti per la Storia di Venezia 5. 1967. Venice.

Toderi G. and Vannel F. 2005. Monete italiane del Museo nazionale del Bargello, Vol. 2: Firenze: Republica. Florence.

235

Robert Kool

CATALOGUE 1. Single Finds (see also Figs. 1-3) The coins below are arranged chronologically. Coins are bronze unless otherwise noted. Coins bearing an asterisk have photographs in Figs. 1–3.

No Barge Item

IAA No.

Diam. Weight (mm) Axis (gm)

Obverse

Reverse

Date (CE)

Mint

Ref.

Nero (?) (54– 68 CE) 1* 31

2

74226 16.27

30

¯

SC within wreath

Head to r.

Cf. Waage Antioch 1952 33–34, No. 348.

Caracalla (198–217 CE) 2* 35

3

74217 8.75

25

¯

AVR[- - -]AV[- - Bull of Serapis -] Bust r., with r., in r. field, wreath. torch

Akko

CHL 2013:12, No. 167.

Akko

CHL 2013: 15, No. 229.

Severus Alexander ( 222–235 CE)

3* 31

5

74229 6.71

22

¯

[IMP C]AE M AV[R SEV ALEXANDER AVG] Bust r., with shield against shoulder.

COLO [-PTOLEM] Emperor in military dress, riding a horse; under foot of horse: caduceus

Philip I (?) (244–249 CE) 4

31

1

74225 7.96

24

¯

Bust r.

Figure st., holding ?

Cf. Kadman Akko (?) 1961:134, No. 212.

5th c. 5

35

1

74215 0.46

10

Illegible

Illegible

Justinian I (?) (527–565 CE)

– Dodecanummium 6* 35

7

31

5

4

74219 4.60

74228 2.39

18

15

¯

¯

Bust r.,

Bust

I †B

Alexan

In ex.: AΛEΞ

-dria

I †B

Alexan -dria

236

Cf. DOC 1:157, No. 274.I. Cf. DOC 1: 157, No .274.1.

Coin Finds from the Akko Marina

Byzantine (6/7th c CE) – Follis 8

36

25

74230 2.64

18

9

35

7

74221 2.01

19

¯

M

Bust(?)

Cross above.

Bust(?)

Illegible.

Constans II (641-668 CE) – Gold Solidus

10* 36

2

72847 4.31

20

¯

[dN CONStAN– tIN VS PPAV] Bust facing,, beardless, wearing, chlamys with tablion ornamented by four pellets, and crown with cross on circlet;

VICTORIA AVG I

DOC 2, 2: 641– Constan- 421, Cross potent on 646 tinople No. Ij.I base and three steps. In ex.: CONOB

in r. hand, scepter.

Constans II, – Dodecanummium Emperor st., facing; 11 36

50

74231 8.78

19

¯

in r. hand long staff; in l. globe with cross potent.

I B Between,

cross potent on 641/2– Alexan globe. 645/6 -dria In ex.: AΛE

Umayyad, Anonymous (697–750 CE) – Fals I In margins: ‫ﺿﺮب‬... 12 35

2

74216 3.53

20

In center: ‫اﻟﻠﺔ اﻻ‬ ‫اﻟﻠﺔ اﺣﺪه‬ 237

In center:

... ‫وﺣﺪة‬

Cf. DOC 2, 2: 468, No.106.1.

Robert Kool

In center:

In center: ‫ﻣﺤﻤﺪ‬

‫ﻻاﻟﺔ‬ 13 35

8

74222 4.29

17

‫رﺳﻮل‬ ‫اﻟﻠﺔ‬

‫اﻻﻟﻠﺔ‬ ‫وﺣﺪة‬ In center below: *

In center:

AlUrdûn

In legend around: ‫اﻟﻠﺔ ﺿﺮب‬

‫ﺑﺴﻢ‬

‫ﺑﺎﻻردن‬ In center: pomegranate Legend around:

14* 35

4

74218 3.33

18

‫ﻻاﻟﺔ‬ ‫اﻻﻟﻠﺔ‬

‫ﻣﺤﻤﺪ‬ ‫رﺳﻮل‬

‫وﺣﺪة‬

10

74224 1.65

18

Same

Same

16 35

6

74220 4.37

17

Illegible

Illegible

In center:

In center:

18 31

9

3

74223 3.43

74227 3.29

17

17

‫ﻻاﻟﺔ‬

‫اﻟﻠﺔ اﺣﺪ‬

‫اﻻﻟﻠﺔ‬

‫اﻟﻠﺔ اﻟﺼﻤﺪ‬

‫وﺣﺪة‬

‫ﻟﻢ ﯾﻠﺪ و ﻟﻢ ﯾﻮﻟﺪ‬

In wreath:

In center:

‫ﻻاﻟﺔ‬

‫ﻣﺤﻤﺪ‬

‫اﻻﻟﻠﺔ‬

‫رﺳﻮل‬

‫وﺣﺪة‬

‫اﻟﻠﺔ‬

In center: 19 36

1

74232 3.04

18

‫ﻣﺤﻤﺪ‬ ‫رﺳﻮل‬ ‫اﻟﻠﺔ‬

238

Cf. Ilisch 1993: 46, Nos. 547– 552.

‫اﻟﻠﺔ‬

15 35

17 35

Cf. Ilisch 1993:32, No. 331– 338.

Illegible

Walker 1956: 223, No. 724 (variant).

Coin Finds from the Akko Marina

Byzantine Empire Basil II and Constantine VIII (976-1025 CE) – Gold

Histamenon Nomismata

20* 41

11

73197 4.22

27

¯

BASILCCONSTANT IhSXPSREXREG ‘bR Facing busts of Basil NANTIhm II l., and Constantine VII r., wearing Facing bust of Christ Pantocrator stemma; Basil wears DOC 3, 2: square patterned 1005- Constanwearing stola and 621, No. loros and holds 25 koloboin; blessing tinople 6a.6. with r., holding in long-shafted cross; l. Gospels. Behind, above him suspended crown; cross nimbate with two crescents in the Constantine wears upper quarters. chlamys with tablion.

Muwahhids, Abu Muhammad ‘Abd-Al-Mu’min Ibn-’Ali (AH 524-528/1130-1163 CE) – Gold Dinar

In center: 21*

72846 2.24

21

¯

In center:

‫ﻻاﻟﺔ اﻻ‬

‫اﻟﻤﮭﺪى اﻣﺎم‬

‫اﻟﻠﺔ ﻣﺤﻤﺪ‬

‫اﻻﻣﺔ اﻟﻘﺎم‬

‫رﺳﻮل اﻟﻠﺔا‬

‫ﺑﺎﻣﺮ اﻟﻠﺔ‬

239

After Hazard AH 1952: 547/ Bijayah 143, No. 1152 447. CE

Robert Kool

CATALOGUE 2. Acre Marina Florin Hoard (see also Fig. 4) The coins below are gold fiorini (florins) arranged chronologically according to mint marks. They are mark as (H.) in the text. Coins bearing an asterisk have photographs in Fig. 4.

No

IAA No.

Wt. (gm)

Diam. Axis (mm)

Obverse

Reverse

Mint mark

Reference

Florence, c. 1252 – Fiorino Series I without symbols

*1

72830

3.48

21



+ FLOR– ENTIA fleur-de-lis

+. S. IOHA– NNES.B. John the Baptist nimbate, st. in front, in a coat of animal hair; in l. hand staff; r. hand raised in benediction



Florence, 1252–1260 – Fiorino Series III with small symbols + S:IOHA– NNES.B. Same but pearled nimbate Same

*2

72817

3.53

21



+ FLOR– ENTIA Same

*3

72831

3.47

20



Same

Bernocchi 1975: 9, No. 69.

Bernocchi 1975: Cluster of 10, No. 76. three pellets Same

Same

Florence, 1260–1267 – Fiorino Series III with small symbols

*4

72838

3.47

20



+. S. IOHA– + FLOR–ENTIA NNES.B Same Same

Small leaf

Bernocchi 1975: 12, No. 109.

Florence, 1252–1291 – Fiorino Series IV with symbols

*5

*6

72836

72832

3.47

3.50

20

20



+ FLOR–ENTIA +. S. IOHA– Same NNES.B.



+ FLOR– ENTIA Same

240

+. S· IOHA– NNES·B Same

Pellet between legs of saint

Ring with pellet

Bernocchi 1975: 12, No. 115.

Bernocchi 1975: 13, No. 119.

Coin Finds from the Akko Marina

*7 *8

*9

*10

*11

*12

72837 72825

72826

72828

72840

72823

3.51 3.47

3.47

3.48

3.47

3.54

20 20

20

20

20

21



Same

Same

Same



+ FLOR– ENTIA Same

+. S. IOHA– NNES.B Same

Bernocchi 1975: Six-pointed 13, No. 122. star



+ FLOR– ENTIA Same

+ . S. IOHA– NNES·B Same

Bernocchi 1975: 14, No. 134.

Same

+. ·S· IOHA– NNES·B Same Same

Bernocchi 1975: 14, No. 136.

+. S. IOHA– NNES.B Same

Cf. Bernocchi 1975: 15, No. 149. New variety with elevated pellet before B.





+ FLOR– ENTIA Same



+ FLOR– ENTIA Same

Crescent with pellet

Five-petal flower

+. S. IOHA– NNES.B Same

Same

Bernocchi 1975:18, No. 189. Key

+. S. IOHA– NNES.B Same

*13

72822

3.53

20



+ FLOR– ENTIA Same

*14

72834

3.50

20



Same

Same



+ FLOR– ENTIA Same

+. S. IOHA– NNES.B Same



+ FLOR– ENTIA Same

+. S. IOHA– NNES.B Same

*15

*16

72819

72816

3.54

3.55

21

20

Bernocchi 1975: Voided cross 19, No. 204. with four pellets Same

Four-petal flower with stem

Same

Bernocchi 1975: 21, No. 228.

Bernocchi 1975: 21, No. 233. Grape leaf

*17

72829

3.50

20



Same

Same

Same

Same

*18

72833

3.51

20



Same

Same

Same

Same



+ FLOR– ENTIA Same

+. S. IOHA– NNES.B Same

*19

72820

3.54

20

241

Two oak leaves with stem

Bernocchi 1975: 21, No. 235.

Robert Kool

*20

*21

72818

72843

3.53

3.52

21

20





+ FLOR– ENTIA Same

+. ·S·IOHA– NNES.B Same

+ FLOR– ENT[IA] Same

+[. S. IO]HA– NNES.B Same

Bernocchi 1975: 24, No. 266. Hammer Bernocchi 1975: 24, No. 277. Shell

*22 *23

*24

*25

*26

72824 72842

72841

72845

72827

3.52 3.48

3.50

3.48

3.51

20 20

20

20

20

+ FLOR– ENTIA Same

+ . S. IOHA– NN ES.B Same



Same

Same



+ FLOR– ENTIA Same

+ . S. IOHA– NNES.B Same



+ FLOR– ENTIA Same

+S IOHA– NNES B Same



+ FLOR– ENTIA Same

+ . S. IOHA– NNES.B Same



Bernocchi 1975: 26, No. 301. Pine cone Same

Sheaf of wheat

Same

Cf. Bernocchi 1975: 28, No. 326.

Cf. Bernocchi 1975: 29, No. 340. Unknown variety Trefoil with without stops stem Bernocchi 1975: 29, No. 346. Hook

Florence, 1252-1291 – Fiorino Series XX with symbols

*27

72821

3.52

20

*28

72844

3.48

20

*29

72835

3.50

20

+ FLOR– ENTIA Same

+ . S. IOHA– NNES.B Same



Same

Same



+ FLOR– ENTIA Same

+ . S.IOHA– NNES.B Same



Bernocchi 1975: 47, No. 543.

Segno 7 Same

Same

Unpublished. Poppy (?)

*30

72839

3.48

20

ã

Same

Unpublished.

Same Poppy (?)

242

Chapter 9

A Fifteenth-Century Harbor Installation with Wooden Pillars in Akko Ehud Galili (1) and Baruch Rosen (2) 1. Israel Antiquities Authority and Zinman Institute of Archaeology, University of Haifa 2. Israel Antiquities Authority Key words: marine archaeology, Acre, wooden pier, wooden quay, Mamluk, Ottoman

Fig. 1: A) Plan of the Akko Harbor in 1977, before the construction of the modern marina; location of the wooden pillar installation (marked by a rectangular frame) (modified after Fried, Glaswein, Werner Engineers 1957), B) Plan of the distribution of the installation’s wooden pillars, C) Possible reconstruction of a pier on pillars attached to the breakwater, D) Possible reconstruction of a pier on pillars detached from the breakwater. (S. Ben-Yehuda, E. Galili)

Introduction During the Akko Marina deepening project (Galili this volume a) a harbor installation made of wooden pillars was discovered southeast of the entrance of the Akko Marina (Fig. 1A). The installation was at a water depth of 3–4m, 20–30m from the modern breakwater tip. It included a group of 11 wooden pillars, embedded upright in the seabed. The rectangular (15 × 20cm) pillars protruded, at the discovery time, to a height of 15–40cm above the seabed. Their upper parts had eroded due to repeated deepening of the harbor and marine erosion. Another

pillar, torn out during deepening of the harbor, was discovered not in situ. That pillar (180 × 15 × 13cm), also rectangular, is pointed on the end, apparently to enable an easy insertion into the seabed. The pointed portion is approximately 50cm long and is rectangular in cross-section (Fig. 2). The pillars were inserted into a triangularly shape area measuring 28 × 28 × 14m (Fig. 1B). On the southwestern part of the triangle is a group of six pillars, four of which (Pillars 5–8) are arranged in a row on a north–south axis and two more (Pillars 9, 10) create a kind of arc emerging from the northern part of the row and facing east. The pillars are

243

Ehud Galili and Baruch Rosen

Discussion and Conclusion The installation was apparently used as a quay or pier constructed on pillars, used to moor ships near the harbor entrance, in relatively deep water. That area was the closest that ships with a 2–3m deep draft could get to the inner quays of the harbor, which was silted up at this time. The installation may have been separated from the remains of the southern breakwater (Fig. 1D). Alternatively, it could have been connected to the ruins of the breakwater (Fig. 1C). It seems that the installation served as a tie-up for ships loading and unloading cargo into lighters or directly onto the southern breakwater. Installations consisting of pillars of a similar size were discovered in Ashkelon and in the northern bay of Atlit. The Atlit structure was dated to the twelfth–thirteenth centuries CE (Galili and Sharvit 1997; Galili et al. 2002). In 1291 Akko was destroyed by the Mamluks and the harbor was blocked to prevent the return of the Christians to the city. The need for safe haven for ships in the northern part of Israel and the special natural characteristics of the Akko Bay led to the continuation of the harbor’s activity in the fourteenth century. Warehouses were still to be found in the ruined city, belonging to Venetian merchants using the harbor to export cotton to Europe (Ashtor 1983: 24; Shor 1990: 161–163). The Mamluks were interested in the export of cotton and pilgrim traffic, since both yielded taxes, thus they could have built mooring facilities in the area of the destructed harbor. The wooden-pillar anchorage whose remains are discussed may have been built in the fifteenth century for the cotton trade and for loading and unloading of cargo and passengers by means of small and medium-size vessels as an alternative to the harbor, which the Mamkuks had destroyed. The construction of maritime installations of wood is atypical to the local material culture, where large trees were an expensive and rare commodity. Maritime structures, including evidence of wood, brought from Europe and used in the construction of breakwaters, have been uncovered in the Roman harbor of Caesarea and the Byzantine harbor of Tiberias. Evidence of wood, used in nautical and coastal installations, has been found at other coastal locations as well (Rosen et al. 2004). Evidence of a wooden dam constructed during the Roman period has recently been discovered at Naḥal Taninim (Sa‘id 2002). Possibly this construction style and material were imported from Europe by the Romans and the Crusaders. The finds from Akko Marina show that later, during the Mamluk and Ottoman rule, the use of wood in marine installations continued. That may indicate a persistence of northern Mediterranean influence in maritime construction, possibly even a European one.

Fig. 2: One of the pointed wooden pillars which was pulled from the bottom by the dredger in the area of the wooden pillar installation (S. Ben-Yehuda , E. Galili) 60–140cm apart. About 10m north of the end of the row of pillars is another pillar (the northernmost one in the installation) and 8m northeast of the group are two more pillars that are 180cm apart (Nos. 2, 3). South–southeast of these two are two lone pillars (Nos. 1, 11). There were probably additional pillars in this installation that were torn out over the years by fishing nets or anchors caught in them, and, by the repeated dredging of the entry channel to the harbor. The pillars have been dated by C14 to the mid-fifteenth century CE (475 ± 50 BP un-calibrated; 1425–1461 CE after calibration, RT 1755). The wood was identified as European oak (Quercus robos) (N. Lipschitz, pers. comm.).

244

A Fifteenth-Century Harbor Installation with Wooden Pillars in Akko

Acknowledgments The authors wish to thank the Israel Antiquities Authority for the institutional support, M. Feinberg Vamosh for the English editing, D. Jacoby for his useful comments on the manuscript, N. Lipschitz for wood identification and S. Ben Yehuda for the drawings.

References Ashtor E. 1983. Levant trade in the Later Middle

Ages Princeton.

Fried, Glaswein, Werner Engineers 1957.

Hydrographical Map of the Akko Port, Report 191–2 for the Akko Municipality, Haifa. (Hebrew). Galili E. and Sharvit J. 1997. Underwater Survey in the Mediterranean Sea 1992–1996. Hadashot Arkheologiyot 107: 138–144. (Hebrew). Galili E. Raban, A. and Sharvit J. 2002. Forty Years of Marine Archaeology in Israel. In H. Tzalas, (ed). Tropis VII, Proceedings of 7th International Symposium on Ship Construction in Antiquity, Pylos, Greece, 1999. Pp. 927–961. Athens. Galili E. This volume a. The Akko Marina Archaeological Project, Introduction. Rosen B., Zvitov R., Galili E. and Nussinovitch A. 2004. Copper Preservation of Pinewood Submerged for 550 Years. Michmanim 18: 9–18, 79–83. Sa‘id A. 2003. The Nahal Taninim Dam, Hadashot Arkheologiyot 114: 38–41. (Hebrew). Shor N., 1990. The History of Akko, Pp. 161–163. Tel Aviv. (Hebrew).

245

Chapter 10

Anchors from the Akko Marina Ehud Galili (1) and Baruch Rosen (2) 1. Israel Antiquities Authority and Zinman Institute of Archaeology, University of Haifa 2. Israel Antiquities Authority Key words: marine archaeology, Acre, stone anchor, lead anchor, iron anchor, Roman, Byzantine, Ottoman

Introduction

stones circumscribed by grooves, for securing a rope, were used as weight anchors to secure water craft and fishing equipment beginning in prehistoric times, and are still used for that purpose. Weight anchors operate on the principle of holding in place the watercraft to which they are tied by causing friction between the tied stone and the seabed. Underwater archaeological studies of shipwrecks, and anchors discovered at shrines and cultic sites in a dateable archaeological context, allow the anchors to be dated and classified by type (Frost 1973). The most prominent characteristics of ancient stone anchors (as opposed to modern stone anchors) are the size of the anchor and the diameter of the hole to which the cable was tied. Traditional fishermen nowadays use stone anchors in anchoring small boats. These anchors are tied by thin, strong ropes and consequently, the hole diameter of a modern stone anchor is small compared to that of a similarsized ancient anchor. Thus, larger, heavier anchors

Anchors are the commonest archaeological find on the seabed. Generally water craft carry more than one anchor. In antiquity, ships sailing the high seas carried as many as 15 or more anchors (Galili 1985). Usually one or two anchors were used for routine work while others were stored below deck and held in reserve. Anchors made of stone and iron were retrieved during the Akko Marina archaeological project (Galili this volume a). They reflect extensive nautical activity in the harbor over the years. Two iron anchors with four particularly large arms (grapnels) were found ca. 1km south of the harbor.

Stone Anchors Stone anchors may be roughly divided into weight anchors (stone perforated by a single hole, Fig. 1a) and composite anchors (stone slabs perforated by two or more holes, one for tying the rope and the others for inserting wooden pegs-Fig. 1b). Perforated stones or

Fig. 1: Schematic drawing of stone anchors: A) one holed weight anchor, B) three holed composite anchors (S. Ben-Yehuda).

Fig. 2: Weight anchor (No. 001/3), scale = 5cm (E. Galili). 246

Anchors from the Akko Marina

diameter, 50cm thick, with a hole in the center, 20cm in diameter (No. 001/5). Its estimated weight is ca. 350kg. The shape of this object resembles a millstone, although it may have been used as a weight anchor or a mooring stone. It was left on the harbor bed. Composite Stone Anchors Three of these flat, rectangular stone slabs were retrieved. Each has two holes and weighs 4–5kg. Anchors of this size were used for small boats (3–5m long) or to anchor fishing gear. They are described in Chapter 12, under fishing gear. (see Galili and Rosen this volume c, fig. 5)

Iron Anchors Iron anchors started appearing during the second century BCE and took the same form as the composite wooden anchors that were in use since the beginning of the seventh century BCE. Composite iron anchors were made of two parts that could be assembled or taken apart for storage. The anchor,

Fig. 3: Large weight anchor (No. 001/4) (E. Galili). (50–300kg) with a large hole (7–15cm) are usually ancient. From the Late Bronze Ageon, weight anchors were upgraded by inserting wooden pegs into holes at their bottom. These “composite” anchors increased holding power and allowed a reduction in the weight/ mass, of the stone anchor. Small Weight Anchor A square kurkar slab with an 18cm hole (No. 001/3, weight 39kg, thickness 18cm (Fig. 2) was discovered on the seabed in the northeastern part of the eastern rampart. Large Weight Anchor This rectangular weight anchor was made of kurkar stone (No. 001/4, length 132cm, width 62cm, thickness 27cm, weight 280kg). It has one hole, 23cm in diameter (Fig. 3). It was discovered in the southwestern part of the harbor (see location in Galili this volume a, fig. 3: 22), during previous dredging. Stone anchors of this type were used mainly in the Late Bronze Age and many like them were discovered along the Israeli coast in shallow water (3–5m), especially in exposed areas (Raban and Galili 1985). The size of this anchor is unusual; it may have served as a mooring stone in the harbor. Round Weight Anchor A stone object was discovered in the northeastern part of Area D (see location in Galili this volume a, fig. 3), consisting of a round kurkar slab, 100cm in

Fig. 4: Two-armed, iron, Roman anchor from Ashkelon (S. Ben-Yehuda, E. Galili).

247

Ehud Galili and Baruch Rosen

1971: 51–61, Pl. X: 6). The latter is displayed in the National Maritime Museum in Haifa (Inv. No. 613.6). A group of 13 anchors of this type were found at Haifa Bay ca. 3km south of Akko at a depth of ca. 11 m (A. Kotzer, pers. comm).

forged as a single unit, consisted of a shank and one or two arms at the bottom. At the shank top were two holes, one to insert the stock and the other to hold a ring to which the anchor cable was attached. The stock (rectangular or round in section, having a stair in its center) was inserted into the shank top, at 90° to the arms, thus assuring that one of the arms would penetrate the seabed while anchoring (Fig. 4). The stock was secured by an iron pin or by a rope that was tied to a ring at it’s end. In a number of cases, groups of iron anchors were found along the coast of Israel together with remnants of shipwrecks, which made it possible to date the anchors (Galili et al. 2014).

a

Middle Byzantine, Y-shaped Iron Anchor with Two Arms This anchor (No. 002/1, length 147cm, reconstructed width ca. 110cm, weight of surviving portion 53.5kg), consisted of a round shank, at the top of which was a circular hole into which the vanished stock was inserted (hole diam. 6cm). Another hole at the shank top (diam. 3.5cm) was intended for a ring or a rope, which did not survive (Fig. 5 a, b). At the shank’s bottom were two broad arms, rectangular in cross-section. One of the arms was preserved almost intact, but had been broken from the anchor. This anchor may be of Type E according to Kapitan’s classification (1984: fig. 8). The arm may have been detached when the ship using it, was wrecked in a storm while in the harbor. Similar anchors were found in shipwreck remains in the Marmara Sea (Günsenin 1999; Pulak 2007) and in a ship that carried a cargo of glass items, wrecked off the coast of southern Turkey (Bass et al., 1984; Bass 1979). They were dated to the middle Byzantine period, the eleventh century CE. Several anchors of that type were discovered along Israel’s coast, one in Ashkelon (Galili et al. 2000) and the other off the Carmel coast (Kapitan 1969-

Fig. 6: A) Fragment of cross-shaped iron anchor from the Akko Marina (no. 001/2, E. Galili)

b

Fig. 5: A) Y-shaped, two-armed iron anchor from the Akko Marina (No. 002/1); B) Proposed reconstruction (S. Ben-Yehuda, E. Galili).

Fig. 6: B) Proposed reconstruction of the anchor (S. Ben-Yehuda, E. Galili). 248

Anchors from the Akko Marina

Byzantine Iron Anchor with Two Arms in the Form of a Cross

A

Also recovered was a fragment of an anchor (No. 001/2, length?, reconstructed width 100cm, thickness 4.5cm, weight of surviving portion 9.8kg). One arm survived along with a part of the shaft (Figs. 6A, 6B). The 90o angle between the shank and the arm and its general form date this anchor to the Byzantine period (Kapitan 1984: fig. 8/D). That anchor type was common from the fourth to the seventh centuries CE. Dozens of similar anchors were found along the Israeli coast, mainly on the northern Carmel coast (between Atlit and Haifa), in the southern anchorage at Dor and in the anchorages of Apollonia, in Caesarea and Ashkelon.

B

Iron Anchors with Four Arms (Grapnels), found south of the Akko Marina Following a report by the fisherman A. Kotzer, a particularly large iron anchor. (Figs. 7a, 8A, 8B), was discovered ca. 1,800m south of the Akko Harbor, about 800 m from shore, on the rocky seabed at a depth of 12m (N 32o 54’ 12”; E 35o 04’ 08”). The anchor, which has four arms, was 480cm long, and had a maximum width of 200cm, with a ring (diam. ca. 50cm) on the upper part. At the ends of the arms were four triangles (the bills) embedding the anchor into the seabed; they measured 50 × 50cm. The wide section of the anchor and the arms were found facing in a generally western direction (at an angle of 300o) and the ring to which the cable would have been attached was facing east. The underwater

Fig. 8: A,B) Diver examining iron grapnel a (E. Galili). survey in this area in 2004 revealed another anchor 150m southeast of the first one, similar in size and shape (length 500cm, width 200cm); however, two arms were broken (Fig. 7b). The latter anchor lies in a general north-south direction (at an angle of 190o) with the arms to the south and the ring to the north. The arms seem to have been broken due to stress on the anchor in a stormy sea. Presumably both anchors were lost by ships during stormy weather. Such anchors were in use during the Crusader period and are usually dated to the tenth–fourteenth centuries (Jacoby 1985). Small anchors of this type (up to 1.5m long) are used on ships to this day and are known as grapnel anchors.

Discussion and Conclusions

Fig. 7: Two iron anchors with four arms discovered south of Akko the Marina (grapnels) (E. Galili).

The large stone anchor (001/4) was apparently used during an early period of the natural anchorage, prior to the construction of the southern breakwater. Although many of the finds discovered in the harbor are dated to the Hellenistic and Roman periods, no lead parts of wooden or iron anchors from these periods were 249

Ehud Galili and Baruch Rosen

Acknowledgments

found in the harbor. It seems that in the Hellenistic and Roman periods the breakwater and the piers in the Akko Harbor functioned in a way that ensured safe mooring. Thus the use of anchors in the harbor in these periods was limited. There is a text written before the Arab conquest, which refers to a shipyard in Acre, implying harbor activity, and shortly afterward the harbor was used by the Arabs (Raban 1993). The iron anchors found in the harbor belonged to ships that anchored there during the Byzantine period. It is possible that during the Byzantine period the harbor was not properly maintained, the breakwaters were in partial ruin and anchoring conditions were poor. The iron anchors that were found in the harbor were apparently broken during attempts to heave them up after they had become embedded in the seabed during a storm, or during a storm that severely rocked the ship and thus broke the anchor. It is also possible that these anchors originate from ships wrecked in the harbor. During the Crusader period the Akko Harbor was the main maritime gateway to the Holy Land and saw extensive military, civilian and commercial activity. That being the case, the small number of finds from the Crusader period in the marina is somewhat surprising. Historical document of that period describes the western basin as a shallow area unsuitable for anchoring large ships. It noted that the large ships anchored in the eastern basin (Jacoby 1985). That statement explains the scarcity of Crusaders remains in the western basin. However, the eastern basin was open to southwest winds, making it dangerous to anchor there during winter storms. Indirect evidence that anchoring conditions in the harbor were poor is provided by another historical document that mentions dozens of iron anchors sent from Venice to Akko by the Venetians to be leased to ship owners (Jacoby 1985; Galili and Rosen 2008). It has been suggested that the lack of suitable conditions for anchoring during severe winter storms forced ships to anchor in the open sea. That created a need for a large number of strong, reliable iron anchors that could hold to the seabed and prevent the ships from being swept ashore and wrecked. The pair of large Crusader iron anchors found about 1,800m south of the Akko Marina illustrates the problem of lack of safe haven for large vessels in the harbor during winter storms in that period. The position in which the anchors were found could indicate that they belonged to large ship(s) anchoring outside the harbor during a storm. It seems that two of the arms of the southern anchor broke during the storm. That caused the anchor to lose its hold in the seabed. Consequently the ship was swept eastward to shore where it was grounded or wrecked. In nautical charts from various periods, an anchor symbol appears in an area where these iron anchors were discovered (Galili et al. this volume a). The symbol indicated a preferred spot to drop anchor in the open sea and fits the historical descriptions and the archaeological finds.

We wish to thank the Israel Antiquities Authority for the institutional support, M. Feinberg Vamosh for the English editing, D. Jacoby for his useful remarks on the manuscript and Howard Rosenstein and Fantasea for the underwater photography equipment.

References Bass G.F. 1979. The Shipwreck at Serçe Limani, Turkey. Archaeology 32: 36–43. Bass G.F., Steffy J.R. and van Doorninck F.H. Jr. 1984. Excavation of an 11th-Century Shipwreck at Serçe Limani, Turkey. National Geographic Society Research Reports 17: 161–182. Frost H. 1973. Anchors, the potsherds of marine archaeology: on the recording of pierced stones from the Mediterranean. In D. Blackman (ed.). Marine Archaeology. Pp. 397–409. London. Galili E. 1985. A Group of Stone Anchors from Newe-Yam. International Journal of Nautical Archaeology 14.2: 43–153. Galili E., Sharvit J. and Dahari U. 2000. Ashkelon, Underwater Survey, Hadashot ArkheologiyotExcavations and Surveys in Israel 110–111: 82–83. Galili E. and Rosen B. 2008. Acco – The Harbor. In E. Stern (ed.). The New Encyclopedia of Archaeological Excavations in the Holy Land 5. Pp. 1558–1561. Jerusalem. Galili E., Bauvais S. and Rosen B. 2014. Cargoes of Medieval Iron Ingots Recovered from Shipwrecks off the Carmel Coast, Israel. Archaeometry: doi: 10.1111/arcm.12077. Galili E. and Rosen B. This volume c. Fishing Gear. Galili E. This volume a. The Akko Marina Archaeological Project, Introduction. Galili E., Zviely D. and Rosen B. This volume a. The Akko Harbor in Numismatic, Epigraphic, Cartographic and Aerial Photographic Evidence. Günsenin N. 1999. From Ganos to Serçe Limani: Social and Economic Activities in the Propontis during Medieval Times Illuminated by Recent Archaeological and Historical Discoveries. The Institute of Nautical Archaeology Quarterly 26.3: 18–23. Jacoby D. 1985. Venetian Anchors for Crusader Acre. The Mariner’s Mirror 71 (1): 5–12.

250

Anchors from the Akko Marina

Kapitan G.F. 1969–1971. Ancient Anchors and Lead Plummets. Sefunim 3: 51–61. Kapitan G.F. 1984. Ancient Anchors, Technology and Classification. International Journal of Nautical Archaeology 13: 40. Pulak C. 2007. The Wrecks of Yenikapi: the Gift of Storm. Arkeoatlas 6: 129–141. Raban A. 1993. Maritime Acco. In E. Stern (ed.). The

New Encyclopedia of Archaeological Excavations in the Holy Land 1. Pp. 29–31. Jerusalem.

Raban A. and Galili E. 1985. Recent Maritime Archaeological Research in Israel – A Preliminary Report. International Journal of Nautical Archaeology 14.4: 321–356.

251

Chapter 11

Ship Fittings and Rigging Devices from the Akko Marina Ehud Galili (1) and Jacob Sharvit (2) 1. Israel Antiquities Authority and Zinman Institute of Archaeology, University of Haifa 2. Israel Antiquities Authority Key words: marine archaeology, Acre, ship nails, lead sheathing, wooden block, lead patch

Fig. 1: 1-8. Various copper-alloy (1-7) and iron (8) nails (IAA)

Introduction

Metal Nails

While deepening the Akko Marina (Galili this volume a), metal and wooden objects associated with ship hulls or riggings were recovered. They included: metal nails, sections of lead sheathing, lead patches and wooden block sheaves. The finds, which may shed some light on the marine activity in the ancient harbor of Akko, are described and discussed below.

Seven used nails made of copper alloy and one iron nail were recovered, most were well preserved and only slightly affected by corrosion. No. 1 (31/93-C-35/487; Fig. 1: 1) A complete nail (9 × 9mm, 190mm total length) with square shaft, slightly bent; the head is round and 252

Ship Fittings and Rigging Devices from the Akko Marina

Lead Sheathing and Patches

conical (27mm in diameter; 10mm high), the top is hammered.

