During the summers of 1961-64, an expedition of the University Museum of the University of Pennsylvania under the direction of George F. Bass excavated the wreck of a 7th-century Byzantine ship that had struck a reef just off the small coastal island of Yassi Ada located between the Turkish mainland and the Greek island of Kos. The wreck lay at an ideal working depth of 32 to 39 m on a moderately steep but fairly even slope and appeared to be well preserved. Then visible was a relatively small amphora mound with a pile of iron anchors at one end and a well-defined area containing kitchen utensils and hearth and roof tiles at the other. (for more information, see Site & Excavation)

After an initial season of considerable experimentation with various available mapping techniques, angle-iron frames, measuring 6 by 2 meters and subdivided into three 2-m² areas, were erected like a series of steps up the slope on which the wreck lay. Each frame, supported by six pipe legs that also helped support the two adjacent frames, could be leveled and set as close to the seabed as possible. This frame system, in conjunction with stereophotogrammetry, yielded a highly accurate three-dimensional plan of the wreck.

Plan of the ship's hull as found on the seabed. (slide# YA7-571) Photo: INA.

An accurate three-dimensional plan of the wreck site proved indispensable to a reconstruction of the ship. When the ship sank, the keel and hull bottom on the port side had come to rest on a steep slope of exposed bedrock, with the bow pointing upslope toward the island of Yassi Ada. The ship then slid further downslope until the hull's after end began to dig into an area of deep sand. The hull's forward half remained resting on exposed bedrock, and for this reason did not survive. The bottom third of the sternpost, the after third of the keel, and roughly 40% of the after third of the hull bottom on the port side was well protected by sand and did survive. Most of the ship's port side up to deck level between amidships and the stern also survived, due to its breaking off the hull and falling into sand while still fairly intact.

Although it seemed unlikely that enough of the hull remained to permit an overall reconstruction, it was possible to make a reliable reconstruction on paper of the overall shape and dimensions of the hull's after half. The reconstruction would not have been possible had it not been for the fact that a highly accurate three-dimensional plan of the wreck-site seabed had been made that included the impression that the after part of the hull on the port side had made in the sand. J. Richard Steffy, best known for his reconstruction of the 4th-century BC Kyrenia ship, then made a long series of scale models designed to test my own reconstruction and project the after hull lines forward into the bow. In the end, only the shape of the very bow itself remained somewhat in doubt.

Interior view of the completed half model showing the hull planking, framing, deck, support stanchions, and galley area. (slide# YA7-640) Photo: INA.

The final model made was a half-model that shows the hull's overall shape and both it exterior and interior construction. (for more information, see Models & Replicas) The hull, designed primarily for speed, was notably narrow and streamlined, its maximum breadth being well aft of midships. Although nothing of the rigging survived, we can say that the ship would probably have performed best with a lateen rig. She was rather small, with an overall length of just under 21 m and a capacity of about 60 tons.

The hull had been built in a manner that is transitional between ancient Mediterranean shell construction, in which outer hull planking was fastened together edge to edge by pegged mortise-and-tenon joints, and "modern" skeletal building, in which some frames are erected before planking is begun, and the frames, not the planking, are the primary determiners of hull shape and the primary source of hull strength.

Hypothetical drawing of the Yassiada 7th century ship under construction. (slide# YA7-607) Photo: INA.

In the ancient Mediterranean world, hulls had been built to last as long as possible, despite the cost. The Yassi Ada hull, on the other hand, was built to last only long enough to make a good profit, with economy taking precedence over the ship's appearance. Wales girdling the sides of the hull for extra strength and a majority of timbers lining the hull interior were little more than half-logs. Mortise-and tenon joints that edge-joined the hull planking were much smaller and more widely spaced than they had been in earlier periods, and were now unpegged and used only up to the waterline. In earlier periods, frames had been fastened to hull planking by wooden trunnels and long nails driven through the trunnels to the inner face of the frames, where they were clenched. In this hull, planking was fastened to frames only by short nails that barely penetrated halfway into frames.(for more information, see Ship's Hull)

Model showing the location of the pile of spare anchors in the bow area (top) and the arrangement of the port bower anchors on the bulwark of the model (bottom). Photo: INA.

