Thousands of years of maritime history lie hidden beneath the waves off the Turkish coast. Bronze Age merchantmen, Byzantine dromons, and Venetian and Ottoman galleys are among the hundreds of vessel types that plied the waters around Turkey in the past. Some of these vessels succumbed to the dangers of the coast, to sink into obscurity in the deep. Discovering and documenting Turkey's nautical heritage is one of INA's long-term goals. To this end, INA has been conducting shipwreck surveys since 1973, and annually since 1980.
The 1993 survey was conducted between September 10th and October 7th under the direction of Cemal Pulak and Donald Frey. Two different teams, one on the INA survey vessel Virazon and the other aboard a Turkish trawler combed the shores of the Provinces of Izmir, Aydin, and Mugla (fig. 1). Violent storms in the Aegean Sea prevented us from investigating the Strait of Samos, where old reports claim a Bronze Age wreck lies. The Virazon waited for the weather to improve for almost a week in the protected waters around Cesme but when the storm failed to abate the search was canceled in that area. Instead, the Virazon sailed south- ward to begin the survey on the Mediterranean coast. Engine failure, however, put an end to that phase of the survey on September 28th.
Nevertheless, in that short time we located six exciting wrecks. Five are near the rocky promontory called either Kizilburun or Inceburun, a dangerous cape just to the west of Teke Burnu. Northwest of this promontory lies the wreck of a steel ship said to have sunk 50 or 60 years ago, which demonstrates how dangerous sailing continues to be in this area.
The most interesting of the five wrecks at Kizilburun is a Byzantine wreck lying at a depth of 39 - 40 m and laden with marble architectural elements. From the distribution of these elements and other remnants over an area only 14 m long by 4.8 m wide, it is clear that the ship came to rest on its keel without being scattered over the seabed. Spiral-fluted columns of different sizes, slabs of white and red marble, a column capital, and a double column of a type used as a partition between windows and naves in churches are among the items on the wreck. It has not been possible to confirm the exact nature of these marble elements which, for the most part, have been partly eroded by the sea and largely covered by encrustation, but the fact that among them are pieces used in churches or other monumental structures suggested at first that the ship was carrying an order of materials to be used in a specific construction.
A similar cargo was discovered some years ago in a wreck dated to the second quarter of the sixth century A.D. off the harbor of Marzamemi in southeast Sicily. The Marzamemi ship was dubbed the "Church Wreck" because its cargo consisted mostly of Marmara (Proconnesian) marble architectural pieces, probably for a basilica under construction in North Africa. By shipping "prefabricated" elements from the center to every corner of the empire, the Byzantines achieved an important standardization in church construction. Such shipments also resulted in great savings in shipping and fabrication costs.
At Kizilburun, 15 to 20 amphoras of a single type, half in good condition. (fig. 2). were found among and around the marble pieces. Iron anchors and single- and double-handled galley pots in the sand at the deep end of the site provided evidence of the ship's bow.
The anchors and amphoras suggest that the ship sank in the 10th century. However, the architectural elements show peculiarities of the sixth century, and, with the exception of the fluted columns, the majority of them are only one of a kind. In addition, they appear to be finely finished, rather than left in a rough state for transportation purposes. We therefore believe that they probably represent a collection of elements dismantled from an older building for reuse.
Seventeen meters northwest of the Byzantine marble wreck, toward shore, at a depth of 36 m, we encountered the remains of a shipment of millstones (fig. 3). Seventeen large millstones approximately one meter in diameter, and a smaller one only half a meter in diameter, were recorded. Each stone has a central hole 10 cm in diameter for the securement of a driveshaft or an axle. The shapes of three Byzantine-type iron anchors found in the northern part of the wreck area could not clearly be determined because they were poorly preserved and partially buried in the sand. It was not possible to ascertain whether these anchors belong to the millstone wreck or to another wreck lying just upslope (see below). The absence of other artifacts around the millstones makes their dating difficult. A photograph of an anchor resting on one of the millstones, however, shows the anchor arm joining the anchor shank at an obtuse angle. If this is not a matter of a break in the anchor's arm, we can date this anchor, and probably the wreck, to the 10th to 12th centuries A.D.
Although the first-century B.C. Roman architect Vitruvius stated that large wheel-shaped stones of the type seen on this wreck were operated by water power, the earliest known archaeological examples belong to a Late Roman period water mill. It is believed that olive oil presses utilizing large circular stones came into use in the 12th century, and windmills were first used in that or the following century. However, little has been learned from archaeological excavations about medieval millstones with the exception of those found in Greece on a 12th-century wreck near Alonnissos in the northern Sporades. If the new wreck proves to belong to the 10th to 12th centuries, it will be of substantial significance insofar as it will provide new information concerning the millstone trade, which played a very important role in the Middle Ages.