The term “lead sheathing” refers to a thin plate of lead affixed to the outer face of wooden-hulled ships. The finds includes nine items: five torn sections of sheathing; two rectangular patches, one round patch and one long patching strip (Table 1).

No. 2 (31/93- C-35/487; Fig. 1: 2) An incomplete copper alloy nail (7 × 7mm, 85mm total length) of which a tip is missing. It has a square shaft, the head is round and conical (20mm in diameter, 5mm high) and the top is hammered.

No. 1 (Fig. 2) A piece of lead sheathing, one round corner was preserved on it while the remaining edges were irregularly torn. Approximately 17–18 nail holes, some square and some irregular, are distributed, in an approximate grid pattern, on the piece of lead.

No. 3 (31/93- F-31/829; Fig. 1: 3) Same as No. 2, incomplete cooper alloy nail with square shaft (7 × 7mm, 65mm total length). The head is 16 mm in diameter and 3mm high. No. 4 (31/93- A-36-38/193; Fig. 1: 4) Same as No. 2, incomplete copper alloy nail with square shaft (6 × 6mm, 40mm total length). The head is 14mm in diameter and 5mm high. No. 5 (31/93- A-36-/384; Fig. 1: 5) A complete copper alloy nail (6 × 6mm, 40mm total length), the head is 14mm in diameter and 5mm high and has a hammered, flat conical section. No. 6 (31/93- C-60-/13; Fig. 1: 6)

Fig. 2: A piece of lead sheathing (IAA)

The artifact is a complete copper alloy spike, having a maximum length (shaft and head) of 110mm. The head is square when viewed from above (14 × 14mm; thickness 4mm), and has a hammered flat section. The shaft is rectangular (8 × 8mm) tapering to a sharp chisel tip.

No. 2 (Fig. 3) A torn piece of lead sheathing, having three copper tacks inserted in it. Two nail holes, one square and one round, are also seen.

No. 7 (31/93- C-1-/843; Fig. 1: 7) An incomplete copper alloy spike or nail with remnants of wood attached to it. It has a maximum length (shaft and head) of 125mm. The head is square (17 × 17mm; 5mm thick), and has a hammered flat section. The shaft is square (11 × 11mm) and is slightly bent. No. 8 (31/93- F-31-/1580; Fig. 1: 8) A complete square (10x10mm) iron nail having a maximum length (shaft and head) of 150mm. The head is round and flat when viewed from above (27mm in diameter).

Fig. 3: A torn piece of lead sheathing (IAA) 253

Ehud Galili and Jacob Sharvit

No. 3 (Fig. 4) A long strip of lead sheathing or patch punctured by 28 round or square nail perforations along its edges. Fourteen perforations are positioned in a straight line along each edge of the patch, with parallel rows facing each other. In some places the negative impression of a nail head is still preserved on the outer (sea-facing) side of the sheathing.

Fig. 4: A long strip of lead sheathing or a patch (IAA)

Fig. 6: A torn lead patch

No. 4 (Fig. 5) A piece of torn lead sheathing with a row of four nail perforations along its intact, straight edge and two holes in the middle of the piece. Three perforations are square, and the others have an irregular shape. Observation of the negative impressions left by the tack heads reveals impressions of the tiny bulges formed on the underside of the heads.

Fig. 7: A broken lead patch (IAA)

Fig. 5: A piece of torn lead sheathing (IAA) No. 5 (Fig. 6)

No. 7 (Fig. 8)

A torn lead patch, having three parallel rows of nail made perforations. Altogether there are eight nail perforations. The perforations are square and of the same dimensions. On the negative impressions left by the tack heads impressions were observed of the tiny bulges formed on the underside of the heads.

A complete rectangular lead patch excellently preserved with ten nail holes on the edges and two holes inside the rectangle. Six small holes are square, two are round and four have an irregular shape. No. 8 (Fig. 9)

No. 6 (Fig. 7)

A complete round lead patch with four equally spaced nail perforations, 3–5mm in diameter; next to one perforation is an additional, fifth perforation. The surface of the patch is decorated with a circumscribed line, equidistant from the rim, crisscrossing each other forming a net-like decoration.

A broken, rectangular lead patch, torn on one corner, with six holes, some round and some square, along the edges.

254

Ship Fittings and Rigging Devices from the Akko Marina

No. 9 (Fig. 10) A broken piece of square lead sheathing. No. 10 (Fig. 11) A section of a lead sheathing, in the form of an elongated strip (3.5m long 0.3m wide), folded accordion-like (Fig. 11). It was recovered on a known

Fig. 8: A complete rectangular lead patch Fig. 10: A piece of square lead sheathing

Fig. 9: A complete round lead patch (IAA)

Fig. 11: Remains of sheathing of a Roman ship scraped from a ship’s keel 255

Ehud Galili and Jacob Sharvit

navigation hazard, a rocky shoal located some 300m southwest of the Akko Marina (Rosen et al. 2012: 173). Copper alloy nails, used to attach the lead strip to the ship keel, are still attached to the strip in parallel rows (see details in Galili this volume f).

Parts of Wooden Blocks Blocks are used for transmitting power or changing the angle of exerted power by means of a rope passing around the movable sheaves (pulleys). Two whole sheaves and one broken were recovered. No. 1 (Fig. 13) A complete, round sheave made of hard wood in an excellent state of preservation. Both sides of the sheave (98mm in diameter) are flat (26mm thick), and have a central, round, straight axial hollow (27mm in diameter). The circumference of the sheave bears a wide groove (3–4mm deep) to guide the rope. On both sides of the sheave are numerous circular scratches caused by the rotating motion and friction between sheave, block-shell and sand or another abrasive.

Fig. 15: A modern block (IAA)

Discussion and Conclusions Nails Shipwrights use nails to fasten boards or timbers to one another, or to fasten fittings to the wooden hull. Nails can be square or round in section and are usually made of iron, copper, or copper alloys, such as bronze. They can be generally classified into three main categories: tacks, nails and spikes, each of which can be further subdivided. Terms such as “diamond-head,” “rose-head,” “square shank” and “chisel point” are often used in shipbuilding literature (de Kerchove 1961: 763; Smith 1974: 96). All the nails recovered from the Akko Marina are worn and seem to have been used. They could have originated in ships wrecked inside the harbor, or they may have fallen from ships being repaired. Most probably they were not thrown overboard as waste, since nails had great value as reusable metal. The size of the nails, though small and unrepresentative could indicate that only small and medium-size ships docked inside the harbor.

No. 2 (Fig. 12) A complete, round sheave made of hard wood, in a good state of preservation. The sheave (140mm in diameter) is flat on both sides (32mm thick) and has a central straight axial hollow (35mm in diameter). On one side of the sheave is a 5mm deep recess in the shape of an equilateral triangle with rounded apexes. A flat, compatible iron fitting is fixed inside the groove and attached with three nails or screws. The iron axial pin is still preserved in the hole. No. 3 (Fig. 14) A broken wooden sheave, similar to No. 2.

Lead Sheathing Lead sheathing was fastened from the keel upward to the first or second strakes above the water line. Woven textiles, soaked in an oily substance, were inserted between the wooden hull and the lead sheathing (Black 1999; Kahanov 1993: 46). The lead sheets were attached overlapping each other and joined with small copper tacks with large, flat, heads, enabling water to flow smoothly over them. The lead sheathing served as a physical barrier against

Figs. 12, 13: Complete round sheaves Fig. 14: A broken wooden sheave (IAA) 256

Ship Fittings and Rigging Devices from the Akko Marina

marine fouling, improved sealing and enhanced the ship's strength and stability. Lead sheathing on ships appears from the sixth century BCE through to the fourth century CE (Kahanov 1993: 242). The use of lead sheathing reached a peak at the end of the fourth century BCE and the beginning of the third century BCE, and from the mid-second century BCE to the mid-first century CE. No lead sheathings were observed on ships dating earlier than the fourth century BCE or later than the second century CE. During the time that lead sheathing was in use, there were two picks, during the first half of the third century and during the first century BCE (Kahanov 1993: 242). The presence of the remnants of lead sheeting and the nails indicates that while some may have fallen from ship that anchored in the harbor, patches of such sheets suggest that maintenance took place onboard the anchoring ships or in the harbor.

Acknowledgments The authors wish to thank the Israel Antiquities Authority for the institutional support, M. Feinberg Vamosh for the English editing, D. Jacoby and B. Rosen for their useful remarks on the manuscript and T. Sagiv and C. Amit for the photographs.

References Black E. 1999. Fibers and textiles used in the construction of the ship’s hulls, In H. Tazalas (ed.). Tropis V, Proceedings of the 5th International Symposium on Ship Construction in Antiquity, Nauplia, Greece, 1993. Pp. 53–64. Athens. De Kerchove R. 1961. International Maritime

Dictionary. New York. Wooden Blocks

Galili E. This volume a. The Akko Marina

Blocks consist of one or several grooved sheaves mounted in a wooden shell fitted with a hook, eye or strap by which it is attached to the hull or to the mast (Fig. 15). They commonly derive their names from the number of sheaves, namely, single, double, triple etc., or from some special shape or construction. They also may be known according to their use or the position they occupy aboard. Often blocks are named after the ropes that are roved through them. They are usually made of elm or oak wood, as both are weather-resistant (but must be kept varnished or painted). The shells are hollow with solid grooved pulleys fixed within them. Sheaves are either of lignum vitae or metal. Lignum vitae is an exceedingly hard wood, dark in color and possessing self-lubricating properties. In the center of the pulley is a hole for a pin or axis, on which it turns. This is frequently reinforced by placing an additional piece of metal, called a coak or bush, inside the hole. Around the circumference of the sheave is a round groove for leading the rope. The sheave is positioned in the shell by a pin that passes through both shell and sheave. The blocks’ sheaves recovered from the Akko Marina are difficult to date by their shapes, since there is currently no known typology for sheaves. However, a wooden sheave similar to No. 1 was found, out of archaeological context, at Quseir al-Qadim (Sabkha trench SIIb, loc. 3) and was speculatively dated to the Roman period (Whitcomb and Johnson 1979: 203, pl.70: u, v). Another sheave block, dated to the Ottoman period, was found in the Sharm el-Sheikh shipwreck (Raban 1984: 14). For more definite dates, C14 testing must be performed. It seems that the sheaves were thrown overboard after they broke or they may have sunk together with anchoring ships that were wrecked in the harbor.

Archaeological Project – Introduction. Kahanov Y. 1993. The sheathing and caulking of ancient ships in the Mediterranean from the 6th century BCE to the 5th century CE. M.A. thesis, University of Haifa (Hebrew). Raban A. and Perry, I. 1994: Undersea Excavations in Sinai Peninsula 1969-1973, Report submitted to Israeli Antiquities Authority. (Hebrew). Rosen B., Galili E. and Zviely D. 2012. The Roman Lighthouse of Akko. International Journal of Nautical Archaeology 41.1: 171–178. Smith P.C.F. 1974. The Frigate Essex Papers. Building the Salem Frigate 1798–1799. Salem. Whitcomb D.S. 1979. Chapter 6: Small objects. In D.S. Whitcomb, and J.H. Johnson (eds.). Quseir AlQadim Preliminary Report. Pp. 201–205. Cairo.

257

Ehud Galili and Jacob Sharvit

Table 1: Details of the lead sheathing

Dimensions

Distance Between Nails

Nail Pattern Arrangement

2–2.5

Not consistent

Grid

120

1

50

Two diagonal rows

710

60-90

1–1.5

35–45

Two parallel lines

31/93-F-66, 67/328

155

75

1–1/5

40–73

Two parallel lines

5

31/93-F66, 67/329

160

90

2

48–50 along the edges, 65 in the middle

Three parallel lines

6

31/93-B-43/156

170

140

1

45–70

7

31/93-C-35/44

210

170

2

Not consistent

Raw around the edges Raw around the edges

8

31/93-F-75, 79/1894

75

75

4/5

30–40

Square, on the edge

9

31/93-F-31/523

120

70

2

No.

Area in the Marina and Item No.

1

Length (mm)

Width (mm)

Thickness (mm)

31/93-F

600

500

2

31/93-F-66, 67/317

180

3

31/93-F

4

258

Chapter 12

Fishing Gear from the Akko Marina Ehud Galili (1) and Baruch Rosen (2) 1. Israel Antiquities Authority and Zinman Institute of Archaeology, University of Haifa 2. Israel Antiquities Authority Key words: marine archaeology, Acre, net sinkers, stone sinkers, fishing nets, fishing hooks

Introduction Metal and stone fishing implements discovered in the framework of the Akko Marina Project (Galili this volume a) include two bronze hooks, a fragment of a stone ring for retrieving fishing equipment, nine stone net sinkers and a few dozen lead sinkers. The finds are described and discussed below. In discussing the sinkers, the typology of sinkers previously discovered along the Israeli coastline is used (Galili et al. 2002) (Figs. 1, 2).

Stone Sinkers These objects were used to hold down the bottom lines of fishing nets (Fig. 3) or to hold down a long line of

hooks on the seabed. Their weight varies from 1kg to 12kg. Four of the nine stone sinkers discovered were irregular and have one perforation (Fig. 4 Type S1.2.1) (Table 1). They had rounded corners but were otherwise without particular form. Three of the four were made of sandstone (Fig. 4: A, B, C) and one was of limestone (Fig. 4: D). Three of the nine stone sinkers consisted of sandstone slabs worked into rectangles with rounded corners and two perforations, either square or circular (Fig. 5 Type S1.2.3). Two of three were broken crosswise in the area of the perforation. The complete sinker had undergone biological erosion but retained its rectangular shape. Wood remains (peg remains) were preserved in one of the holes. These were dated by C14 to 240 ±50 years before present, uncalibrated (RT 1753). The wood was identified as Abies alba (common fir) (N. Liphschitz, pers. comm.).

Fig. 1: Typology of ancient lead sinkers used in fishing on the coast of Israel (modified after Galili et al. 2002) 259

Ehud Galili and Baruch Rosen

Fig. 2: Typology of ancient stone sinkers used in fishing on the coast of Israel (modified after Galili et al. 2002)

Fig. 3: Schematic fishing net with stone and lead sinkers (modified after Galili et al. 2002)

A

Fig. 5: Stone net sinkers with two perforations (Type S1.2.3) (IAA)

C

The dimensions of the unbroken stone sinkers were reconstructed assuming that they were symmetrical. Due to their light weight, they were probably not used as ship’s anchors, but rather to hold down fishing gear. It is reasonable to assume that a rope was drawn through one of the holes and a pointed wooden peg was inserted into the other; thus, by digging into the seabed, they increased the anchor’s holding power. The location of the break shows that the sinkers were broken during use due to stress in the area of the hole in which the peg was inserted. Two other sinkers made of coarse sandstone were formed into a rectangular shape with an angled perforation: One was L-shaped (Fig. 6 right, Type S1.1.2) and the other was U-shaped (Fig. 6 left, Type S1.1.3).

B

D

Fig. 4: Stone net sinkers with a single perforation (Type S1.2.1) (IAA) 260

Fishing Gear from the Akko Marina

B

C

A

Fig. 7 A,B,C: Cast lead sinkers (Type L1.3.1, L1.2, L1.3.2): A. Ring sinker of a gill net; B. Tubular sinker of a coastal seine; C. Sinker for a fishing line (IAA)

Fig. 6: Stone net sinkers: Right – with L-shaped perforation (Type S1.1.2); left – with curved Ushaped perforation (Type S1.1.3) (IAA)

Lead Sinkers for Fishing Nets The assemblage contained 34 lead sinkers for attachment to the bottom line (lead line) of a fishing net (Fig. 3:1) and five cast lead objects of various forms without tying arrangements. Sinkers whose final shape was formed by casting: Four cast rings, plano-convex in cross-section were found (Fig. 7: A, D, Type L1.3.1), two of which were broken (weight: 39–42gr; outer diam. 34– 37mm); a cast sinker resembling a hollow cylinder (Figs. 7B, 8B, Type L1.2) weighting 22gr; outer diam. 22mm); and five cast lead objects, cone-shaped or pyramidal (Fig. 7: C). On one of the pyramidal sinkers from this group several intersecting lines were incised (Fig. 7: C), whose significance is uncertain.

Fig. 7D: Reconstruction of a lead line with cast-ring sinkers (E. Galili)

Sinkers whose final shape was formed by folding: The other sinkers were made of thin lead sheets folded into their final form (Fig. 1, Type L2). Five of these were folded into a tube (Fig. 8, Type L2.2) intended to hold strings of various thicknesses (weight: 11– 36gr; outer diam. 19–33mm; height 19–37mm). The rest of the folded sinkers (24) belong to the folded rectangular sinker type (Fig. 1, Type L2.3). Some of these sinkers bore decorations of various types (Fig. 9 Types A1, A2, A5, B1), and one is decorated with unclear lines converging in different angles (Fig. 10: A1). Among these sinkers two were folded double inward. Twenty-two were folded lengthwise in the center; they are divided into three groups by their size as follows:

Fig. 8A: Sinkers of a coastal seine, made of a folded, bent plate for a tube (Type L2.2) (IAA)

Long folded rectangular sinkers: (weight 76– 125gr; length 66–73mm; width 11–19mm; internal diam. of hole 3–5mm). Nine sinkers were made of a folded rectangular, cast sheet and one was made of thin sheet cut into a rectangle and folded.

Fig. 8B: Reconstruction of a coastal seine lead line (E. Galili) 261

Ehud Galili and Baruch Rosen

Wide, heavy, Short folded rectangular sinkers:

(weight 15–17gr; length 30–38mm; width, bent 14–16mm; diam. of hole 6–7mm). Three sinkers decorated with a frame and oblique lines (Fig. 10: A2); one is decorated with parallel lines across in longitudinal axis (Fig. 10: A3).

Narrow, light, folded rectangular sinkers: (weight,

8–13gr; length 42–48mm; width, folded 9–42mm; inner diam. of hole 1.0–1.5mm). Nine sinkers of this type were discovered. Some feature a frame design (Figs. 9: A1, 10: A4), two are decorated by crossed lines (Fig. 10: A3). One bears worn, oblique lines. Several sinkers of Type L2.3 have rounded corners; some are bent into a crescent shape on their horizontal axis. Based on the diameter of the hole (through which a line was passed), the form of these sinkers and their weight, it seems that the two sinkers folded double, and the nine narrow, lightweight sinkers were used to weigh down cast nets. The rest of the sinkers would have weighed down gill nets and coastal seines. Because fishing-sinker shapes changed little over the millennia, these sinkers cannot be dated by their shape (for detailed discussion see Galili et al. 2002).

Fig. 9: Decoration on lead sinkers of fishing nets recovered from the Akko Marina (modified after Galili et al. 2002) 4

3

Fishing Hooks Two intact hooks, made of copper alloy, apparently bronze, were found (Fig. 11). On the smaller of the two (length 28.3mm; width 16mm; weight 1.5gr) the area to which the fishing line was attached is oblate and broad (resembling a shovel) and the shank crosssection is round (diam. 4mm with corrosion). The hook was covered by corrosion, but was not deformed by use. The larger hook (58mm; width 25mm; weight 3gr; thickness of shank 2.5mm) had been extensively deformed by pulling, which changed the original form. The point at which the line was tied is slightly oblate (rectangular in cross-section) and bears a number of grooves on each side to prevent the line from slipping. In the area of the bow (Fig. 12), where

2

1

Fig. 10A: Rectangular lead sinkers of gill nets and cast nets, made of a folded rectangular sheet (Type L2.3) (IAA)

Fig. 10B: A fisherman using a cast net on the Sea of Galilee (after Oliphant 1976: 155)

Fig. 11: Bronze fishing hooks (IAA) 262

Fishing Gear from the Akko Marina

weighs 1,200gr, which constitutes between a quarter and a third of the weight of the original item, which was between 3kg and 4kg. Stone salvage rings such as these were discovered in Ashkelon (Galili and Sharvit 2000), Akko (Fig. 13: C) and Apollonia (Galili and Dahari 1991). Seemingly such rings were used to retrieve fishing lines caught on underwater obstacles (Fig. 13: D). Similar rings, made of lead, were discovered off Haifa’s southern beaches (Galili and Sharvit 1999). In the eastern Mediterranean and the present-day fishing ports of Akko, Haifa and Jaffa traditional coastal fishermen use such iron rings to retrieve fishing equipment from the seabed (Fig. 14). On the southern Levantine coast this tool is called a ḥalasa in Arabic (Galili and Rosen 2008; Galili et al. 2013).

Fig. 12: Diagram of fishing hook and terminology (S. Ben-Yehuda) bending efforts might be concentrated, the shank was broadened to a rectangular cross-section; indeed, at this point it was not straightened by pulling. The deformation begins in the area where the rectangular cross-section ends and the round cross-section begins. Based on a comparison to contemporary hooks, the small hook could have been used to catch fish weighing between 1.5kg and 4.0kg, and the larger one for fish weighing between 2kg and 7kg. (A. Kotzer, pers. comm.).

Fig. 13B: Reconstruction (S. Ben-Yehuda)

Fragment of Stone Salvage Ring This is a fragment of a limestone ring, which had been eroded and weathered (Fig. 13: A). The original inner diameter of the ring measured 28± 2cm (Fig. 13: B). The item is ovoid in cross-section (5 × 8cm). It

Fig. 13C: Similar stone ring found north of Akko

Fig. 13A: Fragment of stone ring used for retrieving fishing equipment

Fig. 13D: Reconstruction of the use of a salvage ring for releasing entangled fishing gear (E. Galili) 263

Ehud Galili and Baruch Rosen

fishermen used a variety of methods for exploiting the rich marine environment in the Akko region. Beach seines could have been used off the shallow, sandy beaches of Haifa Bay south of Akko and gill nets were probably used in the offshore rocky areas of Haifa Bay and the Galilee coast. Cast nets were used on the rocky coastal waters north of Akko and long lines were used in deep water offshore. These methods are still in use by traditional fishermen in Akko today.

Acknowledgments

Fig. 14: Salvage ring made of metal, part of the equipment for long-line fishing in the Akko Harbor (J. Galili)

We wish to thank the Israel Antiquities Authority for the institutional support, M. Feinberg Vamosh for the English editing, T. Sagiv, C. Amit and J. Galili for the photographs and S. Ben-Yehuda for the drawings.

Discussion The forms of artifacts used by traditional fishermen have changed very little over recent millennia. Thus, if they are not discovered in a clear archaeological context, they can rarely be dated with certainty based on their typology. The many and varied findings discovered in the Akko Marina show extensive fishing activity in and around it. The seabed surrounding Akko features numerous habitats: underwater kurkar ridges in the Haifa Bay and west of Akko, a rocky coast interspersed with abrasion platforms, headlands and fissures north of Akko and sandy beaches and sea bottom in Haifa Bay. These areas constituted a productive marine environment that allowed extensive exploitation of its resources. It is reasonable to assume that most of the fishing implements discovered in the marina came from fishing boats and ships that fished the waters within the harbor, or lost these items randomly while at anchor or during preparation and repair of fishing gear. Some of the findings may have reached the seabed after the sailing vessel that carried them was wrecked while at anchor. Sailors commonly fished to vary their diet while waiting in port or while anchoring at night. Thus, most sailing vessels had fishing equipment aboard (Galili et al. 2002; Galili et al. 2013). It is therefore to be expected that on the bed of a harbor that was active for thousands of years, a variety of fishing equipment would be found by the local fishing fleet or by visiting ships anchored and lose there. The fishing implements recovered from the Akko Marina suggest that fishing activity took place in Akko region in antiquity. Their diversity indicates that the

References Galili E. and Dahari U. 1991. Underwater Surveys along the Coast of Israel. Hadashot Arkheologiyot 97: 88–92. (Hebrew). Galili E. and Sharvit J. 1999. Haifa, Underwater Surveys. Hadashot Arkheologiyot 110: 19–24, 15*–20*. Galili E. and Sharvit J. 2000. Tel Ashkelon, Hadashot

Arkheologiyot 111: 111–114, 83*–85*.

Galili E., Rosen B. and Sharvit J. 2002. Fishing Gear Sinkers Recovered from an Underwater Wreckage Site off the Carmel Coast, Israel, International Journal of Nautical Archaeology 31.2: 182–201. Galili E. and Rosen B. 2008. Ancient Remotely Operating Instruments Recovered Underwater off the Israeli Coast. International Journal of Nautical Archaeology 37.2: 283–294. Galili E., Zemer A. and Rosen B. 2013. Ancient Fishing Gear and Associated artifacts from Underwater Explorations in Israel – a Comparative Study. Archaeofauna 22: 145–166. Galili E. This volume a. The Akko Marina Archaeological Project – Introduction. Oliphant L. 1976. Haifa: Or, Life in the Holy Land, 1882-1885. Jerusalem.

264

Fishing Gear from the Akko Marina

Table 1: properties of the stone sinkers

Diameter or measurements of hole

Length (cm)

Width (cm)

Thickness (cm)

Square Square

3×3 2.5 × ? (two holes)

27

22

9.2

Kurkar/ sandstone

Round Round

3.5 2 (two holes)

27

19

7.5

12

Kurkar/ sandstone

Round Round

3 3 (two holes)

38

27

9

S1.2.1

4.2

Kurkar/ sandstone

Round

1.5

30

27

10

31/93 000/5

S1.2.1

1

Limestone

Round

2

13

9

6.5

31/93 000/6

S1.2.2

8.3

Kurkar/ sandstone

Round

1.5

22

21.5

9

31/93 000/7

S1.2.1

6.2

Kurkar/ sandstone

Round

1

28

24

10

5.5 × 6 4×5

26

20

12

3×6 4×6

22

15

7

No.

Type (Fig. 2)

Weight (kg)

Stone Type

31/93 000/1

S1.2.3 (Broken)

6.9

Kurkar/ Sandstone

31/93 000/2

S1.2.3

3.8

31/93 000/3

S1.2.3

31/93 000/4

Shape of hole

32/90 65/2

S1.1.2

5.9

Sandstone

Square; one carinated hole with square cross-section

32/90 65/3

S1.1.3

4

Kurkar

Rectangular, asymmetrical

265

Chapter 13

Jewelry and Adornments from the Akko Marina Ehud Galili Israel Antiquities Authority and Zinman Institute of Archaeology, University of Haifa Key words: marine archaeology, Acre, stone pendant, bronze figurine, bronze cross

Introduction

The gender is unclear. This part seems to have been attached (soldered) externally to a receptacle of some kind (a bowl?) as a decoration. Bronze figurines with similar face patterns were recovered in Hazor (Negbi 1976: fig. 60-1454, fig. 68-1488) and were dated to the Iron Age I and the Late Bronze Age, retrospectively.

The materials retrieved during the Akko Marina project (Galili this volume a) included six metal and stone items used as decorations and adornments, among them a bronze bracelet, two pendants, one of bronze and one of stone, a gold clasp for a chain and two bronze items bearing the image of a human's head that had been soldered to other items.

Man’s Head with Tall Hat and Horns (002/2-31/93, Figs. 2, 4)

Description of Finds

Front half (face) of human head with tapering neck extends from the bottom. The item was cast in a reddish, copper alloy (apparently bronze). The back is flat and slightly convex; weight: 24gr. The eyebrows are connected, forehead flat and chin prominent.

Face of a man or a god (?) with two small horns, a small beard, with a stylized hat on the head. The item was cast in reddish copper alloy (apparently bronze); weight: 38gr. The back is flat with a perpendicular protrusion used to affix the item to some kind of vessel. Remains of lead or tin solder, used to solder the item to the vessel, can be seen on the back side. The image is similar to some metal Canaanite gods

Fig. 1: Bronze head with tapering neck (IAA)

Fig. 2: Bronze head with horns and beard – oblique view (IAA)

Human Head with Long, Pointed Neck (002/3-31/93, Fig. 1)

266

Jewelry and Adornments from the Akko Marina

Fig. 5: Stone pendant (IAA) Fig. 3: Bronze head with horns and beard – back view (IAA)

Decorated Bracelet (191/36.8–A-31/93, Fig. 6) The item is made of a reddish copper alloy (apparently bronze); weight 52.5gr. It seems to have been cast in the lost-wax technique. The body of the bracelet is elliptical in cross-section (6.8–8.5mm) and oblate at both ends, creating a kind of duckbill shape at the ends. Between the oblate ends and the decorated body of the bracelet are two protrusions on the bracelet outside, each measuring a few millimeters. The exterior and sides of the bracelet feature a ropelike decoration; the inner side is smooth. Based on the interior diameter, it is reasonable to assume that the bracelet was to be worn by a child.

Fig. 4: Bronze head with horns and beard – frontal view (IAA) in “Anatolian” and smiting poses from Syria and Lebanon which were dated to the Late Bronze age and the Iron Age 1. (Negbi 1976: pl. 14-168; plate 21-1317; pl. 27-1393). Stone Pendant (1004/36-A-93/31, Fig. 5) A rectangular, box-shaped object made of gray stone with a perforation in its upper part (thickness 7.5mm, weight 9gr). It seems to have served as a kind of pendant.

Fig. 6: Bronze bracelet with rope decoration (IAA)

267

Ehud Galili

The head of the pin is made of one of the bands that created the pin itself that was wrapped around the end, producing the pin’s thickened head (diam. 3.5mm), which keeps it from slipping free off the ring. The pin is secured by a thin, round-sectioned rod (diam. 0.5mm) that passes from one side of the outer ring, through the pin, to the other side of that ring. This item seems to have been a clasp for a necklace or bracelet.

Cross Pendant (002/1-31/93, Fig. 7) A Maltese cross made of a yellowish alloy (apparently brass); weight: 3.5gr. The cross is made of a sheet 1.5mm thick, whose upper part is oblate and features a perforation for hanging. The item was apparently made by cutting a flat metal sheet into the desired shape. The upper part has been flattened by hammering perpendicular to the rest of the the sheet and then perforated to accommodate a stringing hole. The hammered decoration consists of five circles with a dot in the center of each – one circle at the center of the cross and one on each arm. Pendants of this type were common in the Christian world and can apparently be dated from the Byzantine period on.

Fig. 8: Gold clasp (IAA)

Acknowledgments Fig. 7: Bronze cross-shaped pendant (IAA)

The author wishes to thank the Israel Antiquities Authority for the institutional support, M. Feinberg Vamosh for the English editing, T. Sagiv and C. Amit for the photographs, E. Altmark for cleaning the metal artifacts and Amir Golani and Baruch Rosen for their useful remarks on the manuscript.

Gold Clasp or Chain Link (191/36/8-A-93-31/93, Fig. 8) A rectangular, box-shaped item (thickness 2.1mm). Four rings (diam. 2.1mm) are soldered as a row on its top and three on its bottom. A pin with a flaring head was found inserted into the rings on one side. The box is made of a rectangular, decorative strip which was folded to a box shape and soldered at its ends. The strip is decorated with three lengthwise lines on the upper part, and three additional lines on the opposite side. Between the these sets of lines is a layer of wound gold thread creating hollow circles. The gold pin that passes through the rings is made of elongated rod, plano-convex in cross-section. The rod is bent double 180 degrees (creating two parallel bands with a space between them), making the pin round in cross section (external diam. 2mm).

References Galili E. This volume a. The Akko Marina Archaeological Project – Introduction. Negbi O. 1976. Canaanite Gods in Metal. Tel Aviv.

268

Chapter 14

Metal Objects from the Akko Marina Ehud Galili Israel Antiquities Authority and Zinman Institute of Archaeology, University of Haifa Key words: underwater archaeology, shipwrecks, ancient bronze, cultural heritage, Acre, dredging

Introduction Various metal objects were retrieved during the dredging of the Akko Marina (Galili this volume a). Most were located by a metal detector after the dredged material was deposited in the open sea outside the harbor. Among the metal objects were various fixtures and parts of tools, some of whose functions are not clear. Other artifacts can be securely identified by their form, especially those that constituted a complete finished object. Because of the large number of artifacts and because they did not occur in a clear archaeological context, they are described technically, with no reference to date or origin. The numbers in parentheses are identification numbers of the objects retrieved during underwater surveys or during the deepening of the marina (Galili this volume a).

Description of Objects Buckles Large buckle (31.93-A-17/273; Figs. 1, 2): A particularly large buckle (length 19cm), H-shaped, having two thin, asymmetrical heart-shaped metal plates (0.7mm thick) connected to either end of the form. The form was made of two parallel, flat metal shafts to which a perpendicular, round metal shaft was attached. The two thin, flat metal plates, in a

Fig. 1: Large buckle (IAA)

Fig. 2: Large buckle, decorations (IAA) heart or a broad leaf shape, bear a design of triangles formed by oblique lines. The plates are incised with a geometric design around the edge (two parallel rows of dots interspersed with triangles). The inner space is decorated with a stylized plant. The object (weighing 89gr) is made of a yellowish copper alloy (brass). This buckle was apparently a part of a particularly broad leather or cloth belt associated with a fine ceremonial garment. Belt buckle (005/2-3193; Fig. 3): This buckle consists of two metal pieces connected by means of a hinge, a cast frame consisting of one part that is a stylized, jointed arc (resembling a crab’s leg) and another, slender, straight part, round in cross-section, closing the arc and constituting a hinge. A metal plate is attached to and bent around the hinge, creating two parallel sheets connected to each other by three pins. One spherical-headed pin survived in its original location; there were holes at the points where the other pins would have been. The pins served to tighten the two sheets of metal. Between these sheets was apparently a strip of organic material (cloth? leather?) that has vanished. The object (weighing 22gr) was formed of yellowish copper alloy, apparently brass. A portion of cast buckle (005/4-31/93; Fig. 4): The artifact is composed of a cast frame of an arc-shaped buckle. Its lower part is in the form of

269

Ehud Galili

Hinges/Fasteners Long hinge (31/93A-38/83; Fig. 5): A part of a particularly long hinge (length 18.5cm) made of a cast plate having two round protrusions on one side and one round protrusion on the other. The protrusions are punctured by holes (diam. 4.5mm). This item (weighing 109gr) was of a yellowish alloy (brass). It may have been used as a hinge or a fastener for a box or a door or as part of a decorative chain used to suspend a light fixture.

