A COVE OF MANY SHIPWRECKS: THE 1995 INA/CMS JOINT EXPEDITION TO TANTURA LAGOON
Map of Tantura Lagoon (Drawing: P. Sibella)
The CMS and the Institute of Nautical Archaeology (INA) at Texas A&M University (TAMU) joined forces once again to carry out a second season of exploration in Tantura Lagoon from mid-October through mid-December 1995. The primary aim of the expedition was to complete the in situ study of the vessel discovered in Trench VI in 1994 which, following the completion of this recording, has been renamed the Tantura A shipwreck. A hydraulic probe survey was carried out in the immediate vicinity in search of additional sections of this vessel.
The hull was plotted using a computerized three-dimensional mapping program (WEB), which was specifically created for use in nautical archaeology. Over 250 points on Tantura A and the surrounding artifacts in Trench VI were recorded using this method, employing six fixed datum points. Direct measurements were taken, and some details were also copied at a scale of 1:1, on mylar. All the ship's timbers were photographed from a distance of 30 cm with a 15 mm Nikkor lens and strobe, with 50% overlap. Numerous samples were removed for macrobotanical and palynological examination. During 217 recorded field-related tasks, expedition members accumulated over 600 diver hours.
The surviving hull, situated roughly along a NW to SE axis, represents about 25% of the bottom of a small coaster that is believed to have been about 12 m long. The entire side of the hull north of the keel is missing, with the exception of a short portion of the garboard next to the post, and a loose, minor plank fragment. Virtually all of the SW quadrant is also missing. Only a single articulated garboard strake continues past the broken end of the keel, giving a total preserved hull length of slightly over 9 m.
In 1994 we noted charring on the hull, which led us to theorize that the vessel had suffered an onboard fire. Careful examination of the locations of the charring during the 1995 campaign, however, revealed that it was localized at the extremities of the strakes - at the NW where they join the post, and at the SE end of the garboard. Moreover, the charring continues beneath the framing stations but is not evident on the keel, the post, or the surviving frames. This indicates that the burn marks are not the result of an unintentional fire, as the planks must have been burnt prior to being attached to the frames.
These considerations led J. Richard Steffy to conclude that the burn marks result from a process known as "char-bending," in which water soaked planks are heated over a fire as they are bent to the required shape. Strakes receive their strongest curvature at the ends of a vessel, where they must be bent inwards to meet the posts. This explains why on Tantura A the charring is found only at the ends of the hull. This is the earliest recorded evidence for this process in planked-hull ship construction.
A grayish concretion was originally thought to be 'mastic', a putty-like material occasionally found on ancient shipwrecks between frames, sheathing or planking. Further chemical and paleoethnobotanical examinations suggest, however, that it is simply a marine concretion that formed subsequent to Tantura A's deposition.
The ship must have come to rest in its present position before it broke up entirely. This is evident from the observation that the surviving hull planking is perfectly aligned, despite the fact that most of the frames to which it had been attached were ripped away in a 'zipper-like' manner, when the missing section(s) separated from the hull.
Surprisingly, Tantura A bore no evidence of the unpegged type of mortise-and-tenon joints considered typical for ship construction in the mid-first millennium CE. At first, due to the lack of any traces of edge-joinery, the ship's construction appeared to be later than the overlying smattering of Byzantine period pottery. This raised the possibility that it may be a late shipwreck into which Byzantine period ceramics had been washed by the current. In 1995, however, we noted Byzantine sherds glued fast by resin to the planking. This process could only have occurred prior to, or perhaps during, the sinking event - but not after it. This evidence conforms with the results of three radiocarbon tests, that confirm a mid 5th to mid 6th century CE date for the hull. A detailed report on this hull, by Ya'acov Kahanov and Jeff Royal follows.
Rope was found preserved in several locations underlying the hull. Apparently, this belonged to the ship, which became entangled during the wrecking event.
In conclusion, Tantura A is the oldest recorded shipwreck in the Mediterranean to have been built without mortise-and-tenon joinery, and in the innovative ship construction techniques that were to become standard during medieval times. Thus, at present, this hull represents an important transition point in the evolution of Mediterranean hull construction.
Stratigraphy is rarely found on the seabed. The abundance of antiquities in Tantura Lagoon, however, resulted in our repeatedly finding meaningful stratigraphical configurations. Tantura A, for example, had come to rest on two stone anchoring devices. At its NW end, the hull landed on an upside-down 52 kg stone anchor-stock that predates the ship by about a millennium.