Directly north of the millstone wreck, on a steep rocky slope 28 to 36 m deep, fragments of the spherical amphora types well known from the Yassiada seventh-century wreck were discovered. It is not known if two small marble slabs found at the deep end of the site belong to this wreck or if they fell from the shipment of marble elements that lies in deeper water to the southeast. Presumably, the ship crashed into Kizilburun, splintered into pieces, and spilled its cargo of amphoras over the rocky slope. The scattered nature of the wreckage makes it extremely difficult to define the boundaries of this wreck site, and probing the core will undoubtedly reveal additional amphoras.
At the extreme southern point of Kizilburun, hundreds of amphoras lie scattered on a rocky slope at a depth of 28 to 37 m. Many are broken, but intact amphoras also were encountered, particularly in sand pockets at the deep end of the site and at the base of the rocky slope where it merged with a flat, sandy seabed (fig. 4). The occurrence on the slope of large stones and seabed rubble mixed with the amphoras indicates that after the ship sank part of the slope slid downward. No other artifacts were discovered during our survey of the immediate area. The amphoras are all of the same type, and allow this wreck to be dated to the 10th to 12th centuries A.D.
South of the above mentioned Byzantine wreck, we found a wreck of uncertain date at a depth of 45 to 48 m. In the assemblage are nine column drums (1.75 m in diameter and 0.95 m high), a large marble slab (1.75 by 1.15 by 0.32 m), and an element that cannot yet be identified with certainty because of covering sea growth and heavy encrustation, but which may be a column capital of the Doric order. Additionally, one or two marble slabs were seen on the rocks in the northern part of the wreck. If we assume that all the pieces are drums comprising a single column, then the column, with a total height of 9.3 m, must have belonged to an impressive structure. If we are right in believing that the capital is Doric, then the ship may be dated to the Classical period. However, bearing in mind that monumental Doric structures are rare in Anatolia, and that the few known examples are mostly situated to the north of Kizilburun, we may assume that the column was either lost while being shipped to Greece or being sent north along the Anatolian coast. On the other hand, it is also possible that the so-called capital is not a Doric element, but the base for an Ionic or Corinthian column. To better assess the nature of the site, additional visits to the area are needed, to clean the "capital" element and determine its true function and shape. No other artifacts were found except for a seemingly intrusive intact first-century B.C. amphora next to the column drums.
South of Cesme at Altinkum (Golden Sand) lies a wreck approximately 60 m from shore and 4 to 5 m deep. The vessel's cargo consisted of 40 architectural elements including Corinthian column capitals, each 1.3 m square and 0.90 m high, and 25 column bases that vary between 1.15 and 1.35 m on a side with a height of 0.45 m. The artifacts lying closest to shore have shifted little because they rest on a sandstone seabed, whereas the pieces to the south lie on a sandy bottom and have been much more scattered by the waves. All the pieces were roughly shaped, possibly to be finished after arrival at their destination. From the shape of the capitals we can date this wreck to the fifth or sixth century A.D.
Among the materials recovered during the 1993 survey and presented to the Bodrum Museum of Underwater Archaeology was a small lead anchor stock of Roman type, 0.65 m long, found off Teke Burnu (fig. 5). Raised from a depth of 37 m, it has at its center a square hole for the anchor stock. This type of stock was used roughly from the middle of the second century B.C. until the fourth century A.D. Its most widespread use, however, was in the second quarter of the second century B.C. All lead items seen during our surveys have been retrieved for the Bodrum Museum. Recovery of such artifacts, which have high scrap value, prevents them from being looted and destroyed by divers.
Meanwhile, aboard the trawler, the second team sought the wreck of a ship that had carried the Late Hellenistic bronze statue of a naked athlete and a larger than life bronze arm of a second statue, both now in the Izmir Archaeological Museum. The bronzes had been raised in trawler nets some years earlier off Foca. In order to survey the large area in a detailed and systematic manner, we used a prototype side-scan sonar coupled to a computer and Global Positioning System developed specifically for this type of work by INA Director Martin Wilcox. Because of the water depth in the search area, a remotely operated vehicle (ROV) was used in place of divers for identification of sonar targets. The largest and most striking of these targets, with some additional scattered ones that resembled shipwrecks, were scanned with the television camera of the ROV, but all proved to be submarine rocks or scattered groups of stones lying on the otherwise generally flat and sandy bottom. Other sonar targets could not be investigated due to particularly rough seas. Thus, our attempt to find the shipwreck that had yielded the bronze statues remains inconclusive, but one target, enhanced later by Marty Wilcox, appears to be extremely promising. We hope to be able to investigate this target in the near future. On the other hand, a wreck with a cargo of lead ingots at a depth of 50 m off Eski Foca, reported to us during our 1992 survey, was located by the sonar. Unfortunately, however, bad weather prevented visual inspection of the wreck by either divers or the ROV.
--Cemal Pulak
The 1994 survey was conducted in two phases. The first part began on June 3rd, two weeks prior to the final season of the Uluburun excavation. The purpose of this phase was to visit the Bozburun site to plan for the following summer's excavation, to continue the examination of the Cape Gelidonya wreck site, and to investigate several possible wreck sites.