Fig. 3: Belt buckle (IAA)

Fig. 4: Cast portion of a buckle (IAA) a straight shaft having a round cross-section. The object (weighing 12gr) is made of a reddish copper alloy, apparently bronze. It seems to have been part of a belt made of leather or similar material.

Fig. 5: Part of a long hinge (IAA)

270

Metal Objects from the Akko Marina

Segments of a hinge (31/93-A-36.8/184; 31/93A-36.8/112; 31/93-B-36/1011; Figs. 6, 7, 8): Three items (weighing 38–45gr each, thickness 3.5 to 4.5mm) are made of an orange-colored copper alloy, apparently bronze. They feature one perforated protrusion at the top and two at the bottom; diam. of perforations: 4.0–4.5mm. The objects were cast of one piece; it seems that all three were cast in one mold; however, the depressions and protrusions at the ends differ and therefore it seems to have been hand-finished by filing or planing after casting. The front part of these objects is crossed by a decoration of four parallel lines, two at the top and two at the bottom. The objects fit into one another and they may have been a chain for hanging some object or the hinges of a box or fasteners for wooden boxes.

6

7

Chisel (31/93-C-35/1160; Fig. 10): This item was severely corroded (length 125mm, weight 77gr). Half of it (the handle?) was round in cross-section (diam. 13mm) and half was flat (26mm across, 5.5mm thick). This object may have served to plane wood or as a caulking iron.

8

Figs. 6,7,8: Portions of small hinges (IAA) Iron Items Ring (31/93-C-8/1057; Fig. 9): Ring (diam. 7.7cm, weight 139gr), round cross-section (diam. 10mm), slightly open (width of opening 0.7mm). This ring may have been connected to a pier or a boat and used for tieing.

Fig. 10: Iron chisel (IAA) Bronze Items Key (005/6-31/93; Fig. 11): This key is made of a yellowish alloys (bronze) cast in one piece in a double mold. The key (weighing 28gr) is made of a central shank, elliptical in cross-section (diam. 6.5–8.5cm) (the elliptical cross-section is due to the misalignment of the two halves of the mold). In the upper part of the key is an oblate ring bearing signs of filing, apparently made when finishing the cast blank.

Fig. 9: Iron connecting ring (IAA) 271

Ehud Galili

through which a spiked wheel would have been inserted. The wheel, which would have been held in place by the two protrusions, did not survive. The item was cast in a reddish copper alloy, apparently bronze. Pair of balance pans (005/12-31/83, 005/11-31/93; Figs. 13, 14): The pans were made of copper plating (thickness 1–1.5mm), hammered and concave. Each pan consisted of two strips, the edges of which were cut, comb-like, and hammered together. Three rings were attached to each pan. Each ring consisted of a flat bronze tab with a long protrusion (resembling a spoon) connected to the bottom of the pan by a pin. The tab’s long arm was rolled in such a way as to create a tie-ring protruding above the upper part of the pan. The resemblance between the weight (No. 005/11-31/93 weighs 510gr and No. 0512-31/93 weighs 495gr) and form of the pans indicate that they belonged to one balance. Weighing apparatus (steelyards and balances) were common on sailing vessels and have been recovered from shipwreck sites along Israel’s coast and elsewhere in the Mediterranean. These weights may have been used to weigh fish in boats returning from fishing.

Fig. 11: Key made of copper alloy (IAA) Horseman’s shoe spur (31/93-C-31/1382; Fig. 12): This object (weighing 110gr) is made of two arms at the end of which is a ring apparently intended to attach the spur to a shoe. In the back are two small, parallel protrusions; at the end of each is a hole

Fig. 13: Bronze scale pan (IAA)

Fig. 12: Spur for horseman’s shoe from copper alloy (IAA)

Fig. 14: Bronze scale pan (IAA) 272

Metal Objects from the Akko Marina

Parts of Metal Containers Two rims of containers, one made of lead and the other apparently of tin, were retrieved, as well as the bottom of a lead bowl and part of lead bowl (external diam. ca. 65mm) lacking its bottom. Decorated bowl rim (005/1-31/93; Fig. 15): Rim of a small bowl (diam. 78mm, weight 46gr), made of cast lead. A row of round protrusions appears in relief on the rim’s exterior. The rim was apparently connected to a metal vessel that did not survive.

Bronze bell (005/10-31/93; Fig. 17): A bell cast in a reddish alloy (apparently bronze) with a broad base and an arc-shaped loop at the top divided in two by a vertical part. On the inner, convex part of the bell a ring was cast for attaching the clapper. Fishermen attach bells to a pole protruding from a fishing net buoy in order to find the net at night. Fog bells were used by sailing vessels to prevent collisions at night or when visibility was otherwise compromised.

Fig. 15: Rim of a decorated lead bowl (IAA) Varia Lead strainer (31/93-A-36/1015; Fig. 16): Lead plate (thickness 0.5mm, weight 20gr) with perforations. This object could have been used as a filter for a water pump or a bilge pump, which prevented foreign bodies from getting into the water during pumping. Button (005/8-31/93): A convex button, cast in bronze (diam. 25mm, thickness 0.5mm, weight 7gr). In the center of the inner, concave side, is a loop with a 3mm-eye for affixing a thread; the loop was cast together with the button.

Fig. 16: Lead strainer (IAA)

Fig. 17: Bronze bell (IAA) Part of a bronze padlock (005/7-31/93; Fig. 18): A padlock cylinder consisting of two parts. The part that was discovered was cast in a form of a hollow cylinder to which a right-angled arm was attached. The part discovered (weighing 50gr) was made of reddish copper alloy (apparently bronze). It was closed at the top, except for a cross-shaped opening through which the key was inserted. On the upper part of the arm was a small hole (diam. 2mm) to which a string would have been attached to suspend or carry the lock. The cylinder was the body of the lock, the part into which the key was inserted. The arm that protruded from the cylinder would have been passed through two metal rings, one attached to the door and the other to the threshold, or, in the case of a box, to its cover and receptacle. The rod portion of the lock, which was not retrieved, would have been round in cross-section with protrusions on one end and a ring on the other. To close the padlock, the rod with 273

Ehud Galili

section. Various decorations were incised on one side of the object: On its flat, circular portion the image of a bird was incised, facing left and holding a circular object in its mouth. At the other end, a number of lines are incised. In the long, flat parts on either end are two parallel rows of circles, between which is a central line parallel to the long axis of the object. The images of two schematic fish? facing each other can be seen on this part of the artifact. The object (weighing 30gr) is made of reddish copper alloy (apparently bronze) worked by annealing. It may have been used as a pointer for reading or as a fulcrum for pan scales.

the protrusions would be inserted into the hollow cylinder, so that the rod passed through the ring at the end of the right-angled arm. Iron locks of this type were discovered in a fifth-century shipwreck off the coast of southern Turkey (Pulak 2004). Similar locks, made of bronze, were discovered in the excavations of Corinth (Davidson 1952) and in a shipwreck from the Late Roman period discovered off the Carmel coast (Galili et al. 2010).

Bronze item with pointed tip, resembling a leaf (31/93-A-36/19; Fig. 20): The object, resembling a leaf (weighing 38gr), was made of a thin (thickness 0.7mm) sheet of yellowish copper alloy (apparently brass). The item has three holes, two of which are square (4 × 4mm) and one of which is almost circular (diam. 5.5mm). The holes were used to affix the object by means of pins. This object may have been used as a decoration attached to a piece of wooden equipment or other item. Two connected bronze pipes (31/93-C-59/26; Fig. 21): This item (weight 69gr) consists of two pipes, hexagonal in cross-section, made of yellowish copper alloy (apparently brass). They were soldered together along their length by means of two flat sheets. At the end of each pipe (outer diam. 10mm) a kind of a plug made of bronze was soldered. The plug is cylindrical and domed. At its base is a zigzag decoration and holes. The remains of similar domes can be seen; similar decoration, only part of which survived, can be seen at the other end of the pipes. A small ring (diam. c. 5.5mm), made of tin, was soldered to one of the thin sheets that connected the two pipes. Another part was connected to the opposite side, behind the ring. It is unclear what this item was used for; it may have been a sailor’s whistle or a container for a cosmetic brush, pen, needles, sail needs?

Fig. 18: Bronze padlock (IAA) Unidentified Metal Objects Various objects made of copper alloy (apparently bronze or brass), which were not clearly identified. Bronze rods connected by pins (005/10-31/93): A rod made of reddish copper allow (apparently bronze), square in cross-section (13.5 × 13.5cm, weight 480gr). The rod was annealed and bore clear impressions of hammer blows. Another rod was attached to the end of the first one by means of three bronze pins. The component was attached by cutting both rods obliquely and connecting them by means of the bronze pins. Similar rods were used to affix the planks of large sailing vessels to the hull.

U-shaped bronze clasp (005/5-31/93; Fig. 22): A U-shaped object made of a yellowish copper alloy (apparently brass). The object is rhomboid in crosssection (0.3 × 0.3mm, weight 8.5gr). On each of its sides is an oblate space with three holes, at the end of which is a ring. Remnants of pins can be seen in the holes. The use of the item is unclear.

Bronze item with pointed tip and decorations (31/93-A-17/279; Fig. 19): A flat, elongated item (thickness 1.5mm) with a flat circle at one end, from which emerged a pointed tip that was square in cross-

Fig. 19: Elongated bronze sheet with decoration (IAA) 274

Metal Objects from the Akko Marina

Fig. 20: Leaf-shaped bronze object (IAA)

Fig. 21: Pair of attached pipes (IAA)

A mandolin-shaped bronze object with three holes (002/4-31/93; Fig. 23): An object resembling a bell clapper or pendulum (weight 20gr) made of a yellowish copper alloy (apparently brass). The item consists of a shaft rectangular and flat in crosssection, one end of which is round and oblate with two rectangular holes, and the other of which domeshaped with one hole. Raw Lead for Metalwork Waste from lead casting (00/9-31/93; Figs. 24, 25): Eight amorphous, unworked pieces of lead with a total weight of 256gr, apparently waste from the casting process that took place aboard a ship or in port. It was very common in antiquity for sailors to manufacture their own maritime equipment from lead; this is still

Fig. 22: U-shaped clasp (IAA) 275

Ehud Galili

Acknowkedgments The author wishes to thank the Israel Antiquities Authority for the institutional support, M. Feinberg Vamosh for the English editing, T. Sagiv and C. Amit for the photographs, E. Altmark for cleaning the metal artifacts and the IAA illustrators for drawing the artifacts.

References Davidson G.R. 1952. Corinth – The Minor Objects, Vol. XII. Princeton, New Jersey.

Fig. 23: Mandolin-shaped bronze object (IAA)

Galili E., Rosen B. and Sharvit J. 2010. Artifact Assemblage Recovered From a Roman Shipwreck Off the Carmel Coast, Israel. ‘Atiqot 63: 61–110. Galili E. This volume a. The Akko Marina Archaeological Project – Introduction. Pulak C. 2004. Matthews, J.R. Jr.(eds.). Serçe Shipwreck, Vol.

The Padlocks. In G.F. Bass, S. Steffy and F.H. van Doorninck

Limani: An Eleventh-Century

1: The Ship and its Anchorage, Crew, and Passengers. Pp. 437–452. College Station, Texas.

Rosen B. and Galili E. 2007. Lead Use on Roman Ships and its Environmental Effects. The International Journal of Nautical Archaeology 36.2: 300–307.

Fig. 24: Waste from lead-casting (IAA)

Fig. 25: Waste from lead-casting (IAA) the practice among traditional fishermen and sailors. Lead is easily worked; it can be cut and cast aboard into various shapes. On fishing boats it is used in net sinkers and in sounding weights (depth-measuring devices) (Rosen and Galili 2007). Remnants of lead were collected, melted down and recast for various uses on ships, including to weigh down anchors and oars and to seal holes on deck (Galili et al. 2010).

276

Chapter 15

Copper Bun-shaped Ingot from the Akko Marina Ehud Galili (1) and Jacob Sharvit (2) 1. Israel Antiquities Authority and Zinman Institute of Archaeology, University of Haifa 2. Israel Antiquities Authority Key words: marine archaeology, underwater archaeology, Acre, Late Bronze Age, metal ingot

Introduction

The Ingot and Similar Finds

During August 1996, following the Akko Marina deepening project (Galili this volume a), an underwater survey was carried out in the dumping area of barge no. 7 (unpublished IAA diving report no. 37/96-32/1). The barge carried material collected from the entrance to the Akko Marina (Galili this volume a: fig. 3 Area C). During the survey an oval, bun-shaped copper ingot was recovered.

The ingot (Fig. 1) is plano-convex (length 26cm, width 13.5cm, weight 4.56gr). Its convex side is relatively smooth, while its other side is mostly flat, except for three bulges adjacent to the edges. Scores of similar bun-shaped ingots, one-holed stone anchors and associated artifacts were recovered during underwater surveys off Neve-Yam on the Carmel coast, some 30km southwest of Akko (Fig. 2) (Galili et al. 1996; 2011). The Neve-Yam ingots were dated to the Late Bronze Age–Early Iron Age, according to the one-holed stone anchors and a bronze chisel recovered together with the ingots and probably originating from the same wreck. A lead isotope analysis of the Neve-Yam ingots suggested that they originated in the Araba Valley mines (Yahalom-Mack et al. 2014).

Fig. 1: Bun-shaped copper ingot from the Akko Marina (E. Galili)

Fig. 2: Cargo of Late Bronze Age-Iron Age copper ingots and stone anchor from Neve-Yam (E. Galili) 277

Ehud Galili and Jacob Sharvit

Another Bronze Age cargo of copper oxide and bun-shaped ingots and associated one-holed stone anchors was recovered from the Uluburun shipwreck, Southern Turkey, and was dated to the fourteenth– twelfth centuries BCE (Pulak 1998: 194, fig. 7). Similar ingots were recovered from excavations at Alghero, near the northwestern coast of Sardinia (Lo Schiavo 1981: 281, fig. 284; 285, fig. 297).

References Artzy M. 1992. Tel Nami. In E. Stern (ed.). The New

Encyclopedia of Archaeological Excavations in the Holy Land 3. Pp. 1095–1098. Jerusalem.

Artzy M. 1997. Nami, Tell Abu-Hawam and the Nomads of the Sea, West and East: Connections

between the Western and Eastern Mediterranean at the End of the Late Bronze and Beginning of the Iron Age – New Evidence – Colloquium. University of

Discussion and Conclusions

Haifa. Haifa.

The large number of bun-shaped copper ingots from shipwrecks discovered off the Carmel coast and Uluburun indicates that during the Bronze Age ships trading with south Levantine coastal sites played an important role in the “international” trade in copper ingots around the Eastern Mediterranean basin. Remnants of metal-working activities, including crucibles and a furnace, were found in late thirteen century BCE levels on the MB IIA ramparts of Akko. Imported Aegean-style (Mgc III6) and Cypriot ceramics were also found at Tel Akko (Artzy 1997). It is possible that the copper bun-shaped ingot recovered from the Akko Marina was part of a cargo imported from Cyprus or other Mediterranean source, e.g., the Arava mines, and was destined for the local Bronze Age industry in Akko or was shipped abroad from Akko port. Traces of a Late Bronze Age copper industry, including finished products and metal casting waste, were found at Tel Nami, some 32km southwestof Akko (Artzy 1992: 1027). The Akko ingot and the copper ingots from Neve-Yam (Galili et al. 1996; 2011; YahalomMack et al. 2014) indicate that several coastal sites along the northern coast of Israel were active in the “international” metal trade that took place during the Late Bronze and Early Iron Ages.

Galili E., Sharvit J. and Shifrony A. 1996. NevéYam: Underwater Survey. Hadashot Arkeologiyot 106: 54–56. (Hebrew). Galili E., Gale N. and Rosen B. 2011. Bronze Age Metal Cargoes off the Israeli Coast. Skyllis 11: 64–73. Galili E. This volume a. The Akko Marina Archaeological Project – Introduction. Lo Schiavo F. 1981. Economia e Società Nell’età dei Nuraghi. In G.P. Carratelli (ed.). Ichnussa La Sardegna dalle origini all’età classica. Pp. 281–285. Milano. Pulak C. 1998. The Uluburun Shipwreck: an Overview. International Journal of Nautical Archaeology 27.3: 188–224. Yahalom-Mack N., Galili E., Segal I., Eliyahu-Behar A., Boaretto A., Shilstein S. and Finkelstein I. 2014. New Insights into Levantine Copper Trade: Analysis of Ingots from the Bronze and Iron Ages in Israel. Journal of Archaeological Science: 45: 159–177.

Acknowledgments The authors wish to thank the Israel Antiquities Authority for the institutional support, M. Feinberg Vamosh for the English editing and B. Rosen for his useful remarks.

278

Chapter 16

Stone Finds from the Akko Marina Ehud Galili Israel Antiquities Authority and Zinman Institute of Archaeology, University of Haifa Key words: marine archaeology, Acre, stone bowl, ashlars, marble, column

Introduction

marine organisms. Two small, symmetrical crosses were incised on the flat, well-worked part of the slab. Each cross consists of four triangles connected at their apex with a swallowtail base (a version of a Maltese cross). A straight groove is incised along the length of the slab between the crosses. At the top of the groove is a circular, symmetrical depression. The straight groove may have been part of a larger cross that decorated the original object. Similar decorations on marble slabs usually appear on chancel screens, railings and ecclesiastical architectural elements. The elliptical object seems to have been made from a larger slab, perhaps a decorative tablet, used for another purpose. The original artifact could have been depicted with a large cross, having four smaller crosses, depicted between its arms. It is known as “Holy land Cross” or “Jerusalem Cross”, which was widely used by the Castodia Terra Sanctae. The finish of the original slab and the decorations it bears, which were carefully and aesthetically executed, show that it was probably chiseled by an expert craftsman, possibly from abroad. However, the elliptical outlines of the object and its preparation for secondary use were poorly executed with simple tools, apparently by a local stonemason. The symmetrical outlines of the object and the location of the decorations, intentionally and symmetrically centered, show that the original artifact was reshaped for some secondary use associated with the decorations. An elliptical stone slab, very similar in decoration, outline and dimensions, was found in the excavations of the Galei Kinneret Hotel on the shore of the Sea of Galilee. That slab was found in the ruins of an Umayyad structure that was destroyed in the earthquake of 749 CE (Moshe Hartal, pers. comm. 2008). It seems that the Tiberias slab predates 749 CE. Marble chancel screens were commonly imported for use in churches in Provincia Arabia and Palaestina Tertia until the end of the Byzantine period and the early Umayyad period, which saw a sharp decline in the imports of such items (Habas 2008). In underwater surveys off the coast of Israel anchors made of marble were discovered that had been reworked for secondary use from church chancel

Various stone items emerged from the material dredged from the marina and during underwater surveys (Galili this volume a). Among the finds were dressed building stones, architectural elements and finished items of various types. The finds are described below. Other stone finds are described elsewhere in this volume (stone anchors – Galili and Rosen this volume b; stone cannon balls – Galili and Rosen this volume d; stone fishing gear – Galili and Rosen this volume c).

The Finds Marble Slab Decorated with Crosses (004/1-31/93, Fig. 1) This slab was found in Area C (Galili this volume a: fig. 3), near the entrance to the marina, at a depth of 3m. It weighs 2.9kg and was elliptical in shape (length 30cm, width 19cm, thickness 2.5cm). Its upper part was flat and carefully worked and its lower part was roughly worked. The item was pocked with holes and seems to have been subject to erosion by

Fig. 1: Marble slab decorated with crosses (E. Galili) 279

Ehud Galili

food processing onboard. Similar stone bowls were discovered in underwater surveys in the northern anchorage at Atlit (Israel Antiquities Authority Diving Report No. 61/1-15/91, unpublished).

screens or tombstones. A stone anchor with three perforations, decorated with crosses and a vegetal design, was found at the anchorage at Yavneh-Yam (Galili et al. 1998). Another anchor, made of a marble slab in secondary use, was found near the Byzantine city of Asheklon-Maiumas, west of the northern neighborhoods of Ashkelon (Galili et al. 2012). The use of the elliptical slabs found at Tiberias and Akko is unclear. However, their resemblance to each other is not coincidental; it may reasonably be posited that they were made for a clearly defined, ritual, ceremonial or other useful purpose.

Bowl Made of Reddish Stone (004/3-31/93, Fig. 3) This bowl (diam. 49.5cm, width of surviving portion 3.5cm, thickness of wall 2.0–2.5cm, weight 13.5kg) was discovered during a survey dive in Area B (Galili this volume a: fig. 3) at a depth of 2.5 m. The base of the bowl was flat and its sides are broken. The vessel was made of reddish, non-local stone, whose origin is unknown.

Rectangular Bowl (59/2-26/92, Fig. 2) This bowl, made of kurkar (length 32cm, width 26cm, height 23cm, weight 36.5kg), was discovered on the surface during a survey dive, east of Area A,

Column Made of Kurkar (004/4-31/93) A column drum with a round cross-section (length, 100cm, diam. 60cm), was discovered in Area F near the modern northeastern wooden pier, at a depth of 3m (Galili this volume a: fig. 3). Ashlars (004/5-31/93) Dozens of kurkar ashlars were found among the material removed from the marina. The stones were of various sizes (length 80–150cm, width 40–70cm, thickness 40–60cm). Most were discovered in Area F (Galili this volume a: fig. 3). On some of the stones measuring 80 × 40 × 40cm, grooves were carved to attach metal clamps. In the southwestern corner of the marina a group of six ashlars were discovered measuring 150 × 40 × 40cm. Among the stones are some of dressed marble, along with blocks of raw stone (Galili this volume a, fig. 3: 21).

Fig. 2: Rectangular stone bowl (J. Galili) Millstone (004/6-31/93) some 30m northwest of the Island of Flies (Galili this volume a: fig. 2). This item may have been used as a basin for washing or purification before prayers or for

A kurkar slab worked into a semi-circular disk (length 70cm, width 35cm, height 20cm) apparently served as part of a round millstone consisting of two half-discs connected by means of a circumscribing metal band. The stone was found in an area in which material removed from the marina was placed. A similar millstone was found in an underwater survey conducted north of Ashkelon (IAA Dive Report 45/237/96, unpublished).

Discussion and Conclusions We may reasonably assume that some of the ashlars discovered in and around the marina were part of marine and coastal structures that had collapsed. Other stones, after having been dismantled from structures at various coastal sites, were apparently

Fig. 3: Bowl made of red stone (J. Galili)

280

Stone Finds from the Akko Marina

transported by ship to the city of Akko for secondary use and for construction. Ancient architectural elements were reused in some of the Crusader and Ottoman structures in the coastal cities. In the sea wall of Ashkelon, in the harbor fortress in Caesarea and in Khan al-Umdan in Akko, granite columns are incorporated that were taken from Roman structures. Historical evidence reveals trade in building stones dismantled from ancient coastal cities during the Ottoman period. Cargoes of building stones for secondary use have been found in various locations along the coast of Israel: a cargo of columns and large ashlars were found west of Akko at a depth of ca. 16m (Galili 1987; Galili this volume f). At the Israel Shipyards near the Kishon River estuary, a cargo of kurkar stones that had been dismantled from a structure was found. The stones were probably intended for transport by sea (Sharvit and Galili 2001). On the northern Carmel coast near Haifa, a cargo of dozens of dressed kurkar stones bearing remnants of plaster were found, intended for secondary use (Galili and Sharvit 1999). Additional cargoes of dressed kurkar stones that were transported by sea were found at the south anchorage of Dor, at the Neve Yam bay, at Kfar Galim and at the Apollonia anchorage (Galili et al. 1993). Dressed marble and raw marble blocks, apparently imported by sea from Anatolia or Europe were also found. The findings from the Akko Marina and the cargoes of stone discovered in underwater surveys along Israel’s coast show extensive commerce in imported, or locally quarried, building stones in ancient times.

References Galili E. 1987. Underwater Survey. Hadashot

Arkheologiyot 89: 59–60. (Hebrew).

Galili E., Dahari U. and Sharvit Y. 1993. Apollonia.

Hadashot Arkheologiyot 91: 97. (Hebrew).

Galili E., Sharvit, Y. and Bahat-Zilberstein N. 1998. Yavneh-Yam, Underwater Survey. Hadashot Arkheologiyot 106: 117–118. (Hebrew). Galili E. and Sharvit Y. 1999. Underwater Surveys.

Hadashot Arkheologiyot 110: 19–24. (Hebrew).

Galili E., Rosen B. and Finkielsztejn G. 2012. Stone Anchors Bearing Inscriptions from Ashkelon. ‘Atiqot 71: 113–127, 121–122. Galili E. This volume a. The Akko Marina Archaeological Project – Introduction. Galili E. and Rosen B. This volume b. Anchors from the Akko Marina. Galili E. and Rosen B. This volume c. Fishing Gear from the Akko Marina. Galili E. and Rosen B. This volume d. Weapons and Ammunition from the Akko Marina. Galili E. This volume f. Ships and Cargoes in and around the Akko Marina. Habas L. 2008. The Art of Imported Marble Chancel

Acknowledgments The author wishes to thank the Israel Antiquities Authority for the institutional support, M. Feinberg Vamosh for the English editing and B. Rosen for his useful remarks.

Screens and its Influence on Local Production in the Churches of Provincia Arabia and Palaestina Tertia: a Case Study. Proceedings of the Soma 12th Symposium on Mediterranean Archaeology. Famagusta. Sharvit Y. and Galili E. 2001. Israel Shipyards. Hadashot Arkheologiyot 113: 172. (Hebrew).

281

Chapter 17

Weapons and Ammunition from the Akko Marina Ehud Galili (1) and Baruch Rosen (2) 1. Israel Antiquities Authority and Zinman Institute of Archaeology, University of Haifa 2. Israel Antiquities Authority Key words: marine archaeology, Acre, cannon balls, lead shot, iron shell

Introduction

This metal piece (thickness 0.3mm), corroded and partly covered with marine encrustation, was made of a copper alloy. Remains of the weapon, which was made of iron and had completely oxidized, could still be seen inside the scabbard.

Akko, which was the most important fortified city in northern Israel, was the scene of land and sea battles throughout its history. The archaeological project conducted during the deepening of the Akko Marina (Galili this volume a) revealed weapons and ammunition attesting to military actions in the area of the marina and reinforcing historical accounts. Historical documents describe a number of battles in and around Akko: From March to May of 1799, during Napoleon’s siege on the city, thousands of shells and cannon balls were fired, some of which fell in the harbor (Shor 1990: 253–254; Kahanov et al. 2014 a; Kahanov et al. 2014 b). The last naval battle on the coast of Akko, which certainly would have left signs and remnants of ammunition in the marina, took place in late 1840, when shells from the cannons of the British royal fleet blew up Akko’s main ammunition depot, leading to the city’s surrender (Anderson 1952: 561–565). Some of the projectiles found in the marina are likely to have come from these battles.

Fig. 1: Metal-covered wooden scabbard for dagger or sword (IAA)

Ammunition

Weapons

Stone Spheres

Two Parts of a Sword or a Dagger

Two stone spheres were discovered. The larger one (Fig. 2) was made of limestone (weight 16.8kg, diam. 23cm) and the other was made of basalt (diam. 5.8– 6.0cm, weight ca. 220gr). Similar stone shot made of limestone and basalt was found in a shipwreck from the fifteenth century discovered in the northern anchorage of Atlit (Galili et al. 2002: 242; Galili 1986; Galili and Rosen 2014). This item may have been a ballista ball or any early cannon shot. The small basalt sphere may have also been a shot, fired from a small cannon. Stone cannon shots were in use beginning in the fourteenth century; their replacement by iron shots was gradual, which means that such artifacts not discovered in a defined archaeological context cannot be dated with certainty.

One part (length 33–34cm, width 5cm, thickness 0.8–1.2cm) is a relatively well-preserved scabbard’s midsection. It consisted of two parallel wooden slats held together by an organic adhesive that had eroded away. Between the two slats was a space for the weapon. In that space were pieces of oxidized iron – remnants of the weapon. On one end of the scabbard was a corroded band made of a copper alloy, which held the two wooden slats together. The scabbard was covered with a thick layer of marine encrustation. The second item (length 12.5cm, width 3.0–3.5cm, thickness ca. 0.8cm) is the point of the scabbard. It is made of wood that was preserved within the metal piece protecting the weapon’s tip (Fig. 1).

282

Weapons and Ammunition from the Akko Marina

Fig. 4: Hollow iron cannon shells (E. Galili) Hollow Projectile with a Wooden Fuse (Fig. 5): A hollow projectile made of cast iron, which had disintegrated due either to oxidization or from impact after being fired but not exploding. The fuse, inserted in the projectile, was made of hollow wood and contained a yellow substance, perhaps remnants of an explosive. The projectile was ca. 11cm in diameter; the outer diameter of the fuse was 10mm. Such shells were in occasional use on warships beginning in the late eighteenth century. After about 1830 they became the main projectiles fired from naval cannons (Hough 1969: 209–213).

Fig. 2: Cannon shot or ballista missile made of limestone (IAA) Iron Cannon Balls Thirty-two iron spheres of varying diameters were found, some of which were intact and some broken. Among these, 20 were hollow (projectiles containing explosives) and the rest were solid shot (Fig. 3). Tables 1 and 2 present a representative sample of the data on the iron shells and shot that were found, whose state of preservation allowed them to be weighed and measured. It should be noted that corrosion and marine incrustation caused slight changes in their dimensions and weights. We may reasonably assume that most of the iron shells and shot were used in the Akko area mainly from the eighteenth century on. They attest to naval battles that took place around the city beginning Napoleon’s times. Hollow iron cannon shells (Fig. 4, Table 1) appeared only in the nineteenth century. The variety of diameters of shells found (at least five different diameters) shows that numerous cannon types were being used by city attackers and by its defenders.

Fig. 5: Hollow shell that disintegrated without exploding. The remains of the wooden fuse can be seen (E. Galili) Lead Shot This ammunition was intended for small-caliber personal weapons (pistols and rifles). The finds indicate that in addition to cannons, light weapons, fired at relatively short range, were also used in

Fig. 3: Solid iron shells (IAA) 283

Ehud Galili and Baruch Rosen

16–17mm, weight 26–27gr); medium-size (diam. 13.0–14.5mm, weight 14.0–18.5gr); and small (11.0–12.5mm, weight 1–12gr). Some of them bore casting signs, a ridge surrounding the projectile, left at the joint of the mold parts and a ‘tail’ marking the lead casting funnel. Few small lead shot were found, apparently from muskets, pistols or shrapnel from cannons. Marine corrosion probably caused slight changes in the shots’ weight and dimensions. Based on the diameter of the projectiles, it appears that at least three types of light weapons were used in the harbor and/or in the city of Akko during battles.

battle in and around the city. Intact lead shot were found (with no evidence that they had hit a target), apparently after having fallen into the water from anchoring ships or from ships wrecked in the harbor. They may also have been fired and sunk after hitting the water (Fig. 6). Other lead shot were damaged and crushed when they struck a hard target and then sunk into the sea (Fig. 7). Some of these shots bore signs that they had struck wood (Fig. 8), and they may have been fired at ships from the shore. Some 45 shots were found that had been fired and crushed (ricochets). These shots had lost some of their weight as a result of material lost on impact. They were heavily distorted, precluding measurements of their original diameters. Their average weight was between 12 and 13gr. The intact lead shot found, which had preserved their form, were divided into three groups: large (diam.

Acknowledgments The authors wish to thank the Israel Antiquities Authority for the institutional support, M. Feinberg Vamosh for the English editing and T. Sagiv and C. Amit for the photographs.

References Anderson R.C. 1952. Naval Wars in the Levant. Liverpool. Galili E. 1986. Atlit Coast – Ship’s Cargo. Hadashot

Arkheologiyot 88: 10. (Hebrew).

Fig. 6: Lead shot not crushed (IAA)

Galili E., Raban A. and Sharvit J. 2002. Forty Years of Marine Archaeology in Israel. In H. Tzalas (ed.). Tropis VII, Proceedings of 7th International Symposium on Ship Construction in Antiquity, Pylos, Greece. 1999. Pp. 927–961. Galili E. and Rosen B. 2014. A 15th-Century Wreck of an Ordnance-Carrying Ship from Atlit North Bay, Israel. International Journal of Nautical Archaeology 43.1: 115–127.

Fig. 7: Lead shot crushed after firing and striking a hard surface (IAA)

Galili E. This volume a. The Akko Marina Archaeological Project – Introduction. Hough R. 1969. Fighting Ships. New York. Kahanov Y., Stern E., Stern A., Ronen R., Cvikel D., Ashkenazi D. 2014 a. ‘What ship? Who fired the cannonballs at the wall in Akko? An archaeometallurgical and historical study. Historical Metallurgy 46: 98–110. Kahanov Y., Stern E., Cvikel D. and Me-Bar Y. 2014 b. Between Shoal and Wall: the Naval Bombardment of Akko 1840. Mariners Mirror 100: 147–167. Shor N. 1990. History of Akko. Tel Aviv.

Fig. 8: Impression of wooden texture on a lead shot that was fired and srtuck wood (perhaps a ship) (E. Galili) 284

Weapons and Ammunition from the Akko Marina

Table 1: Sample of data on hollow iron projectiles Inv. No.