Three stone anchors were found adjacent to Tantura A in the SE area of Trench VI. As these pierced stones vary in size, shape and stone type, it is unclear whether or not they were deposited as a group. One limestone anchor, rounded in shape, was found partially underlying the keel. A second, made of a crumbly green stone, weighs 83.5 kg and has an asymmetrical shape generally associated with Bronze Age stone anchors. It was found, however, lying upon Persian period pottery, indicating that the anchor cannot predate that chronological period. This requires us to reconsider the generally early date given to stone anchors of a similar shape found on the seabed without a stratigraphical context. A third anchor is made of limestone and weighs 55.8 kg.
Tantura A constitutes a relatively small portion of the entire ship. In search of additional sections, Andrew Lacovara carried out a hydraulic probe survey in its immediate vicinity. Unfortunately, no additional parts of this ship were found. And yet, the results were far from disappointing. The survey revealed two other coherent hulls, as well as the loose timbers of at least three more vessels. All of these hull remains were located in an area smaller than a basketball court.
A dozen meters NW of Tantura A, and at a depth of only 1.5 m below MSL, a collection of loose timbers - ten frames and a single plank - was found. All of the frames were naturally curved branches.
This site also contained basketry, long spans of rope, and even some dyed cloth. One rope was knotted while another ended in an eye-splice. The manner of deposition suggests that all these artifacts derive from a single vessel. One frame has been radiocarbon dated to about the 18th century CE.
Only 8 m west of Tantura A, another hull was located in Trench VIII. This vessel is also situated along a NW-SE axis. At the latter extremity, the ship has been broken and its keel tortuously rip-twisted apart. At the SE end, the hull continues into the sandbank.
This hull has a massive center-line timber overlying the keel and frames, into which is cut a long rectangular mast-step heel (40.0 x 8.6 x 6.5 cm). The unusual shape of the heel may indicate that this ship had carried a lateen rig. At the SE extremity of the trench, a smaller center-line timber, the identity of which remains to be determined, is secured above the first. No evidence for mortise-and-tenon joints was noted, although we examined only a limited length of planking edges.
The portion of the hull retrieved from the sand overburden lacked datable artifacts. It has been radiocarbon dated, however, to the late 7th through mid 9th centuries (680- 850 CE). Thus, the study of this hull could make an outstanding contribution to the evolution of Mediterranean ship construction, filling a critical gap in our knowledge, as it falls between the early 7th century shipwreck at Yassiada and the late 9th or early 10th century CE vessel now being excavated at Bozburun.
A single short plank, bearing two pegged mortise-and-tenon joints, was found amongst the shattered timbers at the NW end of the trench. This plank fragment was associated with quantities of Roman period (2nd century CE) ceramics, consisting primarily of small casseroles and juglets in mint condition. This assembly of artifacts strongly suggests the existence of yet another, earlier, shipwreck nearby.
Another large hull was uncovered 12 m NE of Tantura A, in Trench IX. This hull was situated in the channel that runs through the cove. Some of the upper parts of the preserved hull were damaged by teredo, due to exposure in the channel. The Trench IX vessel was uncovered during the final days of the field season and was, therefore, only briefly studied.
Visible at one extremity is a short section of the keel with a rabbet to accept the garboards. A false keel continues forward of the keel and is slotted into it by means of a tongue-and-groove assembly. No mortise-and-tenon joinery was noted in the limited portion of planking edge and keel available for study.
A transverse timber with a slot along its upper surface may have served as the seat for a removable bulkhead. Ceiling planking predominates throughout the area of the exposed hull. Between the transverse slotted timber and the western extremity of the hull, the ceiling planking lies roughly parallel to the keel. Amidships of the transverse timber, however, the planking is aligned perpendicular to the keel. Two carved graffiti - one depicting a cross covered by an arc, and the other a 'delta' - were noted on the ceiling planking.
Two large kurkar ashlar blocks have retained their original alignment, directly inboard of the transverse timber. These blocks may have served as ballast or alternately, constituted part of the ship's cargo.
Quantities of broken ceramics, lying primarily at the ship's extremity, date this hull firmly to the Byzantine period. This date is further confirmed by radiocarbon testing which provided a mid 6th to mid 7th century (553-645 CE) date. The Trench IX hull, therefore, appears to be roughly contemporaneous to the 7th century CE Yassiada shipwreck, and in a better state of preservation.
At Float 2, a site also located in the channel, a single, disarticulated frame was found. Preliminary inspection suggests that it does not conform to any of the other ship/timber assemblages uncovered by us at Tantura. This site also revealed a potpourri of ceramics dating to a variety of different periods. Particularly worthy of note is the top of an Early Iron Age 'Canaanite jar' with angular shoulders, dating to the 11th or 10th century BCE. This is the third example of this jar type that has been retrieved from the cove.