The ship was carrying 11 anchors when she sank. Seven were compactly stacked on the deck midway between the bow and midships, their shanks laying perpendicular to the vessel's longitudinal axis. Four bower anchors, ready for use, were on the bulwarks close by, two on either side. Three of the bower anchors had each weighed about 250 Byzantine pounds (1 pound=315 gm), or 78.75 kg. The other bower anchor was a best bower weighing about 350 Byzantine pounds (110.25 kg). A complete set of spares with the same weights lay uppermost in the anchor pile on the deck. The bottom three anchors in the pile, each weighing about 450 Byzantine pounds (141.75 kg), were evidently sheet anchors to be used as a last resort in stormy seas.(for more information, see Anchors)

The anchors have straight arms set perpendicular to the shank. Such cruciform anchors were used in the Mediterranean from the 4th to 10th centuries. The arms at their outer end curve upward and outward, terminating in a spade-like tooth, and near the top of the shank, there is a round aperture for the stock. The bower anchors had wooden stocks when the ship sank, but three iron stocks, stowed with the spare anchors, would have been used when greater anchor weight was desired. Each anchor had been made by hand-forging together at least 20 separate pieces of iron, and the cross-sectional areas of both shanks and arms were kept as small as possible in the interest of making strong welds. Since thin shanks occasionally broke, it was only prudent to carry a full set of spare anchors.

View of a full-scale replica of the ship's stern area. (Photo: INA)

The ship had an exceptionally well-equipped cooking and storage facility located in the stern. This galley complex is of such unusual interest that a full-scale replica of the stern part of the ship has recently been built at the Bodrum Museum of Underwater Archaeology in Turkey so that the modern public can visit the galley in person. (see Models & Replicas) Thanks to the fact that the find spot of everything of the galley and its contents to survive was recorded as precisely as possible, the basic location, size, layout and structural features of the galley as reconstructed are firmly supported by archaeological evidence.

The galley floor was set down low within the hull at a level that maximized the amount of space within the galley. A tile firebox with an grill consisting of movable iron bars occupied the port half of the floor. The hearth tiles had been laid in a matrix of clay reinforced by iron bars and probably enclosed by a wooden box. The galley structure rose far enough above the deck to allow for access and adequate lighting. When openings were closed in bad weather, smoke from the hearth could escape through a large circular hole in one of the tiles of the galley roof.

Some of the ship's utilitarian ware from the galley area. (slide# YA7-355) Photo: INA.

Preparation and cooking utensils, kept in the immediate vicinity of the hearth, included a mortar and pestle, 21 cooking pots in a variety of shapes and sizes, and two cauldrons and a bake pan of copper. (for more information, see Ceramics and Miscellaneous Objects) Food containers included at least 16 terra-cotta pantry jars. Also on the starboard side of the galley opposite the hearth stood the ship's water jar. Serving utensils included several copper or bronze pitchers, a glass bottle, 18 ceramic pitchers and jugs, a half-dozen spouted jars with lids, and four or five settings of a fine tableware, each consisting of a redware plate and dish, a glazed bowl and a one-handled cup. A ceramic pipette, or "wine-thief", was used to transfer wine or water from storage jars to pitchers. The coarse ware pitchers, jugs and jars were kept near the hearth and on the galley's after wall, but the more costly tableware and metal vessels were stowed with the food in the main storage locker, which was secured by a padlock. In this locker were also kept coins, weighing equipment, a tool chest and repair materials, spare lamps, and a bronze censer.

Sixteen gold and some 50 copper coins recovered from the wreck had all been kept within the main storage locker, the gold coins in one container and the copper coins in another. All the gold coins except one were issues of the Emperor Heraclius (610-641), and the latest of the copper coins dates to the year 625/626, indicating that the ship probably sank either in or shortly after 626.(see Miscellaneous Objects)

Bronze steelyard hooks, chains, beam, and Athena bust counterweight. (slide# YA7-448)

The weighing equipment consisted of a large steelyard; two small steelyards, one equipped with a balance pan; a set of silver-inlaid balance-pan weights; and a single glass weight. The large steelyard, with a weighing capacity of about 400 Byzantine pounds (126 kg), whose counterweight takes the form of a bust of the goddess Athena, may be the largest steelyard from antiquity to survive. The balance-pan weights constitute one of the most complete sets of Byzantine weights known.(for more information, see Weighing Implements)

The ship was well supplied with tools and materials for making necessary repairs during the voyage including axes, adzes, an awl, a bow drill and bits, billhooks (for rough trimming of wood), a carpenter's compass, chisels, files, gouges, hammers, knives, punches, one or more saws, a carpenter's belt for carrying tools, netting needles and spare lead weights for repairing fish nets, and an interesting assortment of lead weights and lures for deepwater line fishing. (for more information, see Iron Objects)

There were at least 16 spare lamps in the locker. Eight others in use were kept near the hearth. The ship's censer, whose open-work lid was surmounted by a cross on an orb, probably was used to sanctify both religious ceremonies and business agreements, as well as provide incense at meals.

Two examples of the ship's globular amphoras. (slide# YA7-317) Photo: INA.