Diam. (cm)

31/93-006/19 21/94-54/1 31/93-006/22 31/93-006/21 31/93-006/23 31/93-006/25

18.7 12.0 13.0 16.0 16.5 ~11.0

Diam. of Hole (cm) 2.0 2.0 3.0 2.0 2.5 2.0

Weight (kg) 18.0 5.2 6.3 9.5 10.0 2.5

Thickness of Wall (cm) 2.5 2.5 2.0 2.0 2.0 1.5

Comments Intact Intact Intact Intact Intact Missing ca. 20%

Table 2: Sample of data on solid iron shot

Inv. No. 31/93-006/20 31/93-006/27 31/93-006/26 31/93-006/29 31/93-006/24 31/93-006/28 21/94-54/3 21/94-54/2 31/93-006/31 31/93-006/30

Diam. (cm) 13 13 13.5 ~10 .0 11.5 7.5 7.5 7.3 4.7 4.3

Weight (kg) 7.80 8.50 10.20 4.30 3.40 1.63 1.63 1.23 0.37 0.28

285

Comments Intact Intact Intact Parts Intact Intact Intact Intact Intact Missing ca. 20–30%

Chapter 18

Pig Tusks and Animal Bones from the Akko Marina Liora Kolska Horwitz (1), Ehud Galili (2) and Baruch Rosen (3) 1. National Natural History Collections, Faculty of Life Sciences, The Hebrew University, Jerusalem, Israel 2. Israel Antiquities Authority and Zinman Institute of Archaeology, University of Haifa 3. Israel Antiquities Authority Key words: underwater archaeology, marine archaeology, Acre, faunal remains

Introduction The archaeological project conducted during the deepening of the Akko Marina (Galili this volume a) revealed organic finds including 25 animal remains.

The Findings The most common finds were 16 pig teeth: 14 whole tusks (incisors) from lower jaws (five from the right side, nine from the left), one right lower incisor and one right tusk (canine) from an upper jaw (Fig. 1). Based on their size all the teeth were derived from domesticated pigs (Sus scrofa dom.) rather than wild boar (Sus scrofa fer.) (Table 1). The tusks came from the skulls of juvenile males and represented a minimum number of nine animals. A deep horizontal chop mark was observed near the tip on the buccal aspect of a right upper canine (No. A14/321). It was probably incurred during butchery and skull removal. The tooth appears to have been polished, although this may be the result of natural abrasion. Most of the tusks were found in areas A, B and C near the entrance to the marina (Galili et al. 2010: fig. 4). A few remains of other domestic animals were also found on the seabed of the marina. These comprised a lower tooth (pre-molar) of a donkey, Equus asinus (Fig. 2), two right horncores of young domestic goat (Capra hircus) rams with deep cut marks on the base that indicate they had been severed from the skulls (Fig. 3), and a number of fragments of rib shafts of goat or sheep (Ovis aries/Capra hircus).

Fig. 1: Top of photograph shows five right lower tusks and bottom of photograph shows nine left lower tusks. The lower right incisor is marked by the number 1. Left of it is the right upper tusk (L.K. Horwitz) such adverse conditions. The tusks under discussion were found in areas of the marina in which findings dated mainly to the Middle Ages as well as to the Byzantine, Crusader and Ottoman periods. They were recovered in material from Barges 40, 41 and 41, which removed material from Area B in the marina. Crusader-period gold coins were also found (Kool this volume) with this material. The tusks may have reached the marina bed after local pigs were

Discussion Animal bones are quite spongy and fragile and usually survive poorly under conditions of marine erosion unless buried rapidly (e.g., the submerged site of Newe Yam – Horwitz et al. 2006). In contrast, teeth, such as the tusks found at Akko, are hardier and would have had a better chance of surviving under

286

Pig Tusks and Animal Bones from the Akko Marina

Table 1: Outer length measurements of pig (Sus scrofa) canines (in cm) Locus

Right

Locus

Left

C6/619

13.5

A40/78

9.5

C6/622

9.5

B9/233

10.5

C8/1091

11.3

B9/274

11.0

C24/1442

10.5

C6/620

9.0

F43/39/1639

11.5

C6/621

9.0

F48/54/1099

9.7

N

5

6

Mean

11.2

9.7

Range

9.5-13.5

9.0-11.0

SD

1.479

0.908

Combined L&R Incisors (N=11)

Mean: 10.4

Range: 9.0-13.5

SD: 1.341

salt pork does not contain whole bones (usually only butchery leftovers bits of pork and pork fat). Thus, the pig heads, which in any case contain little meat, would have been absent from salted meat barrels or any other form of preserved meat, e.g. smoked hams, sides or bellies. The large number of tusks rules out the possibility that we are dealing with complete animals. If that were the case, we would have expected to find a similar number of tusks from upper as from lower jaws as well as other skeletal elements – like carpals, tarsals and petrous bones that are extremely dense and could survive water erosion. Seemingly the lower jaws, not containing a great deal of edible material (mainly marrow), were separated during processing from the skull, in which there are far more edible parts. Thus, it may be assumed that these tusks came from slaughterhouse offal. At some time during carcass processing, the lower jaws were thrown into the sea as refuse while most of the upper jaws underwent another kind of processing and perhaps did not reach the seabed, at least not as a single whole piece. Preference for juvenile males, as illustrated here, implies slaughter for meat consumption, while females were usually preferentially kept into adulthood and used for reproduction. The breeding and consumption of pigs in Palestine in the Middle Ages have been discussed by Rosen (1995). In that article, Rosen presents historical evidence for the breeding of pigs in Akko during the Crusader period and the import of live pigs and

slaughtered and consumed in the city, or represent food debris from vessels anchored in the bay, their remains having been washed away or thrown overboard as refuse. It is less likely that the tusks originated in produce such as barrels of salted pork brought to the city. Although from medieval to modern times pork has been commonly preserved and exported in this manner especially for seamen, in the Mediterranean salted fish was standard fare for sailors (Thompson and Tadlock Cowan 1995; Braudel 1979). Moreover,

Fig. 2: Pre-molar of a donkey (L.K. Horwitz) 287

Liora Kolska Horwitz, Ehud Galili and Baruch Rosen

Fig. 3: Horncores of goat rams with cut marks close to base (L.K. Horwitz) of pork to the Crusader kingdom in general and to Akko in particular as a clear cultural marker. Archaeozoological findings from sites in the nearby Mount Carmel region (Horwitz 2009) and at Caesarea (Cope 1999) demonstrate that consumption of both wild boar and domestic pig continued throughout the Islamic period, albeit on a smaller scale than in Crusader times. Historical sources from the Crusader period mention a slaughterhouse on the southern boundary of the city of Akko (Gertwagen 1996: 574). The findings in the Akko Marina may therefore reflect the importance of pigs during the Crusader era and underscore both the written and archaeozoological evidence for their processing and consumption in the city. Another possible explanation is that the bones and teeth represent remains of recent, eighteenth or nineteenth century animal carcasses. In the eighteenth century Acre was a predominantly Christian city which transformed itself into a predominantly Muslim city during the nineteenth century; the size of Jewish community was negligible in both centuries (Philipp 2012:24). Thus, it may be assumed that with increasing numbers of Muslims in the city, consumption of pig would have become lees common over time. A photo taken in the late nineteenth century depicts piles of bones on

the southeastern coast of Akko (Fig. 4). These are probably associated with a nearby slaughterhouse which dumped its waste on the coast.

Acknowledgments The authors wish to thank the Israel Antiquities Authority for the institutional support, M. Feinberg Vamosh for the English editing, M. Lyubov Musiyenko, director, the State Museum of the History of Religion, St. Petersburg and M. Anna Kudrik for providing Fig. 4.

References Braudel F. 1979. The Structures of Everyday Life. New York. Cope C.R. 1999. Faunal remains and Butchery Practices from Byzantine and Islamic Contexts (1993–94 seasons). In K.G. Holum, A. Raban and J. Patrich, Caesarea Papers 2, Journal of Roman Archaeology Supplementary Series 35. Pp. 405–417. Portsmouth, Rhode Island.

288

Pig Tusks and Animal Bones from the Akko Marina

Fig. 4: Piles of animal bones on the southeasern coast of Akko (photo: L. Fiorillo 1890, courtesy of the State Museum of the History of Religion, St. Petersburg) Galili E., Rosen B., Zviely D., Silberstein N. and Finkielsztejn J. 2010. The Evolution of the Akko Harbor and its Trade Links Revealed by Recent Underwater and Coastal Archaeological Research. Journal of Island and Coastal Archaeology 5: 191–211. Galili E. this volume a. The Akko Marina Archaeological Project – Introduction. Gertwagen R. 1996. The Crusader Harbor of Acre: Layout and Problems of Maintenance. Autour de la Première Croisade, Actes du Colloque de la Society for the Study of the Crusades and the Latin East, Clermont-Ferrand juin 1995. Pp. 553–582. Paris. Horwitz L.K., Galili E. and Lernau O. 2006. Fauna from the Pottery Neolithic Site of Newe Yam.

Mitekufat Haeven, Journal of the Israel Prehistoric Society 36: 139–171. Horwitz L.K. 2009. Roman through Ottoman Period Fauna from Horvat Shallale, Mount Carmel. In S. Dar (ed.), Shallale – Ancient City of the Carmel. (BAR International Series 1897). Pp. 321–140. Oxford. Kool R. This volume. A Unique Hoard of Thirteenthcentury Gold Florins and Other Coin Finds. Philipp T. 2012. Acre: The Rise and Fall of a Palestinian City, 1730– 1831. New York. Rosen B. 1995. Swine Breeding in Eretz Israel after the Roman Period. Cathedra 78: 25–42. (Hebrew).

Thompson S.J. and Tadlock Cowan J. 1995. Durable Food Production and Consumption in the WorldEconomy. In P. McMichael (ed.). Food and Agrarian Orders in the World-Economy. Greenwood Press. Pp. 35–52. Westport, CT.

289

Chapter 19

Glass Vessels and Raw Glass Chunks from the Akko Marina Ehud Galili (1), Baruch Rosen (2) and Jacob Sharvit (1) 1. Israel Antiquities Authority and Zinman Institute of Archaeology, University of Haifa 2. Israel Antiquities Authority Key words: marine archaeology, underwater archaeology, Acre, bottles, lamps, bowls

Introduction Numerous broken glass artifacts and raw glass chunks were recovered during the deepening of the Akko Marina (Galili this volume a), namely 120 vessel fragments and 12 chunks of raw glass. All the fragments were recovered from the sediment fill and represent the periods of port use. No complete glass vessels were found; some fragments were too small to enable identification. The main indicative vessels fragments (from open to closed) and the raw glass chunks are briefly described and discussed below. The importance of these finds lies primarily in the discovery of the raw glass chunks that may represent the export of local products, well known from written sources. Such production centers adjacent to Akko are yet to be found in excavations and surveys. The glass finds span more than a millennium and they shed light on glass manufacturing and trade. The fragmented glass vessels recovered from the port could represent broken vessels that were damaged during loading or unloading and then dumped overboard. They may also have been aboard ships wrecked in the harbor, in which case they could have been part of the cargo or personal belongings of the crew and/or passengers.

Fig. 1: Deep-ribbed bowl (IAA)

at Hagoshrim in Upper Galilee (Weinberg 1973: 38–39), Jerusalem and Caesarea. Several specimens were found at Nessana (Harden 1962: 76, 78–79) and Ashdod (Barag 1967: 71–73). Cups with solid base (31/93: A-17/277; C-3/1365; C-4/1756; C-6/1744; B-25/322; A-36/1029; A-B-38/ 1022): This type was common in the Galilee during the fourth century CE and was widely distributed in other regions as well. It features a cylindrical body, flaring, rounded rim, flat, solid base, and a wound trail applied below the rim. Four flattened disc bases of these vessels were found. The heights of the vessels are unknown, since only fragments were found. Fragments of this type were unearthed in large quantities in the factory dump at Jalame, about 20km south of Akko, where they were probably produced (Weinberg and Goldstein 1988: 60−62, figs. 4–23, and see discussion and references therein; GorinRosen 2009: 88, 89). Barag (1970: 143–144, fig. 32: 6, 7) classified them as types 4.6 and 4.7 and dated them to the fourth century CE on the basis of vessels from Samaria and the Mount of Olives.

Glass Vessels Bowls Pillar-molded formed bowls (deep-ribbed bowl) (31/93: F-66, 7/1998; Fig. 1): Two fragments from the upper part of a ‘pillar-molded bowls’ were recovered. The rim is rounded with almost vertical sides, decorated on the outside with friezes of high ribs. The vessel color is transparent light blue, pale aquamarine. Bowls of this type were widely distributed during the Early Roman period in the eastern and western parts of the Mediterranean (Jackson-Tal 2004: 22–23). Possible manufacturing sites of such vessels were identified

290

Glass Vessels and Raw Glass Chunks from the Akko Marina

Stemmed goblets with hollow ring base (31/93: B-72-74/ 127; C-4/1753; C-5/352; C-5/1731; C8/1893; C-6/1772; C- 5/1741; C-5/1733; C- 6/1775; B-9\323; C-8/1403; C-8/1399; C- 24/1714; A- 13/230; A-36/1019; C-5/1740; A-11/578; Fig. 2): Over 23 hollow ring bases of footed wine glasses were found, including 17 significant ones and some fragments. They were made in a single operation without any added elements. This type first occurred in the fourth century. It was very common in Israel and Jordan Valley in Byzantine/Early Islamic contexts (fifth– seventh centuries) (Barag 1970: 146–148, pl. 33/5-1; Wilson and Sa’d 1984: 75: Jennings 1997–1998: figs. 14, 15, 16).

Beakers with tubular ring base (31/93: A-17/ 277; C-3/1365; C-4/1756; C-6/1744; B-25/322; A36/1029; A-B-38/1022): Seven tubular ring bases of beakers were recovered (not illustrated). This category includes open vessels, the height of which is greater than their diameter. The majority are very thin and fragile, the dominant colors are lightgreenish and bluish. Since the body fragments found are extremely small, it is impossible to reconstruct the original beaker shapes from the bases that were found and it is possible that some, or all of them belong to jugs. Beakers of this type were found at numerous sites including Akko (Avshalom-Gorni 1999: 12, fig. 22: 4), Horvat ‘Uza, Lohamei HaGeta’ot, Akhziv, Giv‘at Yasaf, Kh. esh-Shubeika, Ḥammat Gader, Qastra and Caesarea (Gorin-Rosen 2009: 88–89, pl. 2.53: 1–9). They are dated to the 4th century CE.

Stemmed goblets with solid bases 31/93: B-9/327; C-5/1736; C-6/1776; C-3/1392; C-5/1735; B-9/326; C-6/1742; C-6/1769; C-8/1402; C-4/1717; Fig. 3): Ten flat bases were recovered, some with sections of the stem attached. These goblets have a solid stem that, in most cases, was formed by extending the bottom of the inflated bulb. In nearly all cases this was tooled to form a rounded knob. A disc of glass was then added to form the base. The goblets are usually carelessly made. The stem length varied, while the most common form had a short, slightly bulbous stem (Barag 1970: 147; Whitehouse 1997: 104–105 no. 155, dated to the fifth–sixth centuries; Jennings 1997–1998: 116, fig. 16 ).

Wine Glasses/Goblets All fragmented vessels having a base and a stem were placed in this category. No complete profiles have been found thus in most cases they were identified by their distinctive bases. Most vessels were made of bright blue/green glass, a minority were made in green glass. Goblets were used primarily for drinking. Barag (1970: 146) terms these “wine goblets” and presents many examples from assemblages in Israel. Such vessels were commonly used during the Late Roman–Byzantine periods and continue to be used during the Omayyad period.

Fig. 2: Wine glasses with hollow ring bases (IAA) 291

Ehud Galili, Baruch Rosen and Jacob Sharvit

Fig. 3: Stemmed goblets with solid bases (IAA) Oil Lamps cylindrical or beaded, tend to be substantial. Similar lamps were found at the Beirut Souks site (Jennings 1997–1998: 138, fig. 20: 3, 4). Such lamps are typical chiefly of the late Byzantine and Umayyad periods (Katsnelson 1999: 79, fig. 5: 5).

Lamps appear to have been more numerous than most other forms of early Byzantine glass. While the lower part of many lamps is preserved, no complete upper section of this vessel was recovered in the Akko Marina. Two main groups are defined by two different methods of suspension: The first type sat in a metal loop attached to an arm and the second type had three handles to hold chains and would have been free hanging. Lamps with a cup-shaped upper portion and stem or thorn-like extension at the bottom were popular throughout Byzantine, Islamic and Western medieval periods (von Saldern 1980: 49, Type 3). Bowl-shaped lamp with three handles 31/93: 8/1394; C-6/1774; A-30/403; Fig. 4): Three fragments of such vessels were recovered in the Akko Marina. Hanging lamps are usually identified by their rims and handles rather than by their bases. The most common example has three small, simple loop handles attached at the top to a rim with an external fold and at the bottom to a slightly flaring or near vertical wall. Bowl-shaped lamps were found at Hammat Gader (Cohen 1997: pl. II, 13, 14) and at Shavei Zion (Barag 1967: fig. 16, no. 21). Bowl-shaped lamps with solid or hollow stem 31/93: C-16/1765; A-34/957; A-36/199; C-6/1770; A-36/406; A-36/404; C-35/14; C-24/1713; Figs. 5, 6): Seven stemmed lamps (Fig. 5) and one hollow conical lamp (Fig. 6) were found. The stem bases,

Fig. 4: Bowl-shaped lamp with three handles (IAA) 292

Glass Vessels and Raw Glass Chunks from the Akko Marina

5

6

Fig. 7: Square jug with strap handles (IAA)

Jugs and Bottles

ribs respectively. They are made of clear greenish and clear and light bluish green glass respectively. These jugs were introduced during the middle first century and remained in extensive use throughout the second century CE. They were used for storage or transport of liquids in Egypt and the Roman Empire (Harden 1936: 238; Barag 1985: 99, Meyer 1992: 31). They are characterized by an in-folded and flattened rim and are generally dated to the late first and second centuries CE.

Square jugs with strap handle (31/93: F-39, 43/808; F-39, 43/806, Fig. 7): A top of square jug (flagon) and a top of square juglet were recovered. The handles are decorated with four and six vertically combed

Bottles (31/93: C-5/1732; A-36/1027; B-9/330; C-6/ 1747; B-9/329; C-5/1727; C-6/1749; C-5/1737; C-8/ 1758; F-44, 48/1283): The bottles category consists of small unguent or cosmetic bottles with a closed

Figs. 5, 6: Bowl-shaped lamps with solid or hollow stem (IAA)

293

Ehud Galili, Baruch Rosen and Jacob Sharvit

body and neck and storage vessels. Ten fragments of bottles were recovered. They consist of pinched-neck bottles, bottles with funnel-shaped mouths, smooth bottles and bubble-neck bottles. Most of the fragments found are of rims and necks, and the minority consists of bases. The recovered bottles represent several periods. A pinched-neck bottle decorated with threads (Fig. 8) has a constriction at the base of the neck and probably a globular body. According to Barag (1970: 125) this type may be dated from the end of the Byzantine period to the beginning of the Umayyad period (sixth–seventh centuries CE). A few examples were also found in Lebanon at the Beirut Souks site (Jennings 1997–1998: figs. 23, 24) and at Ashqelon (Katsnelson 1999: 74–75 fig. 4/4). The bottles with the open bulge in the middle of the neck (Fig. 9) and those with a long cylindrical neck and a thickened bulging rim (Fig. 10) have parallels in the Serçe Limani shipwreck dated to the first quarter of the eleventh century CE. Thus these bottles may be dated mainly to the tenth or eleventh centuries CE. The cargo of the Serçe Limani shipwreck originated somewhere on the Eastern Mediterranean coast, probably Syria (Bass 1984: 64–69; Bass et al. 2009: 4).

Fig. 9: Bottle with the open bulge in the middle of the neck (IAA)

Fig. 10: Bottles with thickened bulging rim (IAA) Other Finds Beads and glass bracelets, found among the excavated material, belong to the Early Islamic and Crusader periods. Item No. 15/91-89/3 (not illustrated) is a fragment of spirally twisted glass bracelet (40mm long and 8mm in diameter) made of bluish glass. Similar bracelets were found at Yoqne‘am (Lester 1996: 215, fig. 18: 5) dated to the Early Islamic period, eighth–ninth centuries.

Fig. 8: Bottle with a cylindrical neck with constriction at the base of the neck (IAA)

294

Glass Vessels and Raw Glass Chunks from the Akko Marina

Raw Glass Chunks Twelve raw glass chunks of several colors and hues (light bluish, light greenish, opaque green, greenish to yellow-brown) were recovered (Fig. 11). They were broken from larger blocks; some are irregularly shaped flakes. Their weight ranges from 20 to 535gr each and their maximum length varies from 41 to 110mm. The raw glass chunks recovered from the marina are relatively small and are of numerous dissimilar textures and colors. Usually, glass chunks from a single ship wrecked on an open shore or while anchoring at a natural shelter are more uniform in color (Gelili et al. 2015). It may be surmised that such large, uniform chunks of raw glass recovered from shipwrecks came from cargoes carried by glass traders. The small heterogeneous chunks recovered from the bottom of the Akko Marina are therefore rare and might represent parts of glass cargoes that accidentally fell overboard while loading exported local glass. Local glass production in the Akko area is documented in written sources. However these sources mainly derive from the late Hellenistic and Early Roman periods. Strabo (Geographica XVI, 758), Pliny (Nat. Hist. 36.45) and Josephus Flavius (War 2, 10, 2) mention the quartz sand at the mouth of the Belus River near Ptolemais (Akko). That sand, uniquely suitable for glass-making, was used locally at Akko (and see references in Gorin-Rosen 2000: 4). Material indications of glass manufacturing (raw materials and glass vessels) were found in numerous salvage

excavations conducted in the city and its surroundings (Gorin-Rosen, pers. comm. 2015). During medieval times Akko sand was a “brand name” and was exported to Venice as such (Jacoby 2001).

Discussion and Conclusions Glass Vessels The time span covered by the glass vessel finds from the Akko Marina ranges from the Early Roman period (late first century BCE) up to the Ottoman period (sixteenth–seventeenth centuries CE). The highest concentration falls within the Late Roman-Byzantine periods and the Abbasid-Fatimid periods. The most common vessels are wine glasses with various types of bases and shapes, followed by bottles of various types, from storage vessels to small unguent or cosmetic bottles and lamps. Raw Glass Chunks Primary glass was made in large blocks that were broken into chunks delivered to workshops in the region or exported by sea for the manufacturing of glass vessels. Strabo, Pliny and Josephus mentioned a Roman primary glass-manufacturing center at the mouth of the Belus River (the Na‘aman River) near Akko (Gorin-Rosen 2000). Judging by a comparative study of glass cargoes along the Israeli coast (Galili et

Fig. 11: Glass chunks (E. Galili) 295

Ehud Galili, Baruch Rosen and Jacob Sharvit

al. 2015) and the distribution of the finds in the Akko Marina, the glass chunks may be dated to the Late Roman, Byzantine or Early Islamic periods. In the past, massive production of glass on the Israeli coast occurred between the Sharon plain and the Akko Bay (Brill 1967; Freestone and Gorin-Rosen 1999; Freestone et al. 2008; Gorin-Rosen 2000). Underwater finds of glass cargoes originating from shipwrecks (Galili et al. 2015) indicate that the maritime trade in raw glass occurred in that same area. The apparent correlation between the manufacturing zone and the trade zone should indicate that production and trade functioned as one system. Raw glass recovered from underwater sites off the Israeli coast most probably represents local loads of such glass shipped, during the Byzantine and Early Islamic period on vessels sailing from local anchorages (Apollonia/Arsuf, Caesarea, Dor/Tantura, Neve Yam, Atlit) and the port of Akko, to centers of glass-artifact production and trade in coastal Europe and present-day Turkey. The raw glass industry was attracted to that region because of the simultaneous occurrence of the major basic ingredients needed for such industry – quartz sand from the Nile and northern Sinai that continuously accumulates in the Haifa Bay (the northernmost end of the Nile littoral cell) and the Akko region, and timber for fuel from the Mediterranean forests east of the middle and northern sections of that coast. (South of the Mediterranean climatic zone, the Irano-Turanian and Saharo-Sindic vegetation could not serve as dependable energy source.) The area where the sources of quartz sand and fire wood fuel overlapped is where raw glass finds on land indicate the location of glass manufacturing centers (Galili et al. 2015: fig. 10). Glass finds in the above-mentioned ancient anchorages and harbor bottoms along the Israeli coast mark glass-exporting centers. Finds of raw glass in shipwrecks off the coast of central and northern Israel indicate glass-exporting routes by sea.

Barag D. 1970. The glass vessels of the Roman and Byzantine periods in Palestine. Ph.D. diss., Hebrew University of Jerusalem. Jerusalem. (Hebrew). Barag D. 1985. Catalogue of Western Asiatic Glass in the British Museum, Vol.1. Jerusalem. Bass G.F. 1984. The Nature of the Serçe Limani Glass. Journal of Glass Studies 26: 64–69. Brill R.H. 1967. A Great Glass Slab from Ancient Galilee. Archaeology 20: 88–95. Cohen E. 1997. Chap. 10: Roman, Byzantine and Umayyad Glass. In I. Hirschfeld (ed.). The Roman Baths of Hammat Gader. Final Report. Pp. 396–431. Jerusalem. Freestone I.C. and Gorin-Rosen Y. 1999. The Great Glass Slab at Bet She‘arim, Israel: An Early Islamic Glassmaking Experiment? Journal of Glass Studies 41: 105–116. Freestone I.C., Jackson-Tal R.E. and Tal O. 2008. Raw Glass and the Production of Glass Vessels at Late Byzantine Apollonia-Arsuf, Israel. Journal of Glass Studies 50: 67–80. Gorin-Rosen Y. 2000. The Ancient Glass Industry in Israel, Summary of the Finds and New discoveries, In M-D. Nenna (ed.). La route du verre, Ateliers

primaires et secondaires du second millénaire av. J.C. au Moyen Âge. (Travaux de la Maison de l’Orient

Méditerranéen 33). Pp. 49–63. Lyon.

Gorin-Rosen Y. 2009. The Glass Vessels from Strata 9–6. In N. Getzov, D. Avshalom-Gorni, Y. Gorin-Rosen, E.J. Stern, D. Syon and A. Tatcher (eds.). Horbat ‘Uza: The 1991 Excavations, Vol. II: The Late Periods, Chapter 2: Strata 9–6: The Late Roman and Byzantine Periods. (IAA Reports 42). Pp. 78–98. Jerusalem.

Acknowledgments The authors wish to thank the Israel Antiquities Authority for the institutional support, M. Feinberg Vamosh for the English editing, D. Jacoby, R. Pollak and Y. Gorin-Rosen for their useful remarks on the manuscript and T. Sagiv for the photos.

Galili E., Gorin-Rosen Y. and Rosen B. 2015. Glass Cargoes along the Israeli Coast. Hadashot Arkheologiyot, Excavations and Surveys in Israel 127. Galili E. This volume a. The Akko Marina Archaeological Project – Introduction.

References

Harden D.B. 1936. Roman Glass from Karanis Found by the University of Michigan Archaeological Expedition in Egypt, 1924–1929. (University of

Avshalom-Gorni D. 1999. ‘Akko, the Courthouse Parking Lot. Excavations and Surveys in Israel 19: 12.

Michigan Studies, Humanistic Series 41). Ann Arbor.

Barag D. 1967. The Glass. In M.W. Prausnitz (ed). Excavations at Shavei Zion. Pp. 65–70. Rome.

296

Glass Vessels and Raw Glass Chunks from the Akko Marina

Harden D.B. 1962. Glass. In H.D. Colt (ed.). Excavations at Nessana 1. Pp. 76–91. London. Jackson-Tal R.E. 2004. The Late Hellenistic Glass Industry in Syro-Palestine: a Reappraisal. Journal of Glass Studies 2004: 11–32. Jacoby D. 2001. The Supply of War Materials to Egypt in the Crusader Period. Jerusalem Studies in Arabic and Islam 25: 102–132. Jennings S. 1997–1998. The Roman and Early Byzantine Glass from the Souks Excavations: an Interim Statement. Berytus – Archaeological Studies 43: 111–146. Beirut. Katsnelson N. 1999. Glass Vessels from the Painted Tombs at Migdal Ashqelon. ‘Atiqot 37: 68–82. Lester A. 1996. The glass from Yoqne‘am: The Early Islamic, Crusader and Mamluk periods. In A. BenTor, M. Avissar, and Y. Portugali (eds.). Yoqne’am I: The Late Periods. (Qedem Reports 3). Pp. 202–217. Jerusalem. Meyer C. 1992. Glass from Quseir Al-Qadim and the Indian Ocean Trade. (University of Chicago Oriental

Institute Studies in Ancient Oriental Civilizations 53). Chicago. Weinberg G.D. 1973. Notes on Glass from Upper Galilee. Journal of Glass Studies 15: 35–51.

Von Saldern A. 1980. Ancient and Byzantine Glass from Sardis. Cambridge, MA-London. Weinberg G.D. and Goldstein S.M. 1988. The Glass Vessels. In G.D. Weinberg (ed.). Excavations at

Jalame – Site of a Glass Factory in Late Roman Palestine. Pp. 38–102. Columbia, MO. Whitehouse D. 1997. Roman Glass in the Corning Museum of Glass. Corning, New York. Wilson J. and Sa’d, M. 1984. The Domestic Material Culture of Busra’ from the Nabataean to the Umayyad Period. Berytus – Archaeological Studies. Beirut. 32: 75.

297

Chapter 20

Structures and Installations In and Around the Akko Marina Ehud Galili Israel Antiquities Authority and Zinman Institute of Archaeology, University of Haifa Key words: Acre, underwater archaeology, marine archaeology, harbor, lighthouse, breakwater, Hellenistic, Roman, Byzantine, Crusader, Ottoman

Introduction

installations documented in the marina and its surroundings, some of which are mentioned and discussed briefly in other chapters in this volume. It should be noted that structures and installations built at sea east of longitude 35º 04’ 3.36” (the eastern margins of the Akko ridge) were constructed mostly on unconsolidated sediments. These include the southern breakwater, the eastern rampart, the artificial Island of Flies and the southeastern city seawalls. Rock-cut installations and quarries discovered on the western and northern shoreline of Akko have been described and discussed elsewhere (Galili and Sharvit 2002; Sharvit and Galili 2002, Antonioli et al. in press).

The marine archaeological remains discovered over the years during surveys and excavations in and around the Akko Marina (Galili 1987; Galili et al. 1991, 2002, 2003, 2008, 2010), can be divided into several categories: 1. Built coastal and harbor installations; 2. Remains of sailing vessels having wooden hulls; 3. Assemblages of artifacts originating in shipwrecks; 4. Isolated artifacts discovered in the marina area, with no clear context and originating in shipwrecks or human activity such as fishing, ship maintenance, battles and refuse dumping (Galili et al. 2002; Galili and Rosen 2008; Galili this volume a). Below is a brief description of structures and

Fig. 1: Location map (E. Galili) 298

Structures and Installations In and Around the Akko Marina

The Southern Breakwater

at preventing the silting of the harbor by sediments (Raban 1982a). At a certain point in time these openings were blocked by a construction (Fig. 5). Round holes were cut into some stones, and granite columns in secondary use were inserted in them vertically, to be used for mooring watercraft (Fig. 6). In a mariner’s guide (Lo Compasso da navigare) from the thirteenth century CE, a reef is mentioned. According to Jacoby (1979: 8) the reef is actually the remains of the ancient breakwater, which was not usable at that time. The mariner’s guide may refer to a submerged reef located ca. 200m southwest of the harbor, which appears in nineteenth-century Admiralty charts (Rosen et al. 2011: 174–175 Galili et al. this volume a, figs. 26-29). A photograph taken by Gertrude Bell in June 1900 depicts large structure on the eastern end of the breakwater (Fig. 7) (The Gertrude Bell Archive, Album A-515). Two such structures are also depicted in nineteenth-century Admiralty charts (Galili et al. this volume a figs. 23, 24, 26). In 1957 a proposal to develop the Akko Marina was submitted to the Akko municipality (Fried, Glaswein, Werner Engineers 1957) in which the ancient southern breakwater was mentioned (Fig. 8). In aerial photographs taken before the modern harbor’s construction and in photographs from the 1941 British Survey Archive (Fig. 9), the breakwater appears as a row of reefs. Some of these were later used as a foundation for the

The southern breakwater is a manmade structure built in the sea, probably with no natural rock core (Flinder et al. 1993). The foundations of the breakwater (Fig. 1: 13) start at the southeastern end of the rocky peninsula on which the Old City stands (the Akko headland) (for a detailed description of the southern breakwater see Flinder et al. 1993; Raban 1982a). From there the breakwater extends ca. 260m eastward and bends northward at its eastern end. Finds from underwater excavations during 1965 and 1974 revealed that the bottom courses of the breakwater were laid atop a foundation of pebbles bonded by cement. To improve adherence of the lowest course to the seabed, the lower face of the stones was not dressed. Long, narrow ashlars were laid as headers on the foundation top. The southern face, facing the open sea, was built of four courses of large stones while the inner face was constructed of three courses of smaller stones (Fig. 2). The fill between the walls was made of crisscrossed ashlars. This structure formed an upright wall 12m wide, protruding from the coast and forming an area of protected water to its north. During the Roman period, huge kurkar ashlars (2 ×1.2 × 12m) were laid atop the earlier breakwater (Raban 1982b). Placing these stones on their long axis perpendicular to the breakwater, at a distance of 1m apart, created flushing channels (Figs. 3, 4) aimed

Fig. 2: The southern breakwater. See insert detail in Fig. 3 below (modified after Raban 1982a) 299

Ehud Galili

Fig. 3: The Southern breakwater: Insert detail of Fig. 2, large ashlars with spaces between them, seen from above and from the north (modified after Raban 2008a)

Fig. 6: The Southern breakwater: ashlar in which a granite column was embedded, used as a post to moor a vessel or as a foundation for a superstructure, seen from the north (E. Galili)

Fig. 7: Structures on the southern breakwater (1922) (photo courtesy of Gertrude Bell Archive, Newcastle University) Fig. 4: The Southern breakwater, large ashlars with spaces between them, seen from the northwest (E. Galili)

Fig. 8: Map of the Akko area before the building ofthe southern breakwater (modified after Fried, Glaswein, Werner Engineers 1957)

Fig. 5: Structure closing the spaces between the large ashlars on the southern breakwater (E. Galili) 300

Structures and Installations In and Around the Akko Marina

Fig. 9: Remains of Crusader vaulted structures in the sea front of the Pisan Quarter (1941 British Survey Archive, IAA, negative nos. N 23-205, N 25-206) modern breakwater (Galili and Rosen this volume e). Scholars have suggested various periods for the construction of the southern breakwater. According to Jacoby (1979: 8) the breakwater was built in the early Hellenistic period; Flinder et al. (1993) suggested that it was built during the Early Roman period. According to Raban (1983: 141) the harbor was built as early as the sixth century BCE. Kesten (1993: 20), however, dated the breakwater’s construction to the Early Islamic period. Judging by the finds revealed by the Akko Marina Project, it is suggested that the southern breakwater was built during the Hellenistic/ Early Roman Period (Zilberstein et al. this volume; Galili and Rosen this volume e). Broken iron anchors recovered in the western basin suggest that during the Byzantine period the southern breakwater was not functioning and was partially damaged (Galili and Rosen this volume b). It is generally accepted that the harbor was destroyed by the Mamluks in 1291. Nevertheless, findings reported in this volume (Galili and Rosen this volume a) suggest that maritime activity in the harbor continued in the fifteenth century.