A wreck near Dalyan that was first investigated in 1986 was reexamined. Discovered by sponge diver Ahmet Guntas, the wreck assemblage included 18 amphoras in three sizes, all of which have flat bases. An amphora of the largest size was recovered from the wreck (fig. 6). This material, along with amphoras from a second wreck discovered during our survey, plus a collection of similar jars in the Bodrum Museum of Underwater Archaeology, comprise a series of flat-bottomed amphoras with small lug-shaped handles on their rounded shoulders, and narrow mouths. Although stylistically they appear to be of the late Medieval period, we have been unable to find any conclusive evidence for their dating.
Five days were spent at Cape Gelidonya. One of the objectives of the visit was to search for the ship's anchors around a pinnacle of rock on which the ship probably ripped out its bottom. Underwater scooters were used to cover maximum area during this search. While investigating east of the pinnacle Murat Tilev stopped to examine a Byzantine anchor, placing his scooter on a large rock nearby. Upon retrieving his scooter he discovered that the "rock" was actually a stone anchor. A thick layer of seaweed had camouflaged it, rendering it invisible to cursory observation. Examination showed that the hawser hole was aligned toward the rock pinnacle. This, taken together with the anchor's position relative to the area of the shipwreck, suggests that the crew had cast it off as a last resort to prevent the ship from hitting the island. Raising and cleaning the anchor revealed it to be similar to the ones found on the Uluburun wreck (fig. 7). At 219 kg (c. 480 lb), it is heavier than any of the stone anchors from that wreck that have been weighed to date. Like the Uluburun anchors, the example found at Gelidonya was made of sandstone, though of a coarser variety. The lack of any anchors at Cape Gelidonya had always baffled its investigators, but this fortuitous discovery has solved that long-standing mystery.
Additional time was spent trying to free a heavily concreted pithos base, seemingly of Cypriot origin and similar to those at Uluburun, that had frustrated all previous attempts. Several days of chiseling revealed the base to be much larger than expected, and it still could not be removed. The retrieval of this artifact will have to wait for another year.
On October 3rd, three weeks after re turning from Uluburun, INA staff members left for the second phase of the survey. Joining us was Cumhur Ink, a veteran sponge diver who originally discovered the Sheytan Deresi wreck (see AlNA Newsletter 1.1: 3-4 and 2.3: 1-5). His knowledge of the coast and diving conditions proved invaluable to our search.
For the first time, INA divers began carrying small cylinders of oxygen strapped to their air tanks during survey dives. Previously, divers had to re turn to the Virazon or be located by the tender boat so that an oxygen hose could be lowered to them for decompression. Now divers could decompress anywhere on oxygen, increasing the safety factor and permitting longer dives.
Near the town of Sigacik at Teke Burnu, we located the scattered remains of two small Byzantine wrecks. One wreck, containing small piriform amphoras probably intended for wine, lay smashed on a rocky slope. The debris continued down to a depth of 37 m, where the rocky slope merged with a nearby flat sandy seabed. At the base of the slope, on the sand, were two complete amphoras along with sherds of others. One of the intact jars was raised for identification purposes and deposited at the Bodrum Museum. Probing in the sand may reveal additional wreck material, but all that was readily visible amounted to less than 20 amphoras found among the rocks in groups of four to five jars. Cumhur Ink, the site's original discoverer, reminded us, however, of old local rumors that many amphoras from this wreck had been removed for sale in Izmir markets. Based on the sole example recovered from the site, the wreck should be dated to the 11th century A.D.
A second scatter of jumbled amphora sherds nearby appears to contain amphoras of similar size, but differs from them in having wider and longer necks. Because of the shallow depth of this wreck, all amphoras are extensively damaged from pounding waves and concreted to the rocky seabed, a factor that prevented us from collecting examples for later study. Also visible among the wreckage were parts of two y-shaped iron anchors. If contemporary with the amphoras, the anchors are clear indicators of a 10th to 12th century A.D. date for the assemblage. It is possible that the material found here may in fact be from the wreck discussed immediately above. Further investigation in the area may elucidate this possibility.
Our search next took us to Papaz Adasi near the ancient city of Didyma. Ilik had heard of an amphora wreck there that had been discovered 20 years ago and had been impressive enough for foreign divers to photograph. When we located the wreck the only things evident were a few amphoras lying half buried in the sand and numerous fragments strewn about the area. Unfortunately, the wreck obviously had been looted. Those amphoras that we examined had the characteristic double handles of jars made on the island of Kos during the first century A.D. for the export of their famed wines. We are hopeful that much of the wreck is still buried and protected.
In 20 years of surveying, INA has discovered approximately 100 wrecks while searching hundreds of kilometers of the Turkish coast. Future plans call for continuing eastward into the provinces of Icel, Adana, Antakya, and lskendurun. In time, we hope to cover the entire Mediterranean coast of Turkey. The experience gained in the surveys is also being put to use in such places as Bulgaria, Egypt, and Bahrain. In all of these places, INA surveys are aimed at locating vessels that can provide us with information about maritime trade, ship construction, and the history of seafaring.
-Edward Rogers