The ship was carrying approximately 900 amphoras when she sank. Some 700 of these were globular jars with capacities ranging from 10 to some 40 liters; the rest were cylindrical jars with pinched waists that carried from 42 to 142 liters. The globular amphoras had been stacked three deep in the hold, and the cylindrical jars had then been placed between the necks of the top layer of globular jars. Only 110 of the amphoras were raised at the time of the excavation. (for more information, see Amphoras)

Since the amphora interiors were lined with resin, which prevented low-viscosity liquids from seeping through the walls, we concluded that the amphoras very probably had been carrying wine. About 165 amphora stoppers, terra-cotta disks cut from the walls of amphoras, were recovered from the wreck.

There were graffiti (carved inscriptions), often completely hidden by concretion deposits, on many of the raised amphoras. The graffiti appeared to be of sufficient potential importance to justify the raising of the amphoras still remaining on the seabed at Yassi Ada so that all the amphoras might be cleaned and more closely studied. During the 1980s, 570 additional amphoras were raised; about 140 remain on the seabed.

The organic contents of still-intact amphoras raised in the 1980s were examined in an effort to determine the nature of the ship's cargo. A recovery on average of just under a dozen grape seeds from intact amphoras indicates that most, or all, of the amphoras had been full and were carrying low-grade wine.

Some of the graffiti indicated that many of the recently made amphoras had earlier carried olives, possibly preserved in sweet wine, and some of these jars yielded, along with grape seeds, degraded bits of olive pits. Graffiti on several of the older globular jars indicated that they had once held lentils. Several dozen different marks of ownership occur on the globular amphoras; some jars had had more than one owner. Since the amphoras show few signs of prolonged use, it is likely that many of the older ones had served for some time as sedentary storage jars.

It is extremely unlikely that some kind of commercial activity would have brought together on one ship so many different types of reused amphoras having so many different prior owners and varying so greatly in age. However, historical events taking place during the period when the ship sank give us a quite plausible explanation for the assemblage that does not involve commerce. Between 611 and 628, the Byzantine Empire was engaged in a protracted war with the Persians so devastating and financially costly that it became necessary for the church to lend major assistance in the provisioning of the army, partly through levies of produce from church-owned lands. Particularly in view of rather frequent allusions to the Christian faith among the graffiti on the amphoras, it seems likely that the ship's cargo of low-grade wine had been part of this effort. Some of the major types of globular amphoras and many of the lamps on board may have been made somewhere along the coast not too far north of Yassi Ada in the general area Samos and Chios. On Samos, a contemporaneous church complex where cylindrical and globular amphoras much like those from the Yassi Ada ship waited to be filled with oil or wine from nearby presses has been cited as archaeological evidence for the church's role in provisioning military bases at this time.

We are now inclined to believe that the ship, with its priest-captain and tiled galley roof, belonged to the church and had been so well equipped for the preparation and service of food to a considerable number of people in order that it might transport churchmen as well as cargo.

The end of the war against Persia came in 628 with both sides totally exhausted. The power vacuum made possible the successful Arab invasions in the 630s that were to change the Mediterranean world forever. This historical context and the unusual nature of the ship and its cargo makes the 7th-century Yassi Ada wreck one of the most interesting and important excavated shipwrecks in the Mediterranean.

(text adapted from van Doorninck, Jr., F.H., "Kirkens skib?-Yassi Adavraget fra 600-tallet," Hvad Middelhavet gemmer [Arhus 1997] 105-120.)

Bibliography

Bass, George F., and Frederick H. van Doorninck, Jr., Yassi Ada 1: A Seventh-Century Byzantine Shipwreck (College Station, Texas 1982). The definitive excavation report.

Bass, George F. Archaeology Beneath the Sea   (New York 1975). A popular account.

Bass, George F., "Underwater Archaeology: Key to History's Warehouse," National Geographic 124.1 (1963) 138-156. A popular account of the seventh-century wreck excavation.

Throckmorton, Peter. History from the Sea: Shipwrecks and Archaeology (London 1987). The American edition is titled The Sea Remembers: Shipwrecks and Archaeology (New York 1991). Describes and illustrates some of the finds from the seventh-century and Ottoman shipwrecks at Yassiada.

van Doorninck, Jr., F.H., "Kirkens skib?-Yassi Adavraget fra 600-tallet," Hvad Middelhavet gemmer (Århus 1997) 105-120. The most recent analysis of the 7th-century shipwreck.

Institute of Nautical Archaeology
2000, Yassiada 7th Century Byzantine Shipwreck Excavation
URL, http://ina.tamu.edu/yassiada7.htm