The Eastern Rampart/Eastern Breakwater (?) A submerged broad and elongated structure, a “rampart,” extends north–south, from near the shoreline (Hasusim Beach) to ca. 15m north of the small artificial island known as the Island of Flies (or Tower of Flies) (Fig. 1: 15 and see a detailed description below). The rampart is ca. 300m long with an average width of ca. 35m. Its highest points in shallow water (0.5–1.0m deep) block the passage of water craft with a low draft. A gap of about 15m, where the water is 2–3m deep, exists between the southern end of the rampart and the island. That gap allows passage of vessels with a 1.0–1.5m draft. Written historical sources dated to the Early Islamic Period describe in detail the building of a breakwater by Ahmed Ibn Tulun (868–884 CE). The architect who planned and built it was the Jerusalemite Abu Bakker. The construction was described in a book by the

Muslim geographer Al-Muqaddasi (1906: 153), the grandson of the architect. Below is a translation by Le Strange (1890: 328, 329): “So Ibn Tulun wrote to his Lieutenant in Jerusalem commanding that he should dispatch my grandfather to him; and on his arrival they laid the affair before him. `The matter is easy` said my grandfather `let them bring such sycamore beams as be large and strong`. These beams he set to float on the surface of the water, a prolongation of the town walls (seawards), and he bound them one to the other; while toward the west he left the opening for a mighty gateway. And upon these beams he raised a structure with stones and cement [lime]. After every five courses he strengthened the same by setting in great columns. At length the beams become so weighted that they began to sink down; but this was little by little, and finally they rested on the sand. Then they ceased building for a whole year that the construction might consolidate itself, after which, returning, they began again to build. And from where it had been left off, continuing, my grandfather made junction between this and the ancient city walls, bringing the new work right up into the old, and causing the two to join together. Across the western gate of the port he built a bridge, and every night when the ships had come within the harbour they drew across the water-gate a chain, even as was the case at Tyre. It is reported that my father received for his matter the sum 1,000 Dinars besides robes of honour, horses, and other gifts, and his name was inscribed over the work. Now, before this harbour had been made the enemy were wont to take advantage of the ships lying here to do them grievous damage.” It has been suggested that this source actually describes the construction of the eastern, now submerged, eastern rampart (Gertwagen 1996). However, the location of the structure Muqaddasi describes is uncertain, and scholars disagree over its function. It is also unclear whether the text describes construction of the eastern rampart or of another rampart located at the entrance to the marina, at the eastern end of the southern breakwater. The location of the Early Islamic harbor gate, which was closed by a chain, is also unclear. The text that describes the building of the harbor states that the gate was in the west; however, this direction does not conform to conditions in that area (Galili and Rosen 2008; Galili et al. 2010; Galili and Rosen this volume e). Test excavations at the rampart revealed that it is built of undressed stones and gravel mixed with some ashlars that were dumped randomly. No traces of wooden caissons were recovered, yet the the finds date no later than the tenth century CE, suggesting that the rampart was built in the Early Islamic period (Yurman 2013: 60–65). It was suggested that the rampart was the breakwater described by Muqaddasi, that aimed at protecting the harbor from the moderate eastern winds (Gertwagen

301

Ehud Galili

laid as headers on the sea bottom from 7m depth up to ca. 0.7m above the present sea level (Raban 1982a). The early-phase remains are visible mainly in the western, flooded half of the island. The earliest phase was dated to the Persian period based on stone types, the header construction method used and similarities with Phoenician ports. It was proposed that the island, being outside the harbor, was used as a dock unconnected to the shoreline (Raban 1982a; Linder and Raban 1965). During the Hellenistic period the structure on the island was strengthened and heightened. Remains of this construction can be seen at sea level on the eastern part of the island. Possibly at this stage a lighthouse was installed on the island by Ptolemaic rulers (Raban 1982b). During the Crusader period a tower was built on the island; its remains are still visible. The Crusader construction is fortified by iron clamps (Raban 1983). Given the finds recovered in the current Akko Marina project, it is proposed that the island was built during the Hellenistic/ Early Roman periods, when the southern breakwater was constructed (Galili and Rosen this volume e).

1996). This assumption is dubious, because no waves generally come from the east due to the short fetch (see discussion in Galili and Rosen this volume e).

The Tower of Flies The Tower of Flies or Al Manara (“lighthouse” in Arabic) designates a structure on a small artificial island located ca. 150m east of the southern breakwater’s end (Fig. 1: 14). The structure’s name – Tour des Mouches (“Tower of Flies”) dates back to the Crusader period and may have come from the identification of Akko with Philistine Ekron, whose god was Baal Zebub (Prawer 1962: 54). On the island stands a polygonal structure, protruding 6.5m above the water’s surface (Fig. 10). In an excavation of the island by the Underwater Exploration Society of Israel in the early 1970s a number of building phases were identified (for detailed description see Flinder et al. 1993). The island was built as a rectangular structure, 60m long and 12m wide; its western part is currently underwater. The foundation platform of the artificial island was built of huge ashlars (4 × 1 ×1m)

Fig. 10: Isometric view of the Island of Flies (the coastline is marked with a thick line) and the submerged courses (modified after Raban 1993) 302

Structures and Installations In and Around the Akko Marina

The Crusader Harbor Chain During the Early Islamic Period the harbor was closed by a chain at night (see above). Historical descriptions dated to the Crusader period reported that the Akko Harbor could be closed by a chain, installed by the Genoese in 1258 (Jacoby 1979: 10; Kedar 2012). Remains of a structure, described by some archaeologists as a square tower, were identified near the southeastern end of the southern breakwater, submerged in about 1m of water. An underwater archeological survey conducted by R. Gertwagen found that the stones of the tower were attached to one another by iron clamps. Similar clamps appear, as noted above, on the Crusader tower built on the Island of Flies and on the round Crusader tower located in the northeastern corner of the marina. It was proposed that the square tower protected the entrance to the harbor and the iron chain was extended from it to the Tower of Flies (Gertwagen 1996). Others posited that the chain was installed at the opening of the western basin and extended from the southern breakwater northward to the round tower on the shoreline (Raban 1982a; Flinder et al. 1993; Riley-Smith 1990; Jacoby 1979; Kedar 2012) (for detailed discussion see also Galili and Rosen this volume e).

Fig. 11: A plan depicting the remains of the Hellenistic/Roman breakwater in the area of the Pisan Quarter sea front and a portion of the sea wall built over it (E. Galili)

Group of Large Ashlars Recovered from the Western Basin

Fig. 12: Details of the seawall and the arch covering the gap between two blocks of the ancient breakwater: right – looking south, left – cross-section looking west (E. Galili)

A group of 15 large kurkar ashlars was found while building a pier at the northeastern corner of the marina (Fig. 1: 23). The stones measured 30–66 × 50–60× 120cm. It is reasonable to assume that these stones were part of a Hellenistic/Roman pier or other ancient marine installation, which was destroyed during the construction of the modern marina.

A Sea Wall with a Submerged Arch South of the Pisan Quarter A high wall segment erected on a submerged foundation is situated offshore at the sea front of the Pisan Quarter (Figs. 1: 18, 9, 11 marked with an arrow). The wall partly covers the foundations of the ancient southern breakwater (Figs. 9, 11). A built segmental arch, is located at the lower part of the northern face of that wall (Fig. 12). The arch consists of one layer of stones on the face of the wall and behind it, to the south, the wall is constructed of large ashlars (ca. 90cm long) (Fig. 12). On the upper part of the wall the remains of a staircase can be seen, leading to a niche facing southward, toward the open sea and the Haifa Bay. The south-facing side of the wall is built of large, carefully dressed ashlars, like the building stones in the outer face of the sea wall to

Fig. 13: Portion of sea wall, view from the south west (2004). Note that there is no built arch on the southern face of the wall (E. Galili) the west (Fig. 13). No continuation of the arch is seen on this side of the wall. The northern face of the wall, toward the Pisan Quarter, was built of small ashlars. The fill between the two faces was of fieldstones bonded by mortar. Iron clamps connected the ashlars diagonally on the side of the wall facing the sea. The clamps consisted of U-shaped iron rods, square in cross-section (20 × 20mm). They were fixed in the

303

Ehud Galili

long), irregular, stone blocks, remaining from the ancient breakwater (Figs. 11, 12). It seems that the arch was originally built partly underwater to bridge the space between two blocks of the ancient breakwater, so that the wall could be built over it (Galili and Rosen this volume e). It is unlikely that the arch represents an opening in the sea wall, as such an opening could have endangered the city (Kesten 1993: 51). Thus, the arch was not an entrance and cannot be used as a reliable indicator for tectonic subsidence or a marker for low sea level during the Crusader period (see detailed discussion in Galili and Rosen this volume f). This wall may have been part of the Pisan Quarter public buildings and their walls facing the sea served as sea walls. While building the sea walls the Ottomans most probably used parts of the previous constructions (Roman foundations and Crusader walls) and probably recycled Crusader ashlars.

stones by cast lead. A late nineteenth-century photograph shows that the wall was connected to the shore as part of a large building, whose southern wing formed the sea wall (Fig. 14). The 1941 British survey photograph (Fig. 9) shows that the eastern wing of the structure has vanished, sometime during the decades after the previous photograph. Based on the construction style and the iron clamps, the building was dated to the Crusader period (Gertwagen 1989). Kesten (1992: 51, 52) suggested that the structure dated to the Crusader period and was part of a series of a Crusader colonnade built on the Pisan Quarter sea front. It was reinforced and covered by fieldstones during the Ottoman era. The partly submerged arch was identified by several scholars as an entrance in the wall, suggesting that the area was subject to tectonic subsidence (Neev et al. 1991; Flemming et al. 1978), or that during the ninth–thirteenth centuries CE, the sea level was lower by ca. 2m relative to the present level (Gertwagen 1989). In underwater surveys at the site in 2004, the foundations of this wall were examined and the remains of the ancient breakwater were mapped. The survey revealed that the segment of the abovementioned sea wall was built over large (up to 4m

Foundation of the Southern Breakwater in the Pisan Quarter An array of boulders ca. 90m long and 10m wide tending generally northwest to southeast on a 115o angle, was discovered and mapped during

c b a

Fig. 14: Akko walls in the Pisan quarter sea front, (a) the wall segment seen in Fig. 13 above, (b) the ruins of the ancient breakwater, (c) the Island of flies, looking southeast 1870–1885 (Photo: C. Fiorillo, courtesy of State Museum of the History of Religion, St. Petersburg) 304

Structures and Installations In and Around the Akko Marina

an underwater survey south of the Pisan Quarter (Figs. 1: 18, 9, 11, 14). Forty-five amorphous blocks were documented, of which 21 measured 3–5m and 24 measured 1–3m. The blocks were arrayed in a long, narrow line extending from the base of the modern breakwater westward. Their location and layout suggest that they were part of the connection of the ancient breakwater to the shore. The Crusader/ Ottoman sea wall was built atop these boulders, the remains of the ancient breakwater. Probes by means of water-jet drilling showed that the seabed under most of the boulders is of pebbles and loose sediment 2–3m deep. In the western part of the area examined, the seabed is rocky, part of the kurkar ridge on which the western part of the Old City is built (See Galili and Rosen this volume e). It seems that the ancient breakwater was built mainly on the seabed consisting of loose sediment that had accumulated east of the kurkar ridge. Such massive construction, founded on loose sediment and exposed to energetic waves, may have caused the sinking and destruction of the southern breakwater. That could have led to the collapse of the Ottoman sea wall built on it. Building on unconsolidated sediment also brought about the destruction of the western part of Caesarea’s Roman harbor (Galili and Sharvit 1998).

In contrast, considerable portions of the Akko sea wall built on massive kurkar rock survived the forces of nature (See also Galili and Rosen this volume e).

The Partly Flooded Templar Quarter at Southwest Akko This area is a partly flooded, flat, rocky shelf, (Figs. 15, 16) bordered by Burj el-Qashashaf in the north, and the southwestern city walls and tower, above which the lighthouse is currently positioned in the south. On the east, the area is bordered by the twentieth-century beach promenade. On the west, it is marked by the abrasion platform ending with a 4–5m-high, underwater drop and west of it a flat, rocky sea bottom. About 70% of the area delineated above is a flat kurkar shelf. It is a rock-cut kurkar platform (maximum depth below sea level 0.8m) that was formed by quarrying building stones from the kurkar bedrock down to the lowest possible elevation in low tidal conditions. Remains of a large quarry, rock-cuttings, stone-built sea walls and installations were discovered on that hewn kurkar shelf.

Fig. 15: The rock-cut shelf in the Templar Quarter sea front, looking east: Aerial photo (modified after A. Ohayon) 305

Ehud Galili

Fig. 16: Plan of the archaeological remains on the flat rocky shelf (IAA) protection from the sea and enable quarrying and building activities on the rocky shelf to the east. Kurkar quarries with natural elevated fringes have been referred to in the past as “high fringe quarries.” They are known from various sites on the Levantine shores such as Dor, Rosh Ha-Niqra, Tyre, Sidon and Arwad. Originally the kurkar protrusions on the fringes of the rocky shelf in Akko must have been more continuous, to protect the quarry. The protective kurkar wall was mostly quarried away at a later stage due to the high demand for building stones. The quarry yielded no datable finds; however, some Crusader walls cover the quarry features (see below). Thus it can reasonably be assumed that the quarrying activities there began during, or prior, to the Crusader period.

Fig. 17: Large protrusions of kurkar bedrock left after quarrying, looking southwest (E. Galili) Abrasion Platform and a “High Fringe quarry”:The western fringes of the rocky shelf consist of an elongated, continuous, natural abrasion platform, up to 11m wide. A line of protrusions of kurkar rock, up to 2.3m high, appear in several places on the eastern fringes of this abrasion platform (Figs. 15, 16, 17). These are the remains of the kurkar ridge that were left behind unquarried. The abrasion platform and the rocky protrusions were deliberately preserved, creating an elevated rocky belt on the western, seaward, fringes of the rock platform (Fig. 18). The purpose of this escarpment was to provide

The foundation courses of Crusader walls on the fringes of the rocky platform: Built of kurkar masonry stones (Figs. 19, 20, 23), these foundation courses were observed in several locations on the southwestern and northwestern fringes of the bedrock shelf. Pairs of adjacent depressions and pieces of oxidized iron, the remains of binding clamps to reinforce the foundations, are seen on top of these stones (Figs. 19, 20, 21). Square iron bars (20x20

306

Structures and Installations In and Around the Akko Marina

a

a

b

b

Fig. 18a, 18b: Abrasion platform left as an elongated escarpment on the western fringes of the rocky shelf. The escarpment was left unquarried to protect the quarry and the Templar Quarter sea front: (a) looking southeast (b) looking northwest. Note the rock cuttings east of the escarpment (E. Galili) mm) are perpendicularly stuck into the kurkar, aiming at connecting the ashlars to the bedrock (Fig. 22). Remains of similar walls, common during the Crusader period, were discovered in other places on the bedrock shelf. These wall fragments may thus be dated to the Crusader period and may be the lowest courses of the western Crusader sea wall. A rock-cut line on the west fringes of the rock-cut platform: On the eastern side of the abrasion platform an elongated rock cutting (0.7m deep) form a zigzag pattern along the fringes of the rocky shelf (Fig. 18). In some places, it is cut down to 0.7m below sea level. It has been suggested that these rock cuttings

Fig. 19a, 19b: The foundation course of the Crusader sea wall, built of kurkar masonry stones with cavities of iron clamps, located on the northwestern fringe of the rocky shelf, looking southwes (E. Galili) may represent the foundation trench of the Crusader or Ottoman sea walls, which were not preserved in this area. However, the remaining Crusader wall segments mentioned above are not correlated with these cuttings (Fig. 23). Nevertheless, in areas where the Ottoman sea wall was preserved (e.g., the southwestern sea wall, Fig. 24) there is no correlation between the rock cuttings and the Ottoman sea wall. At some locations the Ottoman walls cut the rockcuttings and are built unconformably on them. Thus, 307

Ehud Galili

Fig. 23: Rock cuttings on the fringes of the rocky shelf and a Crusader wall segment, lying un-conformably, looking northeast (E. Galili)

Fig. 20: Crusader sea wall fragment on the southwest corner of the rocky shelf, looking southwest (E. Galili)

a

Fig. 21: Oxidize iron clamps on top of the ashlars (E. Galili)

b

a

Fig. 24a, 24b: The southwestern Ottoman sea wall and the rock-cuttings on the shelf. Note the unconformity between them (E. Galili)

b

Fig. 22a, 22b: Iron bar protruding from the rocky shelf (E. Galili)

Fig. 25: Rock-cut elongated feature with a channel, looking east (E. Galili) 308

Structures and Installations In and Around the Akko Marina

and drawings. The palace was described in written documents as the most fortified place in the city. Beginning in 1752 the remains of the palace were dismantled down to its foundations. The stones were used during the reign of Dahr el-`Umar for building the city walls.

Fig. 26: Rock-cut elongated feature with a channel, looking west (E. Galili) it seems that these fringe rock-cuttings are associated with the early quarrying activity rather than the Crusader or the Ottoman sea-wall foundations.

The remains of the nineteenth-century lighthouse: Several large lumps of dressed kurkar stones and field stones (lengths 1.0–5.2m), bonded together with cement (Figs. 16, 17, 27), lay in the middle of the bedrock shelf. In an aerial photograph taken by the Bavarian Squadron prior to 1918, a two-story square building is visible in this location, connected to the shore by a long bridge. Late eighteenth/early twentieth-century photographs and maps suggest that the building on the rock platform was a lighthouse, built at the end of the nineteenth century (Fig. 28).

The remains of a large, elongated rock-cut feature: A long structure (length 50m, width 1.5– 2.5m, height 0.85m; Figs. 15, 16, 25, 26), tending to the southwest is located on the southern part of the bedrock shelf. At its center a longitudinal channel is cut (75–95cm wide, 30–50cm deep). Its western end curves southward, where it is blocked by a natural kurkar protrusion (1.5m high). The feature is cut from the kurkar bedrock and protrudes some 85–100cm above the flat rocky surface. Thus it was probably cut during the operation of the large quarry that covers most of the rocky shelf. At its eastern end, lines of foundation trenches were quarried below the water level, indicating that the feature continued some 10–15m to the east. This rock-cut feature may be associated with the quarry or with the Templar palace, which is depicted in this location on maps

Fig. 27: Lumps of dressed kurkar stones and fieldstones bonded with cement, the remains of a late eighteenth-century lighthouse, looking north (the northern lighthouse) (E. Galili)

Fig. 28: A photo taken during 1870-1889 looking west: depicting the northern lighthouse of Akko (photo: L. Fiorillo, courtesy of the State Museum of the History of Religion, St. Petersburg, P 3421) 309

Ehud Galili

This northern lighthouse stopped functioning and around 1913 a new lighthouse was built at the present location (See also Galili et al. this volume a).

(Sharvit et al. 2013; Sharvit and Planer 2014 and see discussion and interpretations in Galili and Rosen this volume e).

Excavations at the Foot of the Southeastern City Wall, 2008–2013

Acknowledgments The authors wish to thank the Israel Antiquities Authority for the institutional support; M. Feinberg Vamosh for the English editing; D. Jacoby for his useful remarks; the Gertrude Bell Archive, Newcastle University, for providing Fig. 7; M. Lyubov Musiyenko, director, the State Museum of the History of Religion, St. Petersburg and M. Anna Kudrik for providing Figs. 15, 23, the IAA Archive, Jerusalem, for providing Fig. 9, Howard Rosenstein and Fantasea for support associated with photography equipment; and A. Ohayon for the aerial photo Fig. 15.

A section of the Ottoman southeastern sea walls of Akko, ca. 250m long, had incurred severe damage by continuous marine erosion (Fig. 1: 9). From 2003 to 2008 the IAA conducted a project aimed at preventing further damage, strengthening and repairing the foundations of the sea walls and filling voids and missing stones caused by that erosion (Sharvit et al. 2013). A temporary rampart and a road were built along the sea wall and a trench was dug along the walls foundations. Archaeological excavations were carried out simultaneously, exposing the ancient deposits down to the bedrock at the foot of the walls. The excavations by the IAA were carried out with the collaboration of the Old Akko Development Company and Rhodes University. The finds revealed Hellenistic structures including, in their Area E, a wall (8.0 × 1.6m) built of five courses of kurkar ashlars that were founded on the kurkar bedrock. In their Area C a pavement 30m long built of flat kurkar slabs (0.15 × 0.4 × 0.8m) was found, laid on kurkar ashlars constructed as headers (Sharvit et al. 2013: fig. 4). The pavement, interpreted as a wharf by the excavators, was dated to the Hellenistic period by pottery and other small finds. In their Area F a rectangular structure included two parallel walls and a tilted pavement between them. This pavement was interpreted as a dry dock by Sharvit et al. (2013: 43). Above these Hellenistic structures was a layer of debris composed of small stones. The debris was overlaid by a layer of horizontal columns, in secondary use, lying perpendicular to the wall. The columns were made of granite, marble and kurkar (Sharvit et al. 2013: fig. 5). A line (80m long) of Ottoman wooden pilings (dated to 1816–1846) was inserted vertically along the Ottoman sea wall. The pilings (made of Pinus brutia), 3m high and 0.2–0.3m in diameter, were driven at a distance of 0.3m apart (Sharvit et al. 2013: figs. 7, 8). An iron tip in the form of a plowshare was affixed to the lower part of each wooden piling. The tops of the pilings were attached by iron nails to horizontal beams. In some places the bottom course of the Ottoman wall was laid on top of these horizontal beams, or on top of the horizontal stone columns. These finds shed light on the foundation methods of the Ottoman city walls. They also confirm the previously suggested foundation period of the harbor and contribute to the discussion regarding its layout

References Al Maqdisi-Shams Ad din Abu cabd Allah Muhammad Ibn Ahmad Ibn Abu Bakr, 1906. In M.J. de Goeje (ed.). Ahsan at Taqasim fi Mac rifat alAqalim. P. 153. Leiden. Antonioli F, Mourtzas N., Anzidei M., Auriemma R., Galili Y., Kolaiti E., Lo Presti V., Mastronuzzi Scicchitano G., Spampinato C., Vacchi M., Vecchio A. 2017. Millstone quarries along the Mediterranean coasts. Chronology, morphological variability and relationships with past sea-levels Quaternary

International.

Ashtor E. 1986. European trade in Palestine during the late Middle Ages. In B.Z. Kedar (ed.). East-West Trade in the Medieval Mediterranean. London. Flemming N., Raban A. and Goetschel C. 1978. Tectonic and Eustatic Changes on the Mediterranean Coast of Israel in the last 9000 Years. In J.C. Gamble and R.A. Yorke (eds.). Progress in Underwater

Science 3 (New Series) of the Report of the Underwater Association, Proceedings of the 11th Symposium of the Underwater Association, 18–19 March 1977. Pp. 33–93. London.

Flinder A., Linder E. and Hall E.T. 1993. Survey of the Ancient Harbour of Akko, 1964–1966. In M. Heltzer, A. Segal, and D. Kaufman (eds.). Studies in Archaeology and History of Ancient Israel. Pp. 199–225. Haifa. Fried, Glaswein, Werner Engineers 1957. Report on

310

Structures and Installations In and Around the Akko Marina

the Vote to Restore and Develop the Akko Harbor. Report 191–2, prepared for the Akko Municipality. 12 pages and maps. Haifa. (Hebrew).

Gertwagen R. 1989. The Southern Shore of the Old City of Akko – ‘The Pisan Harbor.’ C.M.S. News, Rep. No. 16. Haifa.

Galili E. 1987. The Underwater Survey–1986, Hadashot Arkheologiyot 89: 59–60. (Hebrew).

Jacoby D. 1979. Crusader Acre in the Thirteenth Century: Urban Layout and Topography. (Estratto dagli Studi Medievali, 3a Serie, XX, I). Spoleto.

Galili E., Dahari U., Sharvit J., Aviam M. and Stern E. 1991. Akko Marina Survey. Final Report. Submitted to the Akko Development Corporation by the Israel Antiquities Authority. Pp. 1–54. (Hebrew).

Kedar B. Z. 1991. Looking twice at the Land of Israel, Aerial Photographs of 1917–1918 and 1987–91.

Galili E. and Sharvit J. 1998. Ancient Coastal Installations and the Tectonic Stability of the Israeli Coast in Historical Times. In I.S. Stewart and C. VitaFinzi (eds.). Coastal Tectonics. (Geological Society Special Publications 146). Pp. 147–163. London. Galili E., Sharvit J., Dahari U., Bahat-Zilberstein N., Finkielsztejn G., Stern E., Kool R., Kahanov Y. and Rosen B. 2002. Akko Harbor, Underwater Survey. Hadashot Arkheologiyot 114.

Kedar B.Z. 2012. Prolegomena to the world history of harbor and river chains. In R. Gertwagen (ed.). Shipping, Trade and Crusade in the Medieval Mediterranean. Pp. 3–38. Burlington, U.K. 1993. The Old City of Akko, Reconsideration, Criticisms. Akko. (Hebrew).

Kesten A.

Le Strange G. 1890. Palestine under the Moslems. Pp. 328, 329. London.

Galili E and Sharvit J. 2002. Millstone Quarry on the Coast of Akko. ‘Atiqot 42: 73–78 (Hebrew).

Linder E. and Raban A. 1965. Underwater Survey in the Akko Harbor (1994), Western Galilee and the Galilee Coast, 19th Annual Israel Exploration Society Conference. Pp. 180–193. Jerusalem. (Hebrew).

Galili E., Rosen B. and Zviely D. 2003. Akko and its Surroundings on Coins and Graphic Representations. Horizons in Geography 62: 113–130. (Hebrew).

Neev D., Bakler N. and Emery K.O. 1991.

Galili E. and Rosen B. 2008. The Harbor (Acco). In E. Stern (ed.). The New Encyclopedia of Archaeological Excavations in the Holy Land 5. Pp. 1558–1561. Jerusalem. Galili E., Rosen B., Zviely D., Silberstein N., Finkielsztejn J. 2010. The Evolution of the Akko Harbour and its Trade Links Revealed by Recent Underwater and Coastal Archaeological Research. Journal of Island and Coastal Archaeology 5: 191–211. Galili E. This volume a. The Akko Marina Archaeological Project – Introduction. Galili E. and Rosen B. This volume e. Summary. Galili E., Zviely D. and Rosen B. This volume a. The Akko Harbor in Numismatic, Epigraphic, Cartographic and Photographic Evidence. Galili E. and Rosen B. This volume a. FifteenthCentury Harbor Installation with Wooden Pillars in Akko. Gertwagen R. 1996. The Crusader Harbor of Acre: Layout and Problems of Maintenance. In M. Balard (ed.). Autour de la Premiere Croisade, Actes du Colloque de la Society for the Study of the Crusaders and the Latin East, Clermont-Ferrand, 22–25 juin 1995. Pp. 553–582. Paris.

Mediterranean Coasts of Israel and North Sinai. Holocene Tectonism from Geology, Geophysics and Archaeology. New York. Prawer J. 1962. A History of the Latin Kingdom of Jerusalem, Vol. II: The Crusades and the Second Kingdom. P. 54. Jerusalem. Raban A. 1982a. The Harbors of Akko. In M. Yeda‘aya (ed.). Antiquities of Western Galilee. Pp. 180–207 Haifa. (Hebrew). Raban A. 1982b. A Fragment of a Ship from the Napoleonic Siege on Akko (1799). In M. Yeda‘aya (ed.). The Antiquities of Western Galilee. Pp. 198– 207. Haifa. (Hebrew). Raban A. 1983. Maritime Acco. In E. Stern (ed.). The New Encyclopedia of Archaeological Excavations in the Holy Land 1. Pp. 29–31. Jerusalem. Raban A. 1983b. Akko Harbor. Kardom 24–25: 141–146. (Hebrew). Riley-Smith Y. 1990. The Atlas of the Crusaders. London. Rosen B., Galili E. and Zviely D. 2011. The Roman lighthouse in Akko, Israel. International Journal of Nautical Archaeology 41: 174–175. Sharvit J., Planer D. and Bauxton B. 2013.

311

Ehud Galili

Preliminary finds from archaeological excavations along the foot of the southern seawall at Akko, 2008– 2012. Michmanim 24: 39–52. (Hebrew). Sharvit J. and Planer D. 2014. Akko, the Southern Seawall, Preliminary Report. Hadashot Arkheologiyot 126. Yurman A. 2013. The Underwater Rampart at Akko, its Construction and Function, an Archaeological and Geomorphological Underwater Research. M.A. thesis, University of Haifa, Faculty of Humanities, Department of Maritime Civilizations. Zviely D., Galili E. and Rosen B. 2003. The Akko Harbor in Modern Nautical Maps, Horizons in Geography 56: 52–77. (Hebrew).

312

Chapter 21

Shipwrecks and Cargoes in and around the Akko Marina Ehud Galili Israel Antiquities Authority and Zinman Institute of Archaeology, University of Haifa Key words: Acre, underwater archaeology, marine archaeology, harbor, breakwater, Hellenisic, Roman, Byzantine, Crusader

Introduction

Shipwrecks in the Akko Marina (the Western Basin)

Marine archaeological remains discovered during surveys and excavations over the years, in and near the Akko Marina, can be divided into several categories: 1. Built coastal and harbor installations; 2. Shipwrecked vessels having wooden hulls; 3. Assemblages of artifacts originating in shipwrecks; 4. Isolated finds originating in shipwrecks or anthropogenic activity e.g., fishing, ship maintenance, military activities and refuse dumping (Galili et al. 2002a; 2002b, Galili and Rosen 2008; Galili this volume a; Raban 1982a). Below is a brief description of the shipwreck remnants and cargo remains recovered; some are mentioned briefly in other chapters in this volume.

Shipwreck from the Eleventh–Twelfth Centuries CE at the Marina Entrance Remains of an ancient wooden hull were identified at a depth of ca. 4m near the marina entrance (Fig. 1: 16). A portion of the hull side, measuring ca. 3 × 5m, survived, containing a number of wooden planks and frames, between which was cement ballast. The wood was dated by C14 to the Middle Ages (1062–1250 CE, uncalibrated RT 1422). These were probably the remains of a wooden ship ca. 25m long wrecked in the harbor (Galili et al. 1991; 2002a; 2002b).

Fig. 1: Location map (E. Galili) 313

Ehud Galili

Byzantine Shipwreck Assemblage at the Marina Entrance

all originating in Europe. A wooden carpenter’s square was also recovered from the wreck. In 2006 and 2007 the shipwreck was excavated by Y. Kahanov (University of Haifa). Raban (1982a, 1993) surmised that the wreck represented a ship sunk by the British at the harbor entrance during the siege by Napoleon. A C14 (RT 122) test dated the wreck to the twentieth century. However, the wreck’s location, north of the Island of Flies, away from the harbor entrance, as well as the late dating, do not support Raban’s thesis. (Galili et al. 1991).

An assemblage containing assorted items and wood remnants, originating in a Byzantine-period ship, was discovered at a depth of ca. 4m in the marina entrance (Fig. 1: 21). The hull seems to have been damaged during previous dredging operations. As these were carried out without archaeological supervision, only small, unidentifiable wooden parts were left. Nearby, a group of findings from the Byzantine period were discovered. These included chunks of raw glass (Galili et al. this volume c) fragments of jars (Silberstein et al. this volume), bronze nails and bronze coins, apparently originating from that vessel.