The discovery of so many shipwrecks - some of which were badly damaged during the wrecking event - in such a small area, and in a lagoon that seems ostensibly protected from the weather, requires an explanation. At least some of these ships appear to have sunk while trying to enter the cove during storms. Crossing the shallows between the islands, a ship could be broken on the rocks and afterwards swept into the cove where the swift current that runs there during storms would have 'smeared' the now fragmented hull and its contents across the lagoon and then quickly buried them under a thick layer of preserving sand. Such a scenario would explain Tantura A, as well as the hull remains in Trenches VII, VIII and perhaps IX.
During the 1994 hydraulic survey, two lead-filled wooden anchor stocks were found. In 1995 we returned to study the Trench IA anchor stock. It is broken about a quarter of the way along its 2.10 m length. The stock is loaded with four lead weights. Four holes have been cut into the stock's upper and lower surfaces, and its center is recessed to receive the anchor's shank. Smashed Persian period jars were found in its vicinity. The contents of one jar had flowed over the stock and now adhere to its surface. This strongly suggests that they reached the seabed together. Lead isotope analysis of the stock, by Sophie Stos-Gale reveals a consistency with ores from the Troad, and particularly from the Altinoluk deposit. A second possible source is the eastern Rhodpi Bulgarian deposit of Madjarovo.
Twenty-five kg of organic-rich materials - from bilge mud to resins and residues found inside ceramic containers - were brought back to TAMU's Paleoethnobotanical Laboratory in 1995. This material will form the basis of a masters thesis at Texas A&M University by Steven Butler. While this research is still at a preliminary stage, it has already begun to add a fascinating additional dimension to the shipwreck.
A sample of bilge mud scraped from near the centerline of Tantura A, contained a potpourri of fossil pollen from 19 taxa, while other types remain unidentified. The most common domesticated plant represented is olive (Olea europaea), with cereal pollen second in number. Other significant fossil pollens include grape (Vitis sp.), hazelnut (Corylus), sumac (Rhus), terebinth (Pistacia sp.) and palm (Areacaea), as well as several species of umbels, a family that includes many spices, such as caraway, dill, cumin and celery. The sample also contained three hemp (Cannabis sativa) pollen, perhaps derived from the vessel's cordage. The majority of seeds belonged to wine grape (Vitis vinefera). A single degraded wild grass pollen grain was identified in a sample of the rope found compressed between the keel and the stone anchor stock.
Butler also analyzed the sieved remains of a bag-shaped jar and a resinous concretion from part of a carrot-top amphora that were found together near Tantura A and are believed to have been part of its cargo. The majority of identifiable pollen from the jar were olive and grape.
The trench IA anchor stock ad associated pottery. (Photo: S. Wchsmann)
The abundance of archaeological discoveries made during our 1995 season was very surprising. Each vessel required long hours of recording and study both in and out of the water. Recording them was both an exhilarating, yet exhausting experience for all staff members. This is perhaps best illustrated by something that happened on one of the last days of the excavation. I came back to the children's house at Kibbutz Nahsholim which was our expedition center. I happened to be whistling a tune. Patricia Sibella was busily recording artifacts. A worried look came over her face.
"You haven't found another shipwreck have you?" she asked, concern in her voice.
She was relieved to hear that we had not.
Acknowledgements. The 1995 season of exploration was made possible by generous grants from the National Geographic Society, TAMU's College of Liberal Arts and by the support of the following individuals: Mr. and Mrs. Stanley Chais, Mr. and Mrs. Ted Halpern, Mr. and Mrs. Harry Kahn II, Mrs. Norma Kershaw, Dr. Leon Riebman, Mr. and Mrs. Peter Skinner, and Mr. and Mrs. John Stern.
Senior expedition staff included INA and CMS faculty and staff, as well as independent professionals: Shelley Wachsmann (Project Director and Principal Investigator; Underwater Still Photography); Michael Halpern (Assistant Director); J. Richard Steffy (Advisor on Hull Reconstruction); Stephen Breitstein (Director of Operations); Ya'acov Kahanov (Hull Reconstructor); Patricia Sibella (Ceramicist and Illustrator); Andrew Lacovara (Hydraulic Probe Coordinator).
Additionally, students from TAMU, the University of Haifa, and local
high schools, participated in the excavation, making this expedition a
truly international educational experience. Participants included the following:
Karim Abu-Moach; Michael Aizenberg; Miriam Belmaker; Sheera Baroz; Joe
Breman; Aaron Brody; Steven Butler (Paleoethnobotany; Video Photography;
Dark Room); Jaynie Cox (Land Photography); Bella Davidson; Eyal Glick;
Daniel Goldstone; Eli Haddad; Tali Kazenelson; Hadas Mor; Vered Romeo;
Mira Roditi; Jeff Royal (Assistant Hull Reconstructor); Nimrod Shay; Maya
Shemla; Claude Tibi (Studio Photographer); Yishai Wachsmann.