Shipwreck from the Nineteenth Century in the Eastern Basin Wooden ship remains were found near Hasusim (Horses) Beach in 1.6m of water (Fig. 1: 7). Fragments of a frame and planking, each ca. 3m long, were found; additional parts of that wreckage may be buried in the sediment. The wreck was dated by C14 to the nineteenth century (RT 1426) (Galili et al. 1991).

Remains of a Roman Ship Hull in the Northern Section of the Marina Several ship timber pieces bearing mortise and tenon joining devices were retrieved from the seabed during dredging works in the Akko Marina (Fig. 1: 17). C14 testing, as well as construction features, dated them to between the third and the fifth centuries CE. The relatively closely spaced un-pegged mortise-andtenon joints and hints of a “frame-based” constructed hull mark an early use of unpegged mortise-and-tenon joints and represent a relatively early application of the frame-based technique in ship construction in the eastern Mediterranean (Kahanov this volume).

Shipwreck from the Early Twentieth Century on the Eastern Basin Shoreline Remains of a wooden ship, constructed using iron nails, were discovered on the western edge of Hasusim Beach (Fig. 1: 8). This ship appears in photos from the mid-20th century and in plans of the harbor dating to the 1950s. It seems to have been beached and subsequently disintegrated. Its lower parts become buried in the sediment and are occasionally exposed (Galili et al. 1991).

Shipwrecks in the Eastern Basin Shipwreck Between the Island of Flies and the Southern Breakwater

Shipwrecks on the Eastern and Southeastern Coasts

Remains of a tin-sheathed wooden vessel were discovered in 1966 by the Underwater Exploration Society of Israel (Fig. 1: 10) (see above), and during a survey conducted by E. Linder and P. Throckmorton. The tin was attached to the hull by bronze nails, square in cross-section, and also by bronze tacks having a round cross-section (Raban 1982b; Flinder et al. 1993). Remains of the wood were not dated; however, judging by the construction style and the nails, it was estimated at between 100 and 200 years old.

Shipwreck Dated to the Late Seventeenth Century Remains of a wooden shipwreck were identified, protruding from the seabed near the Naval Academy (Fig. 1: 9). The remains were dated to the late seventeenth/early eighteenth centuries CE (RT 1425). Additional parts of the wreck may be buried in the sand. The vessel was beached and abandoned close to the shoreline (Galili et al. 1991; 2002a; 2002b).

Shipwreck Near the Island of Flies

Shipwreck Dated to the Twentieth Century

This wreck was discovered during the early 1970s, north of the Island of Flies in 4–5m of water (Fig. 1: 11). In October 1975 remains were recovered of planking, frames and copper alloy nails, with a square cross-section and covered by heavy marine encrustation. Wood samples from the ship indicated the presence of three different woods,

A whole wooden hull was discovered in 2.5–3.0m of water east of the eastern rampart/breakwater (Fig. 1: 5). Its remains protruded ca. 40cm above the seabed at that point. It was 23m long and 7m wide, and the bow pointed in the direction of the shoreline at an angle of 30o east. The outer surface of the ship was

314

Shipwrecks and Cargoes in and around the Akko Marina

sheathed by a thin sheet of metal (zinc or a copper alloy). The wooden parts were joined by bronze nails with a square cross-section and by brass tacks with a round cross-section. Trial excavations in the bow area (Fig. 2) and at the hull mid-ship were carried out during the Akko Marina Survey. The wooden parts were dated by means of C14 to the twentieth century CE (RT 427). Among the findings were whole, cylindrical glazed clay bottle, as well as fragments of a porcelain vessel with decorations in blue, a clay pipe and a whole jug (Galili et al. 1991; Grøn et al. 2015). Shipwrecks from the Ottoman Period Near the Southern Beaches Nearly intact wooden hulls of five additional vessels were identified off the sandy beaches southeast of the Naval Academy at a depth of 2–3m (Fig. 1: 2, 3, 4, 6, 21) (Fig. 3). Among them were three vessels dated to the twentieth century and one dated to the nineteenth century. The ships were dated by C14 as follows (Galili et al. 1991): No. 6, twentieth century (RT 1424); No. 2, twentieth century (RT 1430); No. 3, twentieth century (RT 1429); No. 4, nineteenth century (RT 1428) and No. 21 was not dated. The sides of vessels No. 2 and 6 were sheathed by thin sheets of copper alloy. Numerous camel and cattle bones were found near the ships.

Fig. 2: Bow of Ottoman wooden ship (Fig. 1 no. 5) from the early twentieth century (E. Galili)

Fig. 3: Remains of three Ottoman-era shipwrecks (E. Galili) 315

Ehud Galili

Shipwrecks South and West of Akko Cargo of Architectural Items South of the Pisan Quarter Several granite column drums (diam. 0.4–0.6m) and a whole granite column (diam. 0.8m) were found near the base of the modern breakwater, southwest of the sunken Crusader vault (Fig. 1: 18). The columns seemed to have been part of a cargo of construction items from Caesarea or another coastal city. They were supposedly brought by Jazzar Pasha in the nineteenth century CE to build and renovate Akko’s walls and breakwater. Thus, it is impossible to verify whether these Roman columns are associated with a ship that sank or were incorporated in secondary use into Akko’s fortifications and fell into the sea later, after the walls were destroyed, as was the case in Ashkelon and Caesarea. Ballast Stones South of the Entrance to the Marina During the Akko Marina Survey several clusters of non-local schist stones were discovered on the sea bottom about 200m south of the marina entrance (Fig. 1: 1). Water-jet drilling into the seabed and trial excavations at the site revealed a large group of similar stones. A ship seems to have been wrecked in this area whose remains were covered by sand. However, no wooden remains attributable to a ship were found. Alternatively, these stones represent ballast dumped before entering the port, prior to the loading of cargo to be exported from Akko.

Fig. 4: Portion of lead sheathing with nails that covered the keel of a Roman ship; the sheathing was torn off the ship when it ran aground on a reef (E. Galili) nails were discerned. The nails were square in crosssection (4 × 4mm), the diameter of the heads was 14– 15mm and the length of the nails was ca. 37mm. The nails were ca. 9.5cm apart and their rows were 3cm apart. The lead sheathing became corrugated this way when the keel scraped a rock, peeling, folding and tearing the sheath and finally depositing it on the reef. A number of ceramic vessels from various periods were discovered in the area. Among them: the base of a Rhodian amphora from the Island of Kos (second–third centuries BCE), the upper part of a Roman amphora (first–second centuries CE) and a Muslim-era chamber pot. Apparently these findings cannot be associated with the ship losing the lead sheath; it may be reasonably assumed that the ship was damaged but not wrecked at this point. Based on the nails and lead sheathing, the ship can be dated to the Roman period (first century BCE to mid-second century CE) (Y. Kahanov, pers. comm.). Based on the lead sheathing, the width of the keel can be estimated at ca. 30cm. A keel this wide is among the widest known; it could have been part of the keel of a ship ca. 30m long or more. A loaded ship of this size would have weighed a few hundred tons and its draft would have been 3m or more. Obviously the water depth over the reef during the Roman period

Assemblage of a Ship that hit a Shoal Southwest of the Marina A well-known underwater obstacle in the form of a rocky shoal is located a few hundred meters southeast of the Akko lighthouse; its peak is marked on charts at a depth of 3.4m (Galili et al. 2004; Galili et al. this volume a; Zviely et al. 2003) (Fig. 1: 22). Dives at the site revealed that the water is 3m deep at a number of points. In 1982, R. Shchomos, while diving on that shoal, discovered torn strips of a thin lead sheet that had been compressed, accordion style (like corrugated tin), and in which dozens of small bronze nails were embedded (Israel Antiquities Authority, unpublished report, Nov. 21, 1982) (Fig. 4). The total weight of the lead sheet was 57.5kg, and the total length of its parts (in their corrugated “accordion” form) was ca. 3m. The original length, before it was folded accordion-like, was ca. 10m. The sheet’s maximum width was ca. 36cm, its thickness was 1.4mm, and signs of tearing were visible on its sides. The nails attaching the lead to a ship’s keel were inserted in rows and in each fold a number of

316

Shipwrecks and Cargoes in and around the Akko Marina

Fig. 5: Cargo of columns and other architectural items (E. Galili) was similar to that at present. Thus, it may assumed that a large Roman ship sailing near the harbor scraped the above mentioned rocky shoal and thus lost a portion of its lead sheathing (Galili et al. 2002a; 2002b). Cargo of a Ship Carrying Architectural Items in Secondary Use, Found Northwest of Akko In 1990, the fisherman R. Shchomos reported finding a cargo of architectural items at a depth of 15m and a distance of ca. 1,200m west of the northern neighborhoods of Akko (35o 03’ 31” E; 32o 56’ 23” N). A number of survey dives at the site documented nine columns and parts of columns (diam. 60cm) and two particularly large kurkar ashlars (140 × 240 × 50cm) (Figs. 5, 6). Four columns were made of granite and the rest from gray marble. The columns were 200–450cm long. A few were broken into fragments of ca. 100–150cm. The findings were drawn and samples of the stone were taken. A ship bearing these architectural items seems to have been wrecked there (Galili 1987).

Marble Granite Kurkar

Fig. 6: Cargo of columns and other architectural items (plan: IAA) 317

Ehud Galili

Fig. 7: Iron grapnel anchor (E. Galili)

Acknowledgments

Two Iron Anchors with Four Arms (Grapnels) Found South of the Akko Marina

The author wish to thank the Israel Antiquities Authority for the institutional support; M. Feinberg Vamosh for the English editing; S. Ben-Yehuda for the drawing of fig. 6; and Howard Rosenstein and Fantasea for the underwater photography equipment.

Two large iron anchors were identified underwater, ca. 1,800m south of the marina, ca. 800m from shore, at a depth of 12m on the rocky seabed (Galili and Rosen this volume b). They were found lying 150m apart. Both originally had four arms, ending in triangular bills (Fig. 7), and cable rings on top. When discovered, the southern one was missing two arms. One anchor ring pointed east-southeast while the other ring pointed approximately north. Presumably both were lost by anchored ships during a storm. Similar anchors, usually dated to medieval times (tenth–fourteenth centuries), were used during the Crusades.

References Flinder A., Linder E. and Hall E.T. 1993. Survey of the Ancient Harbour of Akko, 1964–1966. In M. Heltzer, A. Segal and D. Kaufman (eds.). Studies in Archaeology and History of Ancient Israel. Pp.199– 225. Haifa. Galili E. 1987. The Underwater Survey–1986.

Hadashot Arkheologiyot 89: 59–60. (Hebrew).

Galili E., Dahari U., Sharvit J., Aviam M. and Stern E. 1991. Akko Marina Survey. Final Report Submitted to the Akko Development Corporation by the Israel Antiquities Authority. Pp. 1–54. (Hebrew).

318

Shipwrecks and Cargoes in and around the Akko Marina

Galili E., Sharvit J., Silberstein N., Zviely D. and Kahanov Y. 2002a. Acre Harbor – New Finds Revealed from Underwater Archaeological Surveys. In B. Galil and Y. Mart. (eds.). The Mediterranean Continental Margin of Israel, 14th Symposium. Pp. 41–42. Haifa. Galili E., Sharvit J., Dahari U., Silberstein N., Finkielsztejn G., Stern Edna, Kool R., Kahanov Y. and Rosen B. 2002b. Akko Harbor, Underwater Survey. Hadashot Arkheologiyot 114: 13–16, 12*–15*.

Silberstein N., Galili E. and Sharvit J. This volume. Hellenistic, Roman and Byzantine Ceramics from the Akko Marina. Zviely D., Galili E. and Rosen B. 2003. The Akko Harbor in Modern Nautical Maps. Horizons in Geography 56: 52–77. (Hebrew).

Galili E., Rosen B. and Zviely D. 2004. Akko and its Surroundings on Coins and Graphic Representations, Horizons in Geography 62: 113–130. (Hebrew). Galili E. and Rosen B. 2008. The Harbor. In E. Stern (ed.). The New Encyclopedia of Archaeological Excavations in the Holy Land 5. Pp. 1558–1561. Jerusalem. Galili E. This volume a. The Akko Marina Archaeological Project – Introduction. Galili E., Zviely D. and Rosen B. This volume a. Numismatic, Epigraphic, Cartographic and Aerial Photographic Evidence of the Akko Harbor. Galili E., Rosen B. and Sharvit J. This volume c. Raw Glass Chunks and Glass Vessels from the Akko Marina. Galili E. and Rosen B. This volume b. Anchors from Akko Marina. Gron O., Boldreel L.O., Cvikelc, D., Kahanov Y., Galili E., Herman J.-P., Nævestad D. and Reitan. M. 2015. Detection and mapping of shipwrecks embedded in sea-floor sediments. Journal of Archaeological Science: Reports 4: 242–251. https:/ /www.infona.pl/resource/bwmeta1.element.elsevierb3a000fb-8e4a-394e-b1fe-738d78503329 Kahanov Y. This volume. Ship Timbers from the Akko Marina: Early Use of Unpegged Mortise-and-tenon Joints and Mixed Shell-and-frame Construction. Raban A. 1993. Maritime Acco. In E. Stern (ed.). The New Encyclopedia of Archaeological Excavations in the Holy Land 1. Pp. 29–31. Jerusalem. Raban A. 1982a. The Harbors of Akko. In M. Yeda‘aya (ed.). The Antiquities of Western Galilee. Pp. 180–207. Tel Aviv. (Hebrew). Raban A. 1982b. A Fragment of a Ship from the Napoleonic Siege on Akko (1799). In M. Yeda‘aya (ed.). The Antiquities of Western Galilee. Pp. 198– 207. Tel Aviv. (Hebrew).

319

Chapter 22

The Akko Marina Archaeological Project – Summary Ehud Galili (1, 2) and Baruch Rosen (1) 1. Israel Antiquities Authority 2. Zinman Institute of Archaeology, University of Haifa Key words: Acre, underwater archaeology, harbor, shipwrecks, Bronze-Age, Persian, Hellenistic, Roman, Byzantine, Early Islamic, Crusader, Ottoman, sea-level changes

Introduction

1979 and references therein). Excavations at Tel Akko uncovered remains dated to the Early Bronze Age and thereafter during most historical periods (Dothan 1993). Abroad, an early mention of Akko (Fig. 2) already appears in an Egyptian inscription from the temple of Amon at Karnak, dated to the

The maritime archaeological complex of the Akko headland and Tel Akko (Fig. 1) reveal ca. 4000 years of alternating building and destruction (Makhouly and Johns 1946; Shore 1990; Galili et al. 2010; Dichter

Fig. 1. The Akko headland and Tel Akko (modified after Google Earth) 320

The Akko Marina Archaeological Project – Summary

North Africa

Fig. 2. Top: Akko headland looking northwest, Bottom: Location of Akko and sites mentioned in the text (photo: A. Ohayon, Drawing: E. Galili) fifteenth century BCE (Cohen 2002: 70; Flinder et al. 1993: 200). The modern traditional city on the

periods. The materials presented and discussed in this volume were collected over time in several archaeological expeditions conducted by the IAA and other institutions. The marine archaeological operations in the Akko Harbor, are detailed in Chapter 1 (Galili this volume a). The latest marineoriented archaeological excavations were conducted at the foot of the southeastern Ottoman sea wall from 2008 to 2013 (Sharvit et al. 2013) and are mentioned briefly below. The finds described in this volume are summarized and discussed chronologically below, along with the relevant historical events and issues involving maritime activities, the harbor functions and its location in the different periods.

headland is a living urban complex. Its marina, a busy maritime center, is built over the ancient harbor; thus, systematic archaeological excavation was not possible there. The archaeological records of the cycles of building and destruction of the ancient harbor have been partly erased by modern construction. Hence, the archaeological materials recovered during the 1992–1993 Akko Marina Archaeological Project and other underwater archaeological projects are crucial for filling the information gap. They add much to the understanding of the maritime activities in the city and the functioning of its harbor in various 321

Ehud Galili and Baruch Rosen

Paleogeography and Coastal Changes in the Akko Region

coastline in the center of the Haifa Bay embayment reached more than 2km east of the recent coastline.

Based on the available archaeological and geological data and on pervious reconstructions, a tentative reconstruction of the paleo-geographical changes in the Akko headland and surroundings is presented. A more detailed reconstruction of the paleogeography of the Akko peninsula should await the processing and studying of additional core drillings to be carried out in the area presently under the Old City.

d) Ca. 4000–3800 BP (the Middle Bronze Age): The mega-embayment of the Haifa Bay reached its maximum penetration eastward to the Zevulun Plain (ca. 3km in the center of the Haifa Bay and up to 5km in the Kishon River area to the south) (Fig. 4). In that period sea level reached its present elevation (Fig. 5). Since that period sea level has been relatively stable (see below) and only horizontal coastal changes took place, mostly due to sedimentation (sand accumulated in the Haifa Bay) and to some extent, due to erosion in modern times. The morphology of the Akko headland during when sea level stabilized at its present elevation, is important for reconstructing the paleo coastline and assessing the possible existence of a sheltered anchorage there. On the western coast of the Akko headland is a north-south oriented kurkar ridge running parallel to the coastline (Sivan 1981: 66, figs. 17, 18, 19). Remains of this ridge can be seen as bedrock protrusions on the sea side edge of the south western rock cut platform (Galili this volume e fig. 33) and on the walls of the Crusader tunnel (see below Figs. 14-16). The Kurkar hill is also seen in a panoramic view of Akko (Galili et al. this volume a: fig. 17) dated to 1686. Kurkar bedrock was also recovered in rescue excavations in the Northern ditch (Tatcher 2007), in the southern promenade (2.5 m above sea level, Porat 2013), in the excavations of the castle (more than 5m above sea level, Danni Syon pers. comm. 2116) and north of the Old city (+ 4m above sea level, Amani Abu-Hamid pers. comm. 2016). Traces of another Kurkar ridge located to the east (+ 5m above sea level) were recovered in excavations at the Courthouse site (Hartal 1997: 22, 23 table 1) and at the knights hotel site (Danny Syon pers. comm. 2016) (Fig. 7). The area beneath the Old City was not mapped by Sivan or by others due to lack of data. However, it seems that there was a shallow trough under the south central part of the Old City. In the trough east of the coastal ridge, the kurkar top is few m below the present sea level (Sivan 1981 figs. 5, 17). The middle Holocene rising sea flooded the trough and may have created a body of water east of the coastal ridge (Fig. 7: a). During the Middle and Late Bronze Ages and in the early Iron Age, this area (today beneath the Khan al-Umdan and the Venice square) was most probably protected from western and northwestern winds. It could have been used as a sheltered anchorage and for hauling small and medium-size water craft during the summer and the transition seasons. Recent core drillings carried out in Khan al Umdan and excavations carried out in Venice Square north of this Khan reveled Crusader debris down to the water table to an unknown depth (Eliezer Stern pers. comm. 2016). These finds may support the proposed reconstruction above.

a) Ca. 9000–8500 BP: The sea level was ca. 16– 18m lower then today, the coastline west of Akko headline was located some 1000m to the west (Fig. 3). About 2km offshore west of Akko was a narrow elongated island (the currently submerged Vernon reef/kurkar ridge) that continued southward, almost continually, until the center of Haifa bay. The currently submerged Akko ledge (kurkar reef) and the Foxhound reef south of the Akko headland were exposed, and formed a massive, wide peninsula. This currently submerged north-south oriented peninsula was crossed by the channels of the Hilazon and the Na‘aman river. The river canyon of the paleo Hilazon River was observed on the seabed southwest of the Akko headland (Fig. 3). The submerged river channel is located on a geological fault known as the Ahihud Fault, constituting the boundary between the Haifa Bay graben and the Galilee coast (Galili and Eitam 1988; Zviely 2006: fig. 90; Galili et al. in press). Between the peninsula (the Early Holocene Akko headland) and the coast, was an elongated coastal lagoon (Fig. 3). The coastline in the center of the Haifa Bay was ca. 1.5km west of the present coastline (reconstruction modified after Zviely 2006; Zviely et al. 2006, Porath et al. 2008). b) Ca. 7500–7000 BP: Sea level was ca. 8–9m lower than today, the coastline west of the Akko headland was some 200–300m to the west and the summits of the Vernon ridge created small shallow reefs some 3km offshore (Zviely et al. 2006: 157, fig. 93). The Akko headland was an elongated, north-south oriented peninsula, extending some 3km south of the recent headland. The coastline and the small Akko Bay at that period were some 2km southwest of Tel Akko. In that period the mega-embayment of the Haifa Bay started to form when the rising sea flooded the Zevulun Plain, and the coastline in the center of the bay reached ca. 1.5–1.7km east of the recent coastline. c) Ca. 6800–6500 BP: Sea level was ca. 4–6m lower than today, the coastline west of the Akko headland was some 200m to the west. The Akko headland was slightly larger and some 200m south of it was a small reef, the summit of a currently submerged kurkar ridge. In that period the coastline and the small Akko Bay were some 1.5km southwest of Tel Akko. The

Paleogeography of the Akko headland: Several scholars have reconstructed the paleogeography of Akko and its surroundings since the 1980s. Based 322

The Akko Marina Archaeological Project – Summary

Fig. 3. The Haifa Bay and Akko region ca. 8500–9000 years BP, when sea level was ca. 15m lower, the coastal rivers and their submerged courses (modified after Galili and Eitam 1988; Zviely 2006; Galili et al. in press)

Fig. 4. Paleo-geography of Haifa Bay ca.4000 BP, with the location of the ancient tell settlements (after Zviely et al. 2006) on core drillings, shells-containing rock and upper clay deposits, Sivan draws three east-west oriented cross-sections in Akko. She proposed that at ca. 4000 BCE (the time of the beginning of the occupation in Tel Akko) two bays reached the fringes of Tel Akko (Sivan 1981: fig. 18, pp. 56, 58, 87). One of these proposed bays was open to the sea at its northern edge (the northern bay) and another was open to the

Fig. 5. Sea-level changes in the Carmel coast region during the Holocene (after Galili and Rosen 2011) sea at its southern edge (the southern bay) (Fig. 6). Dothan who excavated the Middle Bronze Age IIA city gate at the northwestern side of the tell proposed that it was the “sea gate,” which faced the proposed northern bay/lagoon that could have provided safe anchoring conditions (Dothan 1978). In later investigations (Sivan 1996: 123) the shells-containing rock deposit was dated to the last interglacial high sea stand (ca. 120,000 years BP) when global sea level was ca. 6m above present level. The upper clay layer designating the proposed northern bay, was not sufficiently dated by C14. This clay was associated with the same body of water that deposited the shells-containing rock, which later became shallower (Sivan 1981: 84). However, this clay, with its upper surface some 1–2m above sea level, may have been deposited during the early Holocene and even during the late Pleistocene. Given the elevation of the upper clay above sea level, a navigable northern bay could not have existed in the northern area of Akko at 4000 BP. The area of the proposed northern bay could have been a wetland/marshy area, until modern times, as clearly seen in aerial photo taken in 1918 (Kedar 1991: 220: fig. m6). Thus, the above reconstruction of the two bays (Sivan 1981) should not be associated with maritime activity during historical periods. Another reconstruction of the Akko region was proposed by Raban, who suggested that a marine lagoon existed southwest of Tel Akko during the Bronze Age. It was created in the estuary of the Na’aman/Hilazon and the Akko streams, and it could have served as a port during the Hellenistic Period (Raban 1982: 180, 181, fig. 3). He also suggested that another lagoon, that may have served as an ancient harbor, existed in the marshy area north of the old city (Raban 1982: fig. 3). 323

Ehud Galili and Baruch Rosen

been proposed that during the Bronze Age, ships anchored/moored in the relatively protected leeside of kurkar reefs, partly submerged, at some distance offshore (Galili 1986, 2009; Galili and Sharvit 1994). Underwater surveys conducted over the past 50 years along Israel’s coast yielded evidence for such natural shelters at Tell Riddan, Yavne-Yam, Jaffa, Apollonia, Caesarea, Neve Yam and Atlit (Galili et al. 2002b; Galili and Rosen 2008; Galili et al. 2010).

A recent sedimentological and palaeontological study in Tel Akko region (Morhange et al. 2016: fig. 10) confirms Raban's reconstruction and suggests that an open estuary existed south and southwest of Tel Akko (until 2800 BP) and southwest of the Tel until 2000 BP. According to this study, the estuary provided an open anchorage, enabling the existence of a harbor adjacent to the Tel (Morhange et al. 2016: 80). However, so far, archaeological excavations failed to reveal traces of such harbor.

The Zevulun Plain tell settlements and their association with the sea: Some 4,000 years ago global sea level stabilized after it has risen for the previous 17,000 years from -120m at the last glacial maximum (Galili et al. 2005; Galili and Rosen 2011) (Fig. 5). The Paleogeography of the Haifa Bay and the Zevulun Plain during the Holocene was reconstructed by Zviely (2006), Zviely et al. (2006) and Porath et al. (2008). The post-glacial rising sea flooded the Zevulun Plain forming a mega-embayment, and the coastline shifted inland, a few kilometers east of the present coastline. In the Zevulun Plain, a few kilometers inland, there are several ancient archaeological mounds arranged in a line parallel to the coastline of the Haifa Bay (Fig. 4). It has been suggested that these Bronze Age settlements, including Tell Dauk, Tel Afeq, Tell Zibda, Tel Gadora, Tell Zevet, Tell ed-Idham and Tel Naḥal, were coastal settlements relying on marine subsistence (Fleming et al. 1978). However, the geology of the sediments in the Zevulun Plain suggest that these settlements were never on the coast during historical periods. Having no access to the sea, they could not have been marine oriented (Zviely et al. 2006) (Fig. 4). The pattern of their location (in a line parallel to the coastline) was probably because they relied on agriculture, exploiting the productive Zevulun Plain soil and groundwater resources, while avoiding the marshy and sandy areas separating them from the coast.

e) Ca. 3500 – 2800 BP: During the Late Bronze, the Iron Age and the Persian periods sea level remained at its present level, but the coast in the Haifa Bay migrated several kilometers westward due to largescale sedimentation of Nilotic sand in the bay and the formation of sand dunes in the Zevulun Plain. The westward migration of the coastline to its current location continued until the early twentieth century (Zviely 2006: 163). It is assumed that similar process occurred in Akko bay. The flooded trough that was formed in ca. 4000 years BP, gradually silted up with sand and became shallower and smaller with time (Fig. 7). The city on the peninsula was established during the Hellenistic period.

The Archaeological Remains in the Historical Maritime Context The Bronze Age Only meager remains from the Late Bronze Age were recovered during the project. These include two stone anchors and a copper ingot, dated by form only. It is not certain that these remains are to be dated to the Late Bronze Age, because loaf-shaped ingots and stone anchors with one hole were also used after the Bronze Age. The metal figurines found in the harbor (Galili et al. 2002a; Galili this volume b) are similar to Levantine figurines dated by their style to the Late Bronze Age and the Iron Age I, however they may also be of a later period.

Tel Akko harbor/Na‘aman channel anchorage?: Based on core drilling south of Tel Akko, in the northern part of the Zevulun Plain, it was suggested that the sea reached the archaeological mound during the Middle Bronze Age (Zviely 2006; Zviely et al. 2006; Porath et al. 2008; Morhange et al. 2016) (Fig. 4). However, so far no harbor or installations associated with maritime activities has been discovered during the archaeological investigations adjacent to the tell. Given the sedimentation patterns of coastal streams in Israel, it seems that ships could not have entered the Na‘aman channel, as proposed by Raban (1982, 1983a, 1983b, 1985, and see discussion below, Bronze Age). Nevertheless, it cannot be ruled out that small boats sailed to some distance inland in the river channel, with no free passage to the open sea.

Bronze Age anchorages along the Israeli coast: During the Middle Bronze Age, sea level reached its recent height (see above). It has been claimed that the coastal streams of Israel served as inland harbors during the Bronze Age (Raban 1982, 1983a, 1983b, 1985). Yet no archaeological evidence (shipwreck, installations, anchors, etc.) of an anchorage or a harbor has so far been discovered in a channel of an Israeli coastal stream. Given that the main patterns of local coastal sedimentation at that time were similar to the recent ones, we may assume that the inlets of coastal streams were blocked by sand most of the year, as they are today. The river channels were shallow, silted by sand that formed delta barriers prohibiting navigation by seagoing vessels. Therefore, it has

324

The Akko Marina Archaeological Project – Summary

The Iron Age The Akko area is mentioned in the Old Testament in the books of Joshua and Judges as part of the territory of the tribe of Asher (Kashtan 1988). No artifacts or structures discovered in the Akko Marina can be dated to the Iron Age. However, numerous remains from the Iron Age were found at Tel Akko (Dothan 1993). Considering the archaeological evidence for commercial activity in some coastal cities of the Southern Levant during the Iron Age, the lack of such

findings in the Akko Marina is somewhat surprising. During the Bronze Age and the Iron Age the Akko headland served as natural protection for ships sailing along the Levantine coast (see above, Fig. 7: a, 7: b). Watercrafts probably anchored there for the night waiting for a favorable wind, took on food and water, and loaded and unloaded merchandise and passengers. It is possible that remains of shipwrecks, cargoes and anchors dating to these early periods were obliterated by subsequent activities in the port. Such evidence may have also been covered by sediment and could not be reached during the deepening of the marina. Future excavations under the Old City may reveal finds associated with maritime activity from these early periods. The Persian Period

Fig. 6. Paleo-geography of the Akko headland at ca. 4000 BP proposed by Sivan 1981 (modified after Sivan 1981)

Fig. 7. Possible coastal changes in the Akko region: a) at ca. 4000 BP when sea level reached its present elevation; b) in ca. 3000 BP; c) in ca. 2300 BP when the Hellenistic/early Roman harbor was built; d) in ca. 700 BP during the Crusader/Ottoman periods (modified after Raban 1982: 189–190)

In the Book of Judith, dated tentatively to about the 436 BCE, Akko (Okina) is mentioned as one of the coastal cities of the Eastern Mediterranean, the others being Sidon, Tyre, Jamnia, Izotus and Ashkelon. According to written sources, in 374 BCE Akko was the staging ground for the Persian army, numbering 22,000 men and numerous warships, on its way to Egypt (Kashtan 1988). Salvage excavations adjacent Tell Akko suggest that during that period the settlement expanded beyond the perimeter of the tell, creating a 500m wide occupied belt around its eastern, northern and western fringes (Amany Abu -Hamid pers. comm. 2017). The assemblage retrieved from the Akko Marina revealed only a few Persian-period sherds (Silberstein et al. this volume). According to Raban (1982, 1983a, 1983b, 1985) and Linder and Raban (1965), the harbor’s southern breakwater (Figs. 8, 9, 10) was built during the Persian period. This thesis is based on the header method of construction found in the western part of the southern breakwater and on a sherd, dated to the sixth century BCE, bearing a Phoenician inscription, found out of context. Raban (1982, 1983a, 1983b) proposed that the harbor was built for the Persian fleet of Cambyses because the Na‘aman River channel (which, he believed, served as an inland harbor during that time) could not hold such a large fleet (Fig. 1). Based on the header construction of the Island of Flies and the similarity to other Phoenician harbors (an emporium built on a small island near the shoreline), Linder and Raban (1965) suggested that the island was constructed by the Phoenicians during the Persian period. The meager remains from the Persian period in the marina indicate that at this time there was no built harbor in marina area. It is reasonable to assume that Akko maritime activity during the Persian period, as in earlier periods, was based on natural features creating a relatively protected area in the southeast, and possibly in the small bay currently under the Old City (see above) (Fig. 7: b). The existence of yet undiscovered Persian

325

Number of tools

Ehud Galili and Baruch Rosen

Fig. 8. (a) Work areas of the Akko Marina project; (b) Distribution of amphorae from the Hellenistic, Roman and Byzantine periods in the marina work areas; (c) Origin of the amphorae in the Mediterranean (after Galili et al. 2010) 3.36´´ (hence the separation line between rocky sea bottom and sea bottom composed of loose sediments) were constructed mostly on unconsolidated sediments (Figs. 9, 19). These include the southern breakwater, the eastern rampart, the artificial Island of Flies and the southeastern city seawalls. Historical documents show the prominence of Akko during the Hellenistic period (Kashtan 1988). Akko may have flourished during this time, due to the destruction of Tyre Harbor by Alexander the Great (July 332 BCE), replacing Tyre as a center of

remains associated with marin activity in the Akko bay area, or beneath the mediaeval city or on land east of it, cannot be ruled out. The Hellenistic Period The archaeological finds suggest that the first building of maritime structures started during the Hellenistic period. In principle the maritime structures and installations built at sea east of longitude 35º 04´

326

The Akko Marina Archaeological Project – Summary

Sand and gravel Sand and gravel

Sand and gravel

Quarry abrasion platform Island of Flies

Fig. 9. The Akko Harbor region and the location of the main archaeological features and finds (E. Galili) commerce and harbor services (Flinder et al. 1993). The city was conquered in 312 BCE by Ptolemy I Legos, king of Egypt, on his return from Syria, as were the coastal cities of Jaffa and Gaza. In 208 BCE, Akko was given the name Ptolemais by Ptolemy II and the city grew and prospered. Antiochus III occupied Akko during his conquest of Syria in 219–217 BCE. The king and his army wintered in Akko, a sign of the city’s high status at that time. In the second century BCE Akko was a pagan center hostile to the Jews. Its inhabitants demanded that Antiochus IV Epiphanes impose the same edicts on the Jews of Akko as those imposed on the Jews of Jerusalem. The conquest of Akko in 163 BCE by Simon the Maccabee, the leader of the Jewish revolt against the Hellenistic regime, was intended, among other things, to ensure the safety of the city’s Jews. In 52 BCE the city became a royal colony of the governor of Demetrius, Alexander Balas, who married Cleopatra, the daughter of Ptolemy V, in Akko. The pagan inhabitants of the city, which came under the control of the Jewish king Alexander Jannaeus at the end of the second century BCE, appealed to Ptolemy VII Lathyrus, who arrived from Cyprus and landed at Shikmona, aiming to assist in the abolition of Alexander Jannaeus’ rule. According to archaeological findings retrieved from the marina (Figs. 8: a, 9) and the sedimentological patterns, the western basin underwent a change during the Hellenistic period (Galili et al. 2002a; Galili and Rosen 2008) (Fig. 7: c). As noted, only meager finds were discovered from the Persian period and the preceding times. However, thousands

of finds emerged from the Hellenistic and Roman periods, among them amphorae, vessels for everyday use and various other items, many of them from Areas D, E and F (Fig. 8: b). The findings show wideranging links between Akko and the more distant Mediterranean shores, including extensive imports of products from the Aegean. Among the finds from the Hellenistic period are black-slipped vessels, black and red Megarian bowls, Eastern Terra Sigillata vessels, mostly of type A and some of types B, C and D. Western Slope Ware was also discovered, as were Western Terra Sigillata vessels. The amphorae used for maritime transport constitute a significant part of the ceramic assemblage (Silberstein et al. this volume). Among the amphorae, dozens of types were found to have originated throughout the Mediterranean basin, particularly from the Aegean Sea, Italy, North Africa, southern Spain, France and the Black Sea coast (Fig. 8: c). Most of the amphorae (approximately 50%) came from the Aegean, 20% from the Syrian coast and the present Israeli coast, 5% from Egypt and a smaller portion from Italy. About 50 stamped jar handles were discovered during the dredging operation, in addition to many unstamped handles (Finkielsztejn this volume). The amphorae handles are presented in a separate article (Finkielsztejn this volume), detailing the extensive information they furnish. As in other cases in Israel, most of the identified handles originated in Rhodes, and were dated to the first half of the second century BCE. However, handles have also been dated to the second half of that century, as well as to the transition 327

Ehud Galili and Baruch Rosen

E

E

 N

Fig.10. Aerial photograph of the harbor and its surroundings: (a) the proposed location of the chain closing the western basin (according to Raban 1983a, 1993b; Flinder et al. 1993; Jacoby 1979); (b) the proposed location of the chain closing the eastern basin (according to Gertwagen 1996); (c) the eastern rampart; (d) location of the Crusader tunnel leading to the Templar district; (E-E) location of the cross section in Fig. 17 (after Galili et al. this volume a) et al. 2010 and Gambish (2013) the harbor (the

period prior to the first century BCE. More than half the handles were dated from 188 to 166 BCE, the period of major activity on Rhodes (Finkielsztejn this volume). It is noteworthy that no handles from the third century BCE were discovered. The finds from the marina (the stamped handles in particular and the ceramic assemblage in general), show that the southern breakwater was built in the first third of the second century BCE. Finkielsztejn suggested that the large number of handles dating from the begining of the second century could be associated with the battle of Banias (198 BCE), which took place at that time, when supplies for the army of Antiochus III may have been imported via the Akko Harbor. According to Jacoby (1979), Gertwagen (1996), Flinder et al. (1993); Galili and Rosen 2008; Galili

southern breakwater) was built during the Hellenistic/ Early Roman periods. The archaeological findings recovered from Akko Marina and from the area at the foot of the tell, as well as the historical records, confirm that. The recent finds at the foot of the Ottoman wall (Sharvit et al. 2013; Sharvit and Planer 2016) also support that assumption. Kesten (1993) argued that there was no built harbor at Akko during the Hellenistic and Roman periods and that the harbor was built during the Early Islamic era. He claimed that the Akko Harbor could not have existed unless supported by a nearby city, and the settlement on the tell was too distant. His argument is not supported by the archaeological finds recovered from the marina, which date the construction of the southern breakwater

328

The Akko Marina Archaeological Project – Summary

to the beginning of the second century BCE. So far no evidence has been found that the harbor was built during the Persian period, as suggested by Raban (1982, 1983a, 1983b, 1993). However, it should be remembered that the finds discussed here were retrieved by a dredger and it is possible that in lower strata, beneath the sediments, there are earlier remains that were not reached. Also, earlier remains may be overlaid by sediments under the old city and east of it, and in areas of the Akko Bay not yet excavated. The Roman Period According to historical sources Akko was one of the most important coastal cities in the region, a gateway to the sea and an imperial strategic base during the 133 years between the conquest of Pompey (63 BCE) and the end of the Jewish revolt against the Romans (70 BCE) (Kashtan 1988). Dignitaries visited the city and numerous events occurred there stressing its prominence. In 47/8 BCE, Julius Caesar landed at the Akko Harbor on his way from Alexandria. In 29 BCE, King Herod landed at Akko after his return from Rome. After the battle of Actium in 31 BCE it was in Akko that Herod met Octavian, the future Caesar Augustus, and held a grand feast in his honor. After King Herod’s death in 4 BCE, Akko became the main port for the Roman legions controlling Galilee. King Agrippa sailed from the Akko Harbor to Rome after disagreements with Pomponius, the governor of Syria. The apostle Paul arrived by ship to the Akko Harbor on his return from his third journey to Asia Minor. In 56 CE, during the reign of Emperor Nero, the ancient Via Maris was renewed and a road was built along some 300 miles from Antioch, connecting Asia Minor, Syria and Egypt with Akko as an important strategic stop along the way. The Akko Harbor served as a maritime base and a temporary camp in 66–70 CE for the armies of Vespasian and Titus, when the rebellion in the Galilee was suppressed. Findings from the marina show that in addition to its military roles, at that time, extensive commercial ties existed between the harbor and various regions and coastal cities around the Mediterranean. Ceramics of various kinds, used to import foodstuffs, were among the notable items retrieved from underwater. Luxury items such as bowls and Western Terra Sigillata ware, rarely found in land excavations, were also recovered. The rare artifacts were apparently imported to meet the demands of an affluent society. The amphorae found indicate a significant rise in the extent of trade and ties with the entire Mediterranean basin. That whole basin had by then become a unified commercial and political entity under Roman rule. About half of the identified Roman amphorae originated from the Aegean Sea and the Black Sea, about a fifth from Italy and the Western Mediterranean, about 16% from the coast of Syria and Israel, and another 10% from the coast of Africa

(Fig. 8: c). Coins minted in Akko during the Roman period indicate that Akko had a closed harbor, with ship sheds and/or storehouses and apparently a lighthouse as well (Galili et al. this volume a). The remains of the Roman ship discovered in the marina (see Kahanov this volume), the coins (Kool this volume), the ceramic findings (Silberstein et al. this volume; Friedman et al. this volume) and coins minted by the city, all indicate that Akko was a major harbor serving the region and a marine gateway to the, hinterland, as previously proposed by Raban (1982: 183). However, it is not clear why Josephus Flavius, when describing Akko and its surroundings in The Jewish War, did not mention the harbor (Kesten 1993: 20). Kesten argued that Josephus, who described the Caesarea Harbor in detail, would not have ignored the existence of an important built harbor at Akko. Thus, Kesten proposed dating the construction of the harbor to the Early Islamic period (see above). As noted above, the findings in the marina and the coins minted in the city during the Roman period (Galili et al. this volume a: 2, 3) indicate that a built harbor existed in Akko during that era, and so the reason Josephus did not mention it remains unclear. A glimpse of sailing conditions and sea level during the Roman period may be revealed by the remains of a marine accident found on an underwater reef located a few hundred meters south of the Akko Marina. This kurkar ridge, and its peak, ca. 3m below sea level, appears in plans and maritime charts (Galili et al. this volume a, figs. 26, 27, 28). Bronze nails and portions of Roman lead ship sheathing were discovered atop the reef (Galili this volume f). The sheathing was torn from the bottom of a ship’s keel that scraped the reef. Judging by the dimensions of the keel (obtained from the torn lead sheathing) the size and draft of the wrecked ship could be estimated, suggesting that a medium-sized Roman ship with a draft of ca. 2m was endangered by the reef. Given that swells of ca. 1m high are common in the area, it seems that during Roman times the summit of this reef was at the same depth as nowadays. This implies that no major sealevel changes have occurred in the region since the Roman period. The Byzantine Period Historical evidence indicates maritime activity in Akko and even mentions a shipyard (Jacoby this volume). However, the archaeological finds in the Marina suggest that the western basin may have been neglected during the Byzantine period. Nevertheless, the numerous shipwrecks and lost iron anchors from this period discovered along the coast of Israel suggest extensive maritime activity (Galili and Sharvit 1994). Numerous Byzantine ceramics were discovered in the Akko Marina at the harbor’s entrance. Most of them belong to three types only (Silberstein et al. 329

Ehud Galili and Baruch Rosen

ships anchoring in the harbor. Others (Linder and Raban 1965: 190; Raban 1983a, 1993; Jacoby 1979) surmise that in the Early Islamic period the harbor was in the western basin. Since only meager finds from that period were discovered in the western basin, this theory cannot be substantiated. However, the paucity of finds from the Early Islamic period in the western basin may be due to dredging operations prior to the study described here, which could have removed strata containing finds from that period. Notably, trial excavations at the eastern rampart north of the Island of Flies found no remains of construction like those described by Muqaddasi (Flinder et al. 1993; Yurman 2013: 72). Raban accepted that the eastern rampart was built by Ibn Tulun in the Early Islamic period; however, he describes it as a rampart connecting the Tower of Flies to the city wall and stated that it also served to block sediment from the east that could have silted the harbor (Raban 1983a, 1993). Judging by the physical conditions in the harbor area there is no need for a breakwater in the east. The short fetch (an area of ocean surface over which the wind blows in an essentially constant direction, thus generating waves) between the harbor and the eastern coastline prevents the raising of significant waves by easterlies (Galili and Rosen 2008; Galili et al. 2010). Based on the archaeological findings during excavations and underwater surveys of the eastern rampart, it is proposed that this rampart was built of gravel and debris and served as a means of access to the Island of Flies, like the road leading to the Sea Castle at Sidon (Galili et al. this volume a fig. 3). The rampart may have also been intended to impede access of hostile vessels from the east to the harbor entrance. As noted above, some of the debate regarding the Early Islamic harbor stems from the different translations. According to Raban, an arch was built over the harbor entrance, while according to Grabois, a bridge was built over the harbor’s western gate (Grabois 1983). Drori’s translation, presented by Kesten (1993), mentions neither a bridge nor an arch. Another translation involves the attachment of the breakwater described by Muqaddasi to the existing breakwater. According to Raban (1982), the new structure described by Muqaddasi was attached to the existing, earlier breakwater: “and then he renewed the construction by connecting the new structure to the sea walls (breakwater) of the old harbor.” This may suggest that during the Early Islamic period the breakwater was built in the western basin and was connected to the ancient southern breakwater. In contrast, Kesten presents a translation describing a link between the new structure and the ancient wall – not necessarily to the sea wall (the breakwater) of the old harbor “When construction reached the ancient wall, he joined them with a seam.” Although the abovementioned historical sources describe extensive maritime activity in Akko during the Early

this volume). This suggests a single wreckage event of a ship at the harbor entrance, perhaps during a storm. About half the Byzantine jars came from the Aegean Sea and the Black Sea, ca. 40%, came from the coast of Syria and Israel, ca. 3% from North Africa and 6% from the Western Mediterranean (Fig. 8: c). These findings show that during the Byzantine period trade was mainly conducted with the SyriaPalestine coast and that few commercial ties existed with Western Mediterranean countries. The broken iron anchors discovered (Galili and Rosen this volume b) suggest that anchoring conditions in the harbor were poor, and ships moored by casting anchors. The amphorae from the eighth century CE show that maritime trade continued during the Early Islamic period by means of Byzantine merchants, sailors and ships that visited the Akko Harbor.

The Early Islamic Period The eastern rampart and the Early Islamic harbor: A major construction project in the Akko Harbor by Abu Bakr, a Jerusalemite architect employed by the ruler Ibn Tulun (reigned 878–884), was described by the Islamic historian Muqaddasi (945–991) (Galili this volume e). The description does not mention the construction of a breakwater, but rather of a structure attached to the southern end of the city wall (Shor 1990). According to Muqaddasi the structure was built of segments reinforced by pillars and the harbor was similar in shape to the “double harbor” in Tyre. In 1047, the Muslim scholar, traveler and writer Nasir Khursaw described the harbor as resembling a stable whose back portion faced the city and having two fortified walls extending into the sea (Raban 1982; Gertwagen 1996). These historical testimonies, despite their detailed nature, leave a number of important issues unresolved: the location of the harbor and its extent, the function and nature of the eastern rampart and the location of the chain blocking the harbor according to written sources (Kedar 2013). Different translations of the historical sources have contributed to the uncertainty and led to various interpretations of issues concerning the harbor and its installations (Kesten 1993: 17). Muqaddasi’s description states that the harbor was in the west, which does not fit the topography of the harbor area. It is unlikely, and does not conform to theories presented above regarding the harbor’s evolution. A submerged, elongated structure (the eastern rampart) extending from the southeastern end of the city wall southward to the Island of Flies, is clearly seen in aerial photos (Figs. 2, 10: c). Gertwagen (1996) proposed that this rampart was the harbor constructed by Ibn Tulun during the Early Islamic period, the construction of which was described by Muqaddasi. She suggested that it served as a breakwater to protect the harbor from eastern winds that raised waves endangering

330

The Akko Marina Archaeological Project – Summary

Islamic period, only a few sherds were found from that time, mainly imported. They were discovered mainly in the western basin, while the later ones (from the Crusader period and thereafter) were found outside this basin. Numismatic findings from that period include 10 coins (Kool this volume) found in the western basin. The ceramic and numismatic findings suggest that during the Early Islamic period, maritime activity took place in the western basin. Judging by the archaeological finds recovered so far, the location of the harbor whose construction Muqaddasi described cannot be ascertained. In excavations on the coastline north of the western basin no pottery from the Early Islamic period was discovered. However, excavations on land north of the eastern basin yielded numerous pottery vessels from that period (E. J. Stern, pers. comm. 2011). The ceramics from the Early Islamic period found in land excavations suggest that the center of activity at this time was on the eastern basin shore (Arnon Yael pers. comm. 2011). Thus, neither the historical sources nor the archaeological findings in and around the harbor can pinpoint the harbor’s location during the Early Islamic period and the place of the entrance blocked by the chain. To resolve these issues, large-scale underwater and land excavations, as well as core drillings must be carried out at the eastern rampart, at the eastern end of the southern breakwater, at Hasusim Beach, in Khan al-Umdan and on the sea bottom in the Eastern basin. Solving these archaeological questions can assist in determining future planning policies for expanding the existing harbor. A policy document for the coastal waters prepared in 1999 by Israel’s Ministry of the Interior states that no large marinas exceeding the ancient harbor’s boundaries should be built in coastal historic cities. If it turns out that the Early Islamic harbor was within the boundaries of the modern marina, it would not be proper to expand the marina to the eastern basin, as previously planned. The chain that blocked the harbor entrance: The chain that blocked the harbor entrance during the Early Islamic and Crusader periods is another debated issue, because it is associated with the layout of the harbor and the location of the entrance to the harbor in these periods. Muqaddasi’s writings allude to the closing of the harbor at night by a chain pulled across its entrance, as at Tyre. In 1047 Nasir a-Khusraw wrote of a “chain that is lowered until it sinks in the water and when the ship enters the harbor, and thereafter it is stretched to prevent any ship from entering the harbor and attacking ships anchored there.” Based on Khusraw’s description, Gertwagen (1996) and Kedar (2013) deduced that the harbor entrance was respectively 30m or 30.5m wide. A few hundred years later, in 1258, the Genoese built a temporary floating obstacle made of thick wooden beams until a permanent iron chain was installed (Gertwagen 1996). Some scholars

suggest that the chain closed the western basin by being stretched from the eastern end of the southern breakwater, northward to a round tower on the shore next to the northeastern corner of the harbor, a distance of ca. 100m (Fig. 10: a) (Raban 1983a, 1993; Flinder et al. 1993; Jacoby 1979). Gertwagen (1996) suggested that it stretched from the Island of Flies westward to a now submerged structure at the eastern end of the southern breakwater (a distance of ca. 120m) (Fig. 10: b). She noted that the round tower is post-Crusader and could not have served as the chain’s tower. Moreover, Gertwagen (1996) notes that the 30m width of the harbor entrance described by Khusraw, does not fit the two theories proposed above, in which the entrance to the harbor was estimated at 80–120m wide. The use of chains for blocking harbors beginning in the third century BCE is extensively discussed by Kedar (2013). He cites written sources mentioning the use of iron chains during the Early Islamic period in the Eastern Mediterranean harbors of Akko, Tyre and Beirut. Various documents and drawings, mentioning and depicting the use of chains for blocking harbors in the Mediterranean during later periods, are presented as well (Kedar 2013). In some places remains of such chains were preserved and are displayed (Fig. 11). After the founding of the Frankish Kingdom of Jerusalem in 1099 the harbors of Akko, Tyre and

Fig. 11. An original iron chain that blocked the entrance to the Harbor of Pisa, on display (E. Galili) Beirut were conquered by the Crusaders. The Crusaders used the harbor chains which had been installed by the Muslims. This is mentioned in written documents as early as 1104–1124 (Kedar 2013: 7– 9). Stretching a heavy iron chain across a ca. 100m-wide gap is not a simple engineering task even today. To operate such a chain and stretch it in the air, or in the water close to the surface to an almost horizontal position, a large powerful winch built on a massive, elevated foundation is needed at 331

Ehud Galili and Baruch Rosen

Assuming that a human can generate drought of ca. 1/10 of a horse’s draft, which is about 10kg (White 1984: 129; Hopfen 1969: 10–11, table 2), some 150 men or 15 horses were needed to operate such a winch. Operating a chain of 30m long, as deduced from Khusraw’s description, would have been a much easier task. A 30m-long iron chain of the same characteristics requires ca. 1700kg of traction to be lifted in the air over the entrance. To do this with the same winch, a force of only 425kg, generated by four horses was needed, or less if part of the chain was in the water. It should be noted that the two above mentioned harbor entrances are considerably wider (100–120m) than the widths deduced from Khusraw’s description. Blocking the entrance to the harbor by a chain that is submerged required a reduced force (see above). Closing the harbor by means of floating wooden barges, connected with a chain, was more practical than closing it by a suspended chain. Thus, it is proposed that the iron chain of the Akko Harbor, or part of it, was floated on a series of barges or logs (Fig. 13).

one end of the chain. The other end of the chain must be anchored to another massive, elevated structure. During the Early Islamic period, when men and/or animals had to pull such chain, it is unlikely that a winch that could stretch a heavy iron chain over such a long distance would have been practical. However, lowering and lifting a shorter chain of few dozen meters each time a ship entered the harbor (as described by Khusraw in Akko in 1047) would have been possible. Calculations of the tensions created by the chains blocking the entrances to the harbors of Palermo (220m wide) (Zorić 1996: 81–84) and the Golden Gate in Constantinople (750m wide) (Pryor 2007: 381–382) suggest that the chains could not have withstood the tension over such distances. Zorić and Pryor (ibid.) concluded that these chains must have been floated by some means. Similar chain floated on barges blocked the Hudson River near West Point, New York, in North America (Fig. 12). These assumptions are supported by written documents mentioning big blocks of wood used for

Fig. 12. A chain floated on barges blocking the Hudson River near West Point, New York, in North America (modified after Wikipedia and Boynton 1864 ) Fig. 13. Possible reconstruction of the iron chain in the Akko Harbor, floated on a series of barges (E. Galili)

floating the chains. Our calculations regarding Akko indicate that an iron chain 100m long weighing some 20kg per meter would have weighed ca. 2000kg. Assuming that such a chain was suspended in the air over the entrance to the harbor, with at a reasonable angle of 20 degrees from the horizon at its ends, it would bear tension of ca. 6 tons (6000kg). At an angle of 15 degrees the chain would bear tension of 11,500kg and at an angle of 12 degrees it would bear tension of ca. 28,600kg. Suspending the chain, or any part of it, would have lightened the load on the whole system by up to 1/8 (ca. 12.5%). It was calculated that to lift such a chain in the air and close the entrance, a winch (capstan) generating more than 6 tons of traction had to be operated at one end of the chain. Given that the traction was generated by men or animals, a winch having a wheel 1m in radius and arms 4m long could reduce the traction needed to 1500kg, or less, if a part of the chain was in the water.

The Crusader Period Most Crusader remains were found in Areas A, B and C, near the entrance to the western basin and only a few were found in Areas D, E and F within that basin. Among the ceramics discovered in the harbor and attributed to the Crusader period are 50 glazed bowls (some with engravings made after firing), 24 cooking vessels and 31 amphorae. Most of the bowls (47) discovered were imported, mainly from Cyprus and the Aegean; only two were local. The amphorae were also mainly imported (28 imported; 3 local). The origins of the cooking vessels, in contrast, differ: most (75%) are local, and the rest are of Cypriot origin. A comparison of these findings with those

332

The Akko Marina Archaeological Project – Summary

from excavations at the Akko Courthouse and in the Hospitaller complex showed that the ratio of imported to local cooking vessels is similar in both locales. However, on land, the vessels were of Syrian types, not found in the marina, possibly because they arrived at Akko overland through trade and internal import (Galili et al. this volume b). The same phenomenon is true of thirteenth-century Aegean ceramics – more were found in the marina than on land. These vessels seem to have been used by ships’ crews, whereas the imported vessels found on land were used for commerce and marketing (Galili et al. this volume b). From a chronological point of view, the findings in the harbor represent the twelfth and thirteenth centuries equally. In contrast, findings from land excavations represent limited and clearly defined periods. In the thirteenth century there are more imported vessels in the marina (mainly from Cyprus) than in the twelfth century, apparently because there was very little import to the Crusader Kingdom of Jerusalem in the twelfth century. The land excavations at Akko revealed special, locally produced unglazed vessels, which constituted about half of the ceramic findings. These vessels were absent in the ceramic assemblage recovered from the marina. This may indicate that they were made for local domestic use and were not exported or used aboard ships (Galili et al. this volume b). Historical sources indicate that during the Crusader period, Akko was the main harbor of the Kingdom of Jerusalem. There are numerous written testimonies to its military, economic and civilian importance. Findings from recent years in land excavations support the historical evidence (Edelstein and Avissar 1997; Pringle 1997; Stern 1997). Therefore, it is rather surprising that the Crusader remains found on the harbor bed are relatively meager. The contradiction between the written evidence and the finds in the marina may be explained by various processes at the site after the Crusader period. Perhaps, as with the remains from the Early Islamic period, the deepening of the marina during the 1970s and 1980s removed much of the fill strata containing the Crusader remains. Another possibility is that during the Crusader period the western basin was partially blocked by sand and silt. Consequently, except for small boats, most maritime activity took place in the yet un-excavated eastern basin, where the Crusader remains may be overlaid by a thick layer of sand.

1983b) and Flinder et al. (1993) have suggested that an inner harbor existed beneath Khan al-Umdan, with a separate exit southward to the sea. Underwater surveys opposite the Pisan Quarter yielded an almost continues barrier of huge boulders, the remains of the southern Hellenistic-Roman breakwater (see below). Thus, there could not have been a southern entrance to an inner harbor in that area. An inner harbor under Khan al-Umdan: Based on Sivan (1981), Gertwagen suggested (1996: 561) that there could not have been an inner harbor in the area of Khan al-Umdan because the Khan is constructed on solid kurkar rock. However, Sivan noted (1981: 58) that no core drillings were available for the area under the Old City. She mentioned the presence of a low (1.5m high) kurkar ridge under the west side of the Old City. However in a map, depicting the upper surface of the kurkar rock, its elevation under the southeastern area of the Old City, including the Khan al-Umdan area, is ca. 4–2m below sea level (Sivan 1081: fig. 17). Additionally, recent excavations in Venice square, adjacent to Khan al-Umdan (See above) revealed no rock there. Thus, the area under the Khan may have been an inner part of the eastern basin of the harbor. The tunnel (mentioned above) discovered under the city leading from the Khan al-Umdan area to the Templar Quarter on the western shoreline (Avissar and Stern 1996) (Figs. 10: d, 14, 15) could have provided an easy access from the inner port to the Templar quarter. The tunnel is now flooded by ground water mixed with sea water, to a height of ca. 70cm. Given the lack of core drillings in this area (see above) the tunnel provides a rare opportunity to gain information about the sub-surface of the Old City. The lower part of the tunnel wall is cut in the natural kurkar to a height of up to 3.2m above sea level (Figs. 14–16). It is evident that the surface of

b

c

A separate Pisan harbor in the Pisan Quarter? The city plan drawn by Pietro Vesconte around 1320, possibly based on earlier drawings by Marino Sanudo, may provide some evidence regarding Crusader Akko and its harbor shortly before the Mamluk conquest of the city in 1291. Some versions of this plan show a round or elliptical formation, open to the sea, at the Pisan Quarter sea front (Galili et al. this volume a). Linder and Raban (1965), Raban (1982; 1983a,

a Fig. 14. The Crusader tunnel: (a) the lower section of the channel cut in the natural kurkar stone; (b) one of the vertical grooves cut in the natural rock intended for holding wooden partitions used for sealing sections of the channel to enable quarrying under sea level; (c) the upper stone built arch, looking east (E. Galili) 333

Ehud Galili and Baruch Rosen

used for drainage and as sewage disposal. Thus, the Khan al-Umdan area could have been part of an inner harbor during the Crusader period. However, it was probably connected to the main harbor, as seen in city plans (Galili et al. this volume a), rather than being a separate Pisan harbor with a direct access to the sea. Future archeological excavations in the khan area could shed light on these issues. The southern breakwater during the Crusader period: A plan of the Akko Harbor dated to 1323 and drawn by Paulino Veneto, a contemporary of Pietro Vesconte, depicts a tower on the southern breakwater, connected to the shoreline by a number of hatched lines (Jacoby 1979: fig. 2) (Galili et al. this volume a, fig. 10: A). According to Jacoby (1979: 14), this line was intended to represent a chain preventing the entry of vessels to the harbor. However, Jacoby notes that a chain at this point does not make sense. It is proposed that the broken line marks a line of rocks remaining from the ancient breakwater. Apparently, during the Crusader period the southern breakwater of Akko was partly breached, with some sections submerged and others protruding from the water. This proposal is supported by a thirteenthcentury description of the harbor entrance: “The entrance to the harbor is between said tower [the Tower of Flies] and the rocks” (Motzo 1947). These rocks were apparently the remains of the ancient southern breakwater (Galili et al. this volume a).

Fig. 15. The Crusader tunnel flooded with water to a depth of 0.5m, looking east (E. Galili)

Numismatic evidence of the Mamluk conquest of Akko: A hoard of florins minted shortly before the city’s fall in 1291 near the modern harbor entrance (Kool this volume) affords a tangible glimpse into the dramatic events in the city at that time. According to an anonymous eyewitness, who was in the area of the harbor on May 18, 1291, after a long siege, the Mamluk army invaded Akko. Groups of Christian soldiers and civilians gathered in the harbor, attempting to flee by using their jewelry and gold to bribe the owners of small rowboats to ferry them to ships anchored off shore, in the hope of reaching Cyprus. The writer notes that many of those fleeing Akko drowned. Judging by the dating of the latest coin in the hoard, these gold coins may have fallen into the sea and sank during that event (Kool this volume).

the kurkar rock was leveled and the ashlars building the arch were placed atop the natural rock. Vertical grooves were cut in the rock on the tunnel’s sides (Figs. 14, 16), apparently to hold wooden partitions used to keep the water level low during the tunneling. The tunnel may have been intentionally dug partially underwater to create a shallow waterway across the entire densely inhabited area, enabling the transport of heavy supplies from the harbor to the Templar Quarter on floating barges. It could also have been

The Mamluk and Ottoman Periods Soon after the fall of Akko in 1291, the Mamluks, fearing the return of the Christians, razed the city and it remained in that state for hundreds of years thereafter (Ashtor 1983). Akko’s harbor installations were probably in poor condition after the city’s destruction by the Mamluks. A Turkish source describing Palestine’s coastline during the early sixteenth century states that Akko, once a great city, had been destroyed and replaced by a village,

Fig. 16. A vertical groove cut in the natural kurkar stone – detail (E. Galili) 334

The Akko Marina Archaeological Project – Summary

opposite which was a large tower guarding the harbor. The source notes that large ships anchoring outside the harbor, between the Island of Flies and the shoreline, moor by a rope to the island (Heyd 1956). Nevertheless, the presence of merchants, pilgrims and travelers in the harbor was attested to after the fall of Akko, despite its supposedly ruined state. Cotton was exported from the Akko Harbor by Venetian merchants (Ashtor 1983: 379). Ceramic vessels arriving with Italian merchants during the Mamluk period were discovered at Tel Yasaf (Stern 1999). These vessels may have been associated with the cotton trade and the continuing commercial ties with Venice (Galili et al. this volume b). Additionally, abandoned buildings in the city were used as cotton warehouses and the ruined harbor was converted for the transport of cotton. The finds recovered from near the entrance to the marina provide evidence as to the maritime activities in the harbor. Southeast of the entrance to the marina a construction built of wooden pillars embedded upright in the seabed was recovered (Galili and Rosen this volume a). These were apparently part of a quay or a pier used to moor sailing vessels during loading and unloading of merchandise e.g., cotton, and passengers. The structure was dated by C14 to the fifteenth century. The western basin was apparently partially silted at that time and most activity took place in the eastern basin. Most probably, large sailing vessels with deep drafts anchored outside the harbor and goods and passengers were transported by lighters. Hundreds of glazed vessels and decorated clay pipes from the Ottoman period discovered in the marina (Galili et al. this volume b) provide evidence for extensive maritime activity during the Ottoman period, showing trading ties with the eastern Mediterranean coasts: Turkey, Syria, Lebanon and Egypt. The archaeological finds from the marina and Tel Yasaf support the historical descriptions showing that the harbor continued to operate, in one way or another, after the Crusader period. The southern breakwater lay in ruin in the seventeenth century. During that time, medium-size sailing vessels anchored in the eastern basin. Large vessels anchored near the Island of Flies in deeper water. Remnants of large buildings on the breakwater are depicted in nineteenth-century drawings and photographs of Akko (Galili et al. this volume a). These buildings suggest that the southern breakwater was renovated during the eighteenth and nineteenth centuries. Based on the Sanudo drawings, the historical sources and the archaeological finds, it seems that maritime activity in Akko continued after the Crusader period. Napoleon’s siege of Akko: Maps of Akko and its surroundings from the eighteenth and early nineteenth centuries show that the eastern basin was used at that time for anchorage and as a landing place (Galili et al. this volume a). Shipwreck remains

discovered near the Island of Flies were dated to the period of Napoleon’s siege on Akko by their timber and the construction technique (Linder and Raban 1965; Galili this volume f). Raban concluded that the ship was intentionally sunk by the British during the French siege of Akko in order to block the harbor entrance. Cannon balls and rifle bullets found on the seabed support the historical descriptions of the battles that raged in the harbor area and the shelling of the city from ships at sea (Galili and Rosen this volume d).

The Layout of the Harbor in Various Periods The harbor bed constitutes a closed basin in which sediments and debris accumulated over thousands of years. Remains of materials lost accidentally or intentionally dumped overboard were stratified on the sea floor. Such artifacts may provide valuable information about the chronology of the harbor’s construction and operation. They also tell us about the type of activities conducted there and the harbor’s connections with various Mediterranean regions. The marina dredged during the current project was divided into six areas (Fig 8: a, for details see Galili this volume a): Area A was the largest, located in the western part of the eastern basin; Area B was north of the entrance to the western basin; Area C was at the entrance to the western basin; Areas D, E, F were within the western basin. Half of the barges that deposited material into the sea during the deepening of the marina removed material dredged from the western basin (38 barges from Areas D, E and F as opposed to 19 barges from Area A, and ca. 24 barges from Area C). It may be reasonably assumed that the large quantity of material removed from the western basin increased the relative quantity of finds from that area. Contemporary sediments were discovered on the surface of the western basin. These accumulated in recent decades and included silt, modern debris and Ottoman remains that became mixed with contemporary strata during the dredging operations of the harbor before the 1990s (Fig. 17). During the dredging operations carried out from 1992 to 1993 the dredger dug deeper and reached previously undisturbed areas. As a result, the archaeological remains recovered from these lower levels provided valuable information as to the history of the harbor (Fig. 17). The extent of the harbor during the various periods is presented below according to location and time perspective. The western basin: Silt sediments removed from a depth of more than 4m in the western basin contained numerous archaeological remains from the Hellenistic and Roman periods (Zilberstein et al. this volume).

335

Ehud Galili and Baruch Rosen

Southwest

A

East

I

Fig. 17. Schematic cross-section E-E of the archaeological remains and sediments in the Akko Harbor, for location of section see Fig. 10 above (E. Galili) record. Numerous artifacts dating to the Ottoman period were recovered in the western basin despite previous dredging that removed much of the upper strata (Galili et al. this volume b).

Such sediments are typical of closed and protected sedimentation basins, suggesting that at that time, the western basin was protected by the breakwater and was at least 3m deep, thus enabling the anchoring of medium-size and even large vessels. The finds show that the harbor was built at that time and that extensive maritime activity, typical of working harbors, seems to have been underway there in those periods. The finds retrieved from the western basin and those obtained by the recent salvage excavation at the foot of the southeastern sea wall (Sharvit et al. 2013: their areas F, C) make it possible to locate the approximate layout of the Hellenistic-Early Roman harbor (Fig. 7: c). Based on the ceramic findings, the southern breakwater was built during the Hellenistic period in the first half of the second century BCE. Roman ship remains found in Area E (Kahanov this volume) show that medium-size ships entered the harbor during that period. The materials removed by the dredger in the marina entrance and east of it (Fig. 8 Areas A, B, C) yielded only meager finds from these early periods, compared to the large quantities of finds discovered in the western basin. Finds from the harbor and written sources suggest that anchoring and mooring conditions in the western basin during the Byzantine period were poor, probably due to improper maintenance and partial destruction of the breakwater (Galili and Rosen this volume b). Ottoman remains were found in the western basin near the seabed surface, while remains from the Early Islamic and Crusader periods were almost absent there. It seems that during these periods the western basin was filled by sand and was usable by small boats only. Obviously, intensive activity in the western basin during the Early Islamic and Crusader periods would have been reflected in a richer archaeological

The Island of Flies: The artificial foundations of this island (Flinder et al. 1993: 216-222; Galili this volume e) were probably laid during the Hellenistic period, when the southern breakwater was constructed. Because the island is fairly small, it is not likely that it served as an emporium, as proposed by Linder and Raban (1965). More probably it housed a fortified structure protecting and monitoring the harbor entrance. It may have also served as a navigation aid and a lighthouse (Rosen et al. 2011; Galili et al. this volume a). In early periods ships may have anchored in relatively deep water (7–8m deep) south and west of the island, an area affording some shelter when northern winds prevailed. In contrast, when western and southwestern winds prevailed, the island was exposed to waves and did not provide a safe haven. The eastern basin: As noted above, the seabed in the eastern basin, overlaid by a 5m-thick sediment layer, was not excavated other than in its western section near the entrance to the marina (Areas A, B and C). The fill that was removed from this area is essentially different from the material removed from the western basin and includes mainly coarse carbonatic sand and fine-grained quartz sand. These sediments indicate relatively powerful wave energy, suggesting that this area was less protected than the western basin. Most of the eastern basin has always been more exposed to winds and waves from the south and the southwest. Even today water craft cannot safely anchor there during southwestern storms. Although most of the

336

The Akko Marina Archaeological Project – Summary

eastern basin has not been deepened or dredged, based on the latest excavations at the foot of the sea wall (Sharvit et al. 2013), it may be assumed that during the Hellenistic and Roman periods the activity in the eastern basin was limited to the more protected areas, in its western part. The eastern part of this basin, which was more open and less protected, may have served for anchoring during calm sea conditions. Scholars disagree about the location of Ibn Tulun’s harbor of the Early Islamic period and its location remains unclear: Was it in the western basin, the eastern basin, or both (see above). According to Linder and Raban (1965), Ibn Tulun’s harbor was in the western basin. In Raban’s opinion the eastern rampart was built to prevent the penetration of sand into that harbor. Gertwagen (1995: 573–575) suggested that during the Early Islamic and Crusader periods the harbor extended from the eastern rampart westward. The scant remains from the Early Islamic and Crusader periods discovered in the western basin could indicate that maritime activity there was limited during these periods. That state of affairs may support the possibility that the maritime activity during these periods was in the eastern basin. Possibly, remains from these periods are hidden beneath the sand in the eastern basin. Written sources (Gertwagen 1995: 573–575) and graphic depictions, beginning in the Crusader period, reveal that the eastern basin west of the eastern rampart, and apparently the western basin as well, were shallow (as a result of partial siltation) and large ships could not anchor there at that time (Galili et al. this volume a). Most remains found near the entrance to the marina (Areas A, B and C) were dated to the Byzantine and Ottoman periods, with a small percentage dated to the Crusader period. Apparently during these periods, the area functioned as an anchorage. The remains had not been disturbed by the previous dredging of the harbor. Harbor plans and charts from the seventeenth– eighteenth centuries reveal that at that time large ships could anchor only near the Island of Flies and at the harbor entrance. The layout of the harbor seen from a chronological perspective: Most of the findings from the Hellenistic and Roman periods were discovered, as noted, in the western basin, while remains from the Byzantine period were numerous at the marina entrance (Silberstein et al. this volume). Thus, the distribution of archaeological remains shows a tendency toward anchoring in the well-protected western basin during the Hellenistic and Roman periods, when the built harbor functioned (Areas D, E and F and the relatively protected western section of the eastern basin). During the Byzantine period, the anchoring zone retreated eastward toward the entrance of the marina and the more protected areas in the Eastern basin (Areas B and C). Finds from the Crusader period (Stern 2013; Galili et al. this volume

b) and a fifteenth-century Ottoman mole made of wooden piles were discovered near the entrance to the western basin. These finds suggest that during the Crusader and Ottoman periods the anchoring and mooring activities moved toward the eastern basin (Area A). The above findings reveal changes over time in conditions in the western basin that prevented ships with deep drafts from anchoring and mooring safely there. These changes probably led, over the years, to the shifting eastward (to the eastern basin) and southward (to the open sea) of the anchoring zones and activities associated with large sailing vessels. Sea-level Changes and Neo-Tectonic Activity in the Akko region. The Akko headland is the northern end of the Haifa Bay and the northern boundary of the Nile littoral cell (Zviely et al. 2006). Southeast of the headland is the Akko Bay, providing a small natural anchorage protected from northwestern and western storms, a rare feature on the coast of Israel (Fig. 10). Archaeological and geological investigations shade light on the coastal and sea-level changes in the region during the Holocene. It has been suggested that during historical periods there were significant fluctuations in relative sea level at Akko. Scholars suggested that during the Hellenistic, Early Islamic and Crusader periods, sea level was 0.5–1.6m below the present level and during the Byzantine period it was significantly higher (Raban 1982:185; 183a; 1993b; Neev et al. 1987; Flemming et al. 1978; Gertwagen 1996; Nir 1997a, 1997b; Sivan 2010). Changes in sea level in the harbor area can affect the functioning of installations and are therefore of interest. The basic principles of using sea markers were discussed and summarized by Galili et al. (2015). By using these principles and in light of the new finds, the possibility of sea-level changes in the Akko region is reassessed and discussed. The principles of using sea-level markers: Coastal archaeological features are often associated with sea level to enable proper functioning. Geological deposits and erosional features are similar to archives and may preserve the location of the coast and sea level at the time of their formation. Thus, studying archaeological and geological features may provide invaluable information on paleo sea level and vertical earth crust movements. Some markers show the highest possible sea level, others indicate the lowest possible level; still others reveal both the highest and lowest possible levels. Obtaining the ancient sea level using archaeological features is usually done by dating, measuring their elevation relative to current sea level, defining their function and their association with the sea and determining their potential use as

337

Ehud Galili and Baruch Rosen

Underwater research (Galili this volume e) showed that the arch was originally built partly underwater, as a part of the sea wall foundations constructed over the remains of the ancient breakwater (Figs. 18, 19). This arch spans the gap between two large blocks of the ancient breakwater (several meters long each) and was used as a foundation for the wall. In antiquity, building structures under the water was common in coastal sites. Examples of such practices include the foundations of the Island of Flies, the construction of Akko’s southern breakwater and the construction of the breakwaters at the Atlit and Caesarea harbors. Thus, finding the foundation of a building under the water does not necessarily indicate a rise in sea level or tectonic subsidence.

sea-level markers (Blackman 1973; Flemming 1969; Morhange et al. 2006; Auriemma and Solinas 2009; Evelpidou et al. 2012; Antonioli et al. 2006; Galili and Sharvit 1998; Sivan et al. 2004; Galili et al. 1988, 2005, 2015 and references therein).

The foundations of the southeastern Ottoman sea wall: The southwestern and western Ottoman sea walls of Akko were founded directly on the coastal kurkar ridge. The kurkar bedrock was leveled to an elevation slightly higher (20–30cm) than the current sea level (Fig. 20). In the southeastern and eastern sections of the city (from Khan al Umdan to the east) there is a trough and the kurkar bedrock surface lays under the present sea level. The sea bottom opposite (south of) the Pisan Quarter sea front, and to the east, consists of poorly consolidated, coarse, marine and brackish sediment that filled the marshy trough east of the coastal kurkar ridge (Figs. 7, 19). Thus, the Hellenistic-Roman southern breakwater

Fig. 18. Relict of the city wall and a partly submerged stone-built arch in the Pisan Quarter sea front, looking southwest (for location see Fig. 19 below, Modified after D. Zviely). A submerged arch in the Pisan Quarter used as a sea-level marker: South of the Pisan Quarter, a built arch, part of the sea wall, stands partly submerged in the water (Fig. 18) (Galili this volume e). It was suggested that the arch can be used as a marker for sea level lower by 1–2m during the Crusader period (Gertwagen 1996; Flemming et al. 1978).

Sea front of the Pisan Quarter The Pisan Quarter

(See See Fig. 18 above) above

Fig. 19. The relict of the city wall founded on the ruins of the Roman breakwater (E. Galili) 338

The Akko Marina Archaeological Project – Summary

were installed across it. Thus the lower part was intentionally excavated below sea level, to enable flooding and transportation of goods aboard small rafts. Thus, the currently flooded tunnel cannot be used as a reliable marker for sea level lower than today, during the Crusader period, as previously suggested (Sivan 2010).

Fig. 20. Foundations of the southern Akko sea walls built on a solid rocky platform, looking west (E. Galili) and the later harbor constructions and sea walls in this areas were mostly built on poorly consolidated, unstable sediments (see above). Rescue excavations (2008–2013) at the foot of the southeastern Ottoman sea wall, aimed at repairing its foundations and filling voids and missing stones caused by marine erosion (Sharvit et al. 2013) revealed the foundations of that wall. These were found to have been built of boulders and columns in secondary use made of granite, marble and kurkar, lying horizontally perpendicular to the wall. The columns overlay a Hellenistic pavement at elevation of up to 1–2m below sea level. In other section, the wall was founded on a row of 2.5m long wooden piles driven into on the kurkar bedrock some 3m below sea level. The piles bridge the gap between the lower part of the wall and the kurkar bedrock (Sharvit et al. 2013: 43, figs. 4, 5, 7 respectively). Measuring the height of the Ottoman sea wall at its southwestern sections showed that the top of the wall, which was founded on the kurkar substrate at this area, is ca. 9.8m above sea level, while the top of the sea wall in the eastern section is only 7.7m above sea level. The ca. 2m difference may be intentional and represent the original elevations of the walls. However, it is possible that the southeastern sea walls underwent settling due to the poorly consolidated sediments on which they were founded. The Hellenistic masonry found ca. 1–2m below sea level may have originally been built underwater at that elevation, similar to other Hellenistic structures (e.g. the foundations of the Island of flies and the southern breakwater) (see above). Thus, these archaeological finding cannot be used as reliable markers for lower sea levels in these periods. The Templar Tunnel as a sea-level marker: The tunnel leading from the area of Khan al-Umdan to the western shoreline (see above) is now flooded by brackish water to a height of ca. 60cm (Avissar and Stern 1996) (Figs. 14-16). The vertical grooves on the tunnel’s sides suggest that wooden partitions

Rock-cut installations and coastal quarries in Akko used as sea-level markers: Coastal kurkar quarries north of Akko were a convenient source of building stones and millstones (Galili and Sharvit 2002; Sharvit and Galili 2002; Antonioli et al. 2017). They were intensively utilized to supply increasing demands for building materials. The rock was quarried down to sea level and even slightly below it, taking advantage of low tide conditions. Rock-cut pools fed with sea water by gravity were discovered along Israel’s coast. Rock-cut channels on the abrasion platforms convey the sea water and are still functioning today at current sea level, indicating that no major sea-level changes have occurred since the time these pools were in use. The bottom of some of these pools is hewn to as much as 1.5m below sea level, suggesting that it was possible to quarry in shallow water if needed (Galili and Sharvit 1998). However, no rock-cut pools or quarries have been discovered in deeper water (more than 1.5m), either in Akko or anywhere along the Israeli coast. If sea level were 2m lower that the current level, as proposed by Gertwagen (1996), it would be expected that quarries as low as that level would have utilized the kurkar rock, which was in high demand in Akko. Since the elevations of the pools and quarries in Akko region fit the current sea level it is concluded that no fundamental change in sea level beyond the local tidal range has occurred in Akko in the last two millennia (Galili and Sharvit 1998, 2002; Sivan and Galili 1999; Galili et al. 2010; Antonioli et al. in press). Coastal water wells in Akko as sea-level markers: A fresh-water well dug near the coastline can sometimes be used as a sea-level marker at the time it was dug (Galili and Nir 1993; Nir and Eldar 1986). The farther inland the well is, the less precise a sea-level marker it is. Also, the more intensive exploitation of ground water in the region is, the lower the ground water level may be. A Crusader well discovered in the Knights Halls in Akko indicated that sea level at that time conformed, approximately, to the current level (Galili and Sharvit 1998). Another well from Akko located ca. 500m from the shoreline places sea level about 1m lower than the current level during the thirteenth century (Nir 1977a, 1997b, 1997c). Another well, dated to the Ottoman period, could indicate that sea level at that time was 2m lower than today (Nir 1997a, 1997b, 1997c). However, in Nir’s opinion, the depth of the well could indicate late use

339

Ehud Galili and Baruch Rosen

(during the nineteenth century), when a pump could have been employed. Over-pumping of water from the aquifer, in the area of the Akko peninsula, when it was densely populated, could have lowered the ground-water level. Therefore, to reach fresh water, the well had to be deepened below sea level. He also suggest that the low water level in the wells could be the result of gradual subsidence of the of the Akko peninsula (Nir 1997a: 33). Given the unknown influences on their depth, these wells cannot serve as a reliable sea-level indicator.

wharf settling as well as waves pounding, deliberate distraction and improper maintenance. Tectonic stability and sea-level changes in Akko: The archaeological and other findings mentioned above indicate that no fundamental, vertical, changes in relative sea level have taken place over the last 2,000 years and that the Akko area was tectonically stable during this period. Nevertheless, sea-level fluctuations that are within the tidal range are hard to detect using archaeological markers, due to an inability to identify such small vertical changes and the inherent uncertainty of the values of these markers. Thus, the possibility of minor vertical fluctuations that are smaller than the local tidal range cannot be ruled out.

Stone-built structures and installations used as sealevel markers: The foundations of the southwestern and western Ottoman sea wall are laid on the abrasion platforms at an elevation conforming the current sea level (Fig. 20) (see above), indicating that at the time of construction, sea level was similar to the current one. Iron, u-shaped clamps discovered at a depth of about 0.50m at the Island of Flies have been suggested as evidence of a lower sea level during the Crusader period (Gertwagen 1996). However, such iron clamps can be installed underwater at such depth. Thus, they do not necessarily indicate relative sea-level changes.

Acknowledgments The authors wish to thank the Israel Antiquities Authority for the institutional support; D. Jacoby and S. Arenson for their useful remarks on the manuscript; Amani Abu-Hamid, Danny Syon, Eliezer Stern and Nimrod Getzov for the useful information associated with excavations in Old Akko; Howard Rosenstein, Fantasea, for his support in providing photography equipment and Google Earth and A. Ohayon for the aerial photos.

Water depth in the Akko Marina as a sea-level marker: The archaeological findings in the marina indicate that the seabed in the western basin was 3m deep or more, a depth allowing medium-size ships to moor there. That is attested by a Roman shipwreck identified in the harbor (Kahanov this volume). If sea level had been considerably lower at that time, such medium-size ships could not have entered the harbor and moored there. Roman shipkeel sheathing recovered from a known reef south of the harbor suggests that the ship’s draft was about 3m and its keel scraped the reef, which ripped the lead sheathing (Galili this volume f). Presently this reef endangers ships with a draft of 3m. If sea level in the Roman period were higher than it is now, as has been proposed (Raban 1983a, 1983b), the reef would not have scraped that keel.

References Antonioli F., Ferranti L. and Kershaw S. 2006. A glacial isostatic adjustment origin for double MIS 5.5 and Holocene marine notches in the coastline of Italy. Quaternary International 145–146: 19–29. Antonioli F., Mourtzas N., Anzidei M., Auriemma R., Galili E., Kolaiti E., Lo Presti V., Mastronuzzi G., Scicchitano G., Spampinato C., Vacchi M., Vecchio A. 2017. Millstone quarries along the Mediterranean coast: chronology, morphological variability and relationships with past sea levels. Quaternary

Geomorphological markers to determine sea-level changes in Akko: No abrasion platforms and/or geological notches in the Akko area were found on the seabed or above sea level. Also all the notches and abrasion platforms found on the Western Galilee coast conform to current sea level. The archaeological and geomorphological findings discovered in the Akko area indicate a relative stability (± 30cm) of sea level in recent millennia (Galili and Sharvit 1998; Sivan and Galili 1999). Thus, it is unlikely that changes in the function of harbor installations occurred as a result of tectonic subsidence and/or a rise in sea level, as has been suggested (Gertwagen1996; Raban 1983a, 1993b). It thus seems that changes in anchoring and mooring conditions and in the harbor’s functioning, resulted from sediments shifting, harbor siltation,

International.

Auriemma R. and Solinas E. 2009. Archaeological remains as sea level change markers: A review. Quaternary International XXX: 1–13. Avissar M. and Stern E. 1996. Akko, the Old City.

Hadashot Arkhaeologiyot 106: 20–21. (Hebrew).

Ashtor E. 1983. Levant Trade in the Later Middle Ages. P. 379. Princeton, N.J. Blackman D.J. 1973. Evidence of sea level change in ancient harbours and coastal installations. In B.J. Blackman (ed.). Marine Archaeology. Pp. 115–138. London.

340

The Akko Marina Archaeological Project – Summary

Boynton E. C. 1864. History of West Point, and its

military importance during the American Revolution: and the origin and progress of the United States Military Academy (With illustrations and maps) New York. Cohen S.L. 2002. Canaanites, Chronologies, and

Connections: the Relationship of Middle Bronze IIA Canaan to Middle Kingdom Egypt, Vol. 3. Winona Lake.

Dothan M. 1978. Notes and news. Israel Exploration Journal 28: 264–266. Dichter B. 1979. Akko: a Bibliography. Akko. Dothan M. 1993. Acco. In E. Stern (ed.). The New Encyclopedia of Archaeological Excavations in the Holy Land 1. Pp.17–24. Jerusalem. Edelstein G. and Avissar M. 1997. Excavation of the Courthouse Site at Akko: A Sounding in Old Acre. ‘Atiqot 31: 129–136. Evelpidou N., Pirazzoli P., Vassilopoulos A., Spada G., Ruggieri G. and Tomasin A. 2012. Late Holocene sea level reconstructions based on observations of Roman fish tanks, Tyrrhenian coast of Italy. Geoarchaeology 27: 259–277. Finkielsztejn G. This volume. Amphora Stamps from the Akko Marina. Flemming N.C. 1969. Archaeological evidence for eustatic changes of sea level and earth movement in the western Mediterranean in the last 2000 years. (Geological Society of America Special Paper 109). Pp. 1–98. Boulder, CO.

Flemming N.C. 1978. Holocene Eustatic Changes and Coastal Tectonics in the Northeast Mediterranean: Implications for model of crustal consumption.

Philosophical Transactions of the Royal Society of London, Series A, Mathematical and Physical Sciences 289, No. 1372. Pp. 405–458. London.

from the Akko Marina. Flinder A., Linder E. and Hall E.T. 1993. Survey of the Ancient Harbour of Akko, 1964-1966. In M. Heltzer, A. Segal and D. Kaufman (eds.). Studies in

Archaeology and History of Ancient Israel in Honour of Moshe Dothan. Pp. 199–225. Haifa.

Galili E. 1986. Historical Sea Level Changes and Seafaring along the Southeastern Mediterranean Coast. Proceedings of Cities on the Sea – Past and

Present, 1st International Symposium on Harbours, City Ports and Coastal Topography. Pp. 69–73.

Haifa.

Galili E., Weinstein-Evron M. and Ronen A. 1988. Holocene Sea-Level Changes Based on Submerged Archaeological sites off the Northern Carmel Coast in Israel. Quaternary Research 29: 36–42. Galili E. and Eitam Y. 1988. Young Faulting in the Shallow Continental Shelf: Evidence from Northern Israel. Proceedings of the Annual Conference of the Israel Geological Society, Ein Bokek. Pp. 30–31. (English/Hebrew). Galili E. and Nir Y. 1993. The Submerged PrePottery Neolithic Water Well of Atlit-Yam, Northern Israel, and its Palaeoenvironmental Implications. The Holocene 3: 265–270. Galili E. and Sharvit J. 1994. Classification of Underwater Archaeological Sites along the Mediterranean Coast of Israel: Finds from Underwater and Coastal Archaeological Research. In C. Angelova (ed.). Actes du Symposium International Thracia Pontica V: 1991. Pp. 269–296. Sozopol. Galili E. and Sharvit J. 1998. Ancient coastal installations and the tectonic stability of the Israeli coast in historical times. In I.S. Stewart and C. Vita-Finzi (eds.). Coastal Tectonics 146. Pp. 147–163. London. Galili E and Sharvit J. 2002. Millstone Quarry on the Coast of Akko, ‘Atiqot 46: 73–78. (Hebrew).

Flinder A., Linder E. and Hall E.T. 1993. Survey of the Ancient Harbour of Akko, 1964–1966. In M. Heltzer, A. Segal and D. Kaufman (eds.). Studies in

Galili E., Sharvit J., Dahari U., Bahat-Silberstein N., Finkielsztejn G., Stern E.J., Kool R., Kahanov Y. and Rosen B. 2002a. Akko Harbor, Underwater Surveys, Hadashot Arkheologiyot 114: 13–16, 12*–15*. (Hebrew/English*).

Flemming N.C., Raban A. and Goetschel C. 1978. Tectonic and Eustatic Changes in the Mediterranean Coast of Israel in the last 9000 years. In J.C. Gamble and R.A. Yorke (eds.). Progress in Underwater Science 3. Pp. 33–94. London.

Galili E., Raban A. and Sharvit J. 2002b. Forty Years of Marine Archaeology in Israel. In H. Tzalas (ed.). Tropis VII, Proceedings of 7th International Symposium on Ship Construction in Antiquity, Pylos, Greece, 1999. Pp. 927–961. Athens.

Friedman Z., Galili E. and Rosen B. This volume. Maritime Depictions on Roman/Byzantine Ceramics

Galili E., Zviely D. and Weinstein-Evron M. 2005. Holocene Sea-level Changes and Landscape

Archaeology and History of Ancient Israel in Honour of Moshe Dothan. Pp. 199–225. Haifa.

341

Ehud Galili and Baruch Rosen

Cartographic and Photographic Evidence.

Evolution on the Northern Carmel Coast (Israel). Mediterranée 1: 1–8.

Galili E. and Rosen B. This volume b. Anchors from the Akko Marina.

Galili E. and Rosen B. 2008, Acco – The Harbor. In E. Stern (ed.). The New Encyclopedia of Archaeological Excavations in the Holy Land 5. Pp. 1558–1561. Jerusalem.

Galili E., Rosen B. and Sharvit J. This volume b. The Fatimide, Crusader, Mamluk and Ottoman Ceramics from the Akko Marina.

Galili E. 2009. Ancient Harbors and Anchorages along the Israeli Coast after Five Decades of Underwater and Coastal Research. Qadmoniot 327: 2–21. (Hebrew).

Galili E. and Rosen B. This volume d. Weapons and Ammunition from the Akko Marina. Gambish G. 2013. Caesarea Maritima and the Grand Strategy of the Roman Empire. Skyllis 13.1.2013: 53–58.

Galili E., Rosen B., Zviely D., Silberstein N., Finkielsztejn J. 2010. The Evolution of the Akko Harbour and its Trade Links Revealed by Recent Underwater and Coastal Archaeological Research. Journal of Island and Coastal Archaeology 5: 191–211.

Grabois A. 1983. Akko and its Sites. Kardom 19–20: 24–35. (Hebrew).

Galili E., Sevketoglu M., Salamon A., Zviely D., Mienis H.K., Rosen B. and Moshkovitz S. 2015. Late Quaternary morphology, beach deposits, sea–level changes and uplift along the coast of Cyprus and its possible implications on the early colonists. In J. Harff, G. Bailey and F. Lüth (eds.).

Gertwagen R. 1996. The Crusader Harbor of Acre: Layout and Problems of Maintenance. In M. Balard (ed.). Autour de la Premiere Croisade, Actes du

Colloque de la Society for the Study of the Crusaders and the Latin East, Clermont-Ferrand, 22–25 juin 1995. Pp. 553–582. Paris.

Geology and Archaeology: Submerged Landscapes of the Continental Shelf (Special Publications 411). London. 179–218, doi:10.1144/SP411.13.

Hartal M. 1997. Excavations in the Courthouse site at ‘Akko’: The Architecture and Structigraphy in Area TA. ‘Atiqot 31: 22-23 table 1.

Galili E., Nir Y., Vachtman D. and Mart Y. In press. Physical characteristics of the continental shelves of the East Mediterranean Basin, submerged settlements and landscapes – actual finds and potential discoveries. In N. Flemming (ed.). Quaternary

Heyd U. 1956. A Turkish Description of the Coast of Palestine in the Early Sixteenth Century. Israel Exploration Journal 6: 201–216.

Palaeo-environments of the European continental shelf, environments for occupation and conditions for survival or destruction of submerged prehistoric deposits.

Hopfen H.J. 1969. Farm Implements for Arid and Tropical Regions. Rome. Jacoby D. 1979. Crusader Acre in the Thirteenth Century: Urban Layout and Topography. Studi Medievali, 3a Serie, 20, I. Spoleto.

Galili E. This volume a. The Akko Marina Archaeological Project – Introduction.

Kahanov J. This volume. Wood Fragments from a Roman Shipwreck.

Galili E. This volume b. Jewelry and Adornments from the Akko Marina.

Kashtan N. 1988. Akko-Ptolemais: a Maritime Metropolis in the Hellenistic and Early Roman Times, 332 BCE–70 CE, as Seen Through the Literary Sources. In I. Malkin and L. Holfelder (eds.). Mediterranean Cities: Historical Perspectives. Pp. 37–53. London.

Galili E. This volume d. Stone Finds from the Akko Marina. Galili E. This volume e. Structures and Installations in and around the Akko Marina. Galili E. This volume f. Shipwrecks and Cargoes in and around the Akko Marina.

Kedar B. Z. 1991. Looking twice at the Land of Israel, Aerial Photographs of 1917-1918 and 1987-91.

Galili E. and Rosen B. This volume a. A FifteenthCentury Harbor Installation with Wooden Pillars in Akko.

Kedar B.Z. 2013. Prolegomena to the World History of Harbor and River Chains. In R. Gertwagen and E. Jeffreys (eds.). Shipping, Trade and Crusade in the

Medieval Mediterranean: Studies in Honour of John Pryor. Pp. 3–38. Burlington.

Galili E., Zviely D. and Rosen B. This volume a. The Akko Harbor in Numismatic, Epigraphic,

342

The Akko Marina Archaeological Project – Summary

Kesten A. 1993. The Old City of Akko, Reconsideration, Criticisms. Akko. (Hebrew). Kool R. This volume. A Unique Hoard of Thirteenthcentury Gold Florins and Other Coin Finds. Linder E. and Raban A. 1965. Underwater Survey

in the Akko Harbor (1994), Western Galilee and the Galilee Coast, 19th Annual Israel Exploration Society Conference. Pp. 180–193. Jerusalem. (Hebrew).

Makhouly N. and Johns C.N. 1946. Guide to Acre. Government of Palestine, Department of Antiquities, second revised edition. Jerusalem. Morhange C., Pirazzoli P.A., Mariner N., Montaggioni L.F. and Nammour T. 2006. Late Holocene relative sea-level changes in Lebanon, Eastern Mediterranean. Marine Geology 230: 99–114. Morhange, C., Giaime, M., Marriner, N., abu Hamid, A., Bruneton, H., Honnorat, A., Kaniewski, D., Magnin, F., Porotov, A.V., Wante, J., Zviely, D., 2016. Geoarchaeological evolution of Tel Akko’s ancient harbour (Israel). Journal of Archaeological Science. Motzo B.R. 1947. II Compasso da Navigare. Annali della Facoltà di Lettere e Filosofìa della Università di Cagliari. Cagliari. Neev D., Bakler N. and Emery K.O. 1987.

Mediterranean Coasts of Israel and North Sinai. Holocene Tectonism from Geology, Geophysics and Archaeology. New York. Nir Y. 1997a. Excavation of the Courthouse Site at Akko: The Crusader-Period Well in Area TA. ‘Atiqot 31: 31–34. Nir Y. 1997b. Excavation of the Courthouse Site at Akko: The Turkish Well in Area TC. ‘Atiqot 31: 105–108. Nir Y. 1997c. Middle and late Holocene sea-level along the Israeli Mediterranean coast –evidence from ancient water wells. Journal of Quaternary Science 12: 143–151. Nir Y. and Eldar I. 1986. Ancient Ground-Water Table in Old Wells as an Indicator of Paleo Sea-levels and Neotectonic Changes along the Coastal Plain of Israel. Geological Survey of Israel Report 65/34/86.

Hierosolymitani. Studies in Crusades and Medieval Culture in Honour of Benjamin Z. Kedar. Pp. 369–

384. Aldershot.

Pringle D. 1997. Excavation of the Courthouse Site at Akko: Excavations in Acre 1974: The Pottery of the Crusader Period from Site D. ‘Atiqot 31: 137–156. Porath N., Sivan D. and Zviely D. 2008. Late Holocene embayment infill and shoreline migration, the Haifa Bay, Eastern Mediterranean. Israel Journal of Earth Sciences. 57: 21–31. Porat L. 2013. The south promenade. Hadashot Arkheologiyot 125. Raban A. 1982. The Harbors of Akko. In M. Yeda‘aya (ed.). The Antiquities of Western Galilee. Pp. 180–207. Tel Aviv. (Hebrew). Raban A. 1983a. The Biblical Port of Akko on Israel’s Coast. Archaeology 36.1: 60–61. Raban A. 1983b. Akko Harbor. Kardom 24–25: 141–146. (Hebrew). Raban A. 1985. The Ancient Harbours of Israel in Biblical Times. In A. Raban (ed.). Harbour Archaeology,

Proceedings of the First International Workshop on Ancient Mediterranean Harbours, Caesarea Maritima (BAR International Series 257). Pp. 11–44. Oxford.

Raban A. 1993. Maritime Akko. In E. Stern (ed.). The New Encyclopedia of Archaeological Excavations in the Holy Land 1. Pp. 29–31. Jerusalem. Rosen B. Galili E. and Zviely D. 2011. The Roman lighthouse of Akko. International Journal of Nautical Archaeology 41.1: 171–178. Sharvit J. and Galili E. 2002. Akko West – Coastal and Underwater Surveys, Hadashot Arkheologiyot 114:11–13, 10*–12*. (Hebrew/English*). Sharvit J., Planer D. and Buxton B. 2013. Preliminary Findings from Archaeological Excavations along the Foot of the Southern Seawall of Akko, 2008–2012. Michmanim 24: 39–40*, 39–52 (Hebrew/English*). Sharvit J. and Planer D. 2014. The southern sea wall. Hadashot Arkheologiyot 126 (Hebrew/English*) Shor N. 1990. The History of Akko. Pp. 161–163. Tel Aviv. (Hebrew).

Pp. 1–28. Jerusalem. (Hebrew).

Sivan D. 1981. Paleogeomorphology of the Akko Area in the Holocene. M.A. thesis, University of Haifa. (Hebrew).

Pryor J.H. 2007. The Chain of the Golden Horn, 5–7 July 1203 (with Peter R. Wilson). In I. Shagrir, R. Ellenblum and J. Riley-Smith (eds.). In Laudem

Sivan D. and Galili E. 1999. Holocene Tectonic Activity in the Galilee Coast and Shallow Shelf,

343

Ehud Galili and Baruch Rosen

Israel: A Geological and Archaeological Study. Israel

Journal of Earth Science 48: 47–61.

Sivan D., Lambeck K., Toueg R., Raban A., Porath Y. and Shirnam B. 2004. Ancient coastal wells of Caesarea Maritima, Israel, an indicator for relative sea-level changes during the last 2000 years. Earth and Planetary Science Letters 222: 315–330. Sivan D. 2010. What goes down must go up. Tehuda 21.1.2010. University of Haifa press release. Silberstein N., Galili E. and Sharvit J. This volume. The Hellenistic, Roman and Byzantine Ceramics from the Akko Marina. Stern E.J. 1999. The Pottery of the Thirteenth– Fifteenth Centuries from Giv‘at Yasaf (Tell Er-Ras). ‘Atiqot 37: 125–136. (Hebrew). Stern E.J. 1997. Excavation of the Courthouse Site at Akko: The Pottery of the Crusader and Ottoman Periods. ‘Atiqot 31: 35–70. Stern E. 2013. The Fatimid, Crusader and Early Ottoman Ceramic Finds From the `Akko Marina: Some Insights into Medieval Maritime Activity. ‘Atiqot 76: 139–168. Tatcher A. 2007. Akko, the Northern Moat, Final Report. Hadashot Arkheologiyot 119. Yurman A. 2013. The underwater rampart at Akko, its construction and function, an archaeological and geomorphological underwater research. M.A. thesis, University of Haifa. (Hebrew). Zorić V. 1996. La catena portuale. Sulle difese passive dei porti prima e dopo l’adozione generalizzata delle bocche a fuoco. Il caso di Palermo, con alcune noterelle di sua topografia storica. Schede medievali. Rassegna dell’officina di studi medievali 30–31 (gennaio–dicembre 1996): 75–108. Zviely D., Sivan D., Ecker A., Bakler N., Rorlich V., Galili E., Boaretto M., Klein M. and Kit E. 2006. Holocene Evolution of the Haifa Bay Area, and its Influence on Ancient Tell Settlements. The Holocene 16: 842–861.

344