en zh es ja ko pt

Volume 35, Number 4July/August 1984

In This Issue

Back to Table of Contents

Shards of History

Written by John Lawton
Photographed by Donald Frey

In an ancient wreck deep in the Aegean, the largest collection of Islamic glass ever found.

In Corning, New York, this summer, scientists presented some answers to questions raised by the discovery, 10 years ago, of a mysterious medieval wreck and its unusual cargo: nearly a million fragments of brilliantly-colored broken glass.

The story of the "glass wreck" began in 1973, when Mehmet Askin, a retired Turkish sponge diver, guided members of the Texas-based Institute of Nautical Archeology (INA) to Serce Limani — a remote anchorage on a peninsula jutting into the Aegean Sea — and Yuksul Egdemir, a Turkish official assigned to the expedition, went down to see what was there. He surfaced with a rainbow of brightly-colored shards in his hands. "There's glass everywhere," he reported.

Because the INA was busy working on other wrecks then, nothing was done about the glass wreck for four years, and the glass littering the sea bed at Serce Limani remained a riddle. Was it simply jettisoned cargo, or was a wreck lying there too? In the spring of 1977, the institute, determined to find out, assembled a team of 20 Turkish and American divers and scientists and equipment: a wooden diving barge, a one-and-a-half-ton decompression chamber and a "telephone booth" — a submergible plastic dome connected by telephone lines and compressed-air hoses to the surface; divers can rest in the "booth" during seabed excavations.

"I was gambling heavily at Serce Limani," wrote George F. Bass, then the president of INA and now its archeological director. "The cost of such an expedition ... is sizeable. Unless we found a ship along with the glass, the funds would be wasted."

Diving in two's for a maximum of 20 minutes at a stretch, the divers began their search. The first pair uncovered more glass and some amphorae — large clay jars used to carry oil or wine — but no wreck. A second pair went down to continue the search. They too made out several amphorae and a lump or two of glass, but still no ship.

"Then suddenly, at the very end of the dive," said Frederick van Doorninck of Texas A & M University, INA's headquarters, "there it was, almost as if it had materialized from empty sand — the definite lines of a hull." It was lying on its port side in a bank of sand — that had fortuitously preserved it from shipworm — at the base of a rocky slope 33 meters (110 feet) below the surface.

Laying a metal grid over the wreck area to enable the precise plotting of every find, the divers began the slow, painstaking task of salvaging the sunken ship and its contents.

It was difficult — and often painful work. The divers, for example, had to pry thousands of razor-sharp glass fragments out of the wreckage — some of them cemented together by silt and sand and though the risk of cuts and infection was high, the divers could not wear gloves, as touch, says Bass, was as important as sight when working on the murky sea floor.

Because they were working at depths of more than 30 meters (100 feet), they also had to think about the "bends," an often crippling, sometimes fatal malady, caused by nitrogen bubbles forming in the blood stream when a diver surfaces too rapidly.

There were other dangers too. Once, says Bass, a huge octopus descended from its lair in the rocky slope above the wreck, spread its tentacles across a two-meter (six-foot) section of metal grid and remained for a while as if to say "I'm the king of Serce Limani." No one argued.

There were, however, compensations. Once, gently probing the sand in his assigned area of the grid, Bass felt the base of what he took to be another broken bottle. So far the divers had found no glass vessels intact, but now they did: a lovely piece of work rounded at the base — pinched at the sides in an hourglass effect — and, miraculously unbroken.

This was to be rare. Of an estimated 10,000 glass vessels on board, only 80 remained intact — in living areas at the bow and stern of the ship. The rest, stowed in the ship's cargo hold, were smashed in a million pieces — literally. Nevertheless, the divers carefully brought up every fragment — including some that were cemented together and had to be separated with dental tools. Each of the shards was washed, dried and individually labeled with the exact position it was found on the wreck. From the hold, the expedition also salvaged two-and-a-half tons of blue, green and amber "cullet" — heavy chunks of glass, which, melted down, were the raw materials of the medieval glassblowers' trade — plus hundreds of shipboard artifacts: terra-cotta cooking vessels, lead fishing-net weights, a wicker basket of iron tools, some weapons and some wooden chess and backgammon pieces.

Two finds — a dozen copper coins and the 16 glass weights — helped date the wreck to about 1025. Other finds, however, tended to confuse the archeologists. The glass weights, for instance, were traced to the reign of the Fatimid Caliph al-Zahir (1021-1036), suggesting they belonged to an Arab merchant. But Byzantine lead seals, one still unused, pointed to the presence on board of one or more Greek merchants.

Shipboard pottery was also confusing — it included a mixture of Islamic and Byzantine shapes and designs. A bucket with an Arabic inscription was without doubt of Islamic origin, and jewelry found on board was almost certainly Fatimid. But food remains included pig bones, and many of the 900 lead fishing weights were inscribed with Christian symbols. One of the ship's eight iron anchors, on the other hand, is stamped with an Arabic letter. And although the glass was definitely Islamic in design, part of the cargo comprised dozens of Byzantine wine jars scratched with Greek graffiti. So was the ill-fated ship Islamic or Byzantine? And what was the nationality of the crew?

As the excavation progressed, the divers also realized that the wreck's hull was unlike any other they had seen before. "The Serce Limani wreck," wrote INA Ship Reconstructor J. Richard Steffy, "represented an undocumented ship type from a period and area in which comparatively little is known about maritime matters."

Accordingly, the INA set about raising and preserving the ship's waterlogged timbers — by a two-year bath in polyethylene glycol — with a view to reconstructing the hull as accurately as possible. "Hundreds of hours were spent experimenting with research models, making drawings of details, compiling lists of evidence to confirm or deny a process, or just pondering the reason for assembling things the way we found them," says Steffy. "Always the results were the same; only one construction sequence satisfied all the evidence at hand: the Serce Limani shipwright set frames before planks."

This was significant because hulls of earlier vintage were fashioned from the bottom up, a plank at a time, and the framework inserted afterwards. The Serce Limani craft had apparently been built in reverse — by nailing planks onto a skeletal frame of keel and ribs, much the same way as wooden ships are still made today. The result was a stronger, roomier hull, and a saving of time and timber. In fact, says Bass, the Serce Limani wreck is the "remains of the earliest known seagoing hull built with modern notions of naval architecture."

By the end of the summer of 1979, the INA had completed mapping and raising all fragments of glass and wood at Serce Limani, and towed them by barge to an imposing 15th century Crusaders' castle overlooking nearby Bodrum harbor. "The task before us was staggering," says Bass. "Stacked around us were crates containing about two thousand plastic bags of broken glass, each holding from one-to-five hundred fragments ... of glass vessels."

To solve this million-piece jig-saw puzzle, the INA assembled a team of Turkish and American students of nautical archeology at long work tables in the basement of the Bodrum Castle's "English tower," with the shards piled upon them. The shards were arranged according to areas where they had been found — in the belief that joining pieces would be discovered among those fragments that were closer to each other in the wreck.

This, they soon learned, didn't work. So the students tried to catalog tumbler bases, jar rims, bottle necks and other recognizable shapes on the assumption there were few, if any, complete vessels in the cargo. And when this didn't work either, all the glass was packed back in the plastic bags and the search for joining pieces began again. The students sorted the shards into more than a dozen different categories according to color — blue, amber, green, purple and plain — and decorative patterns, both molded and engraved. Finally, by further subdividing individual color into darker and lighter shades, they began to spot "joins."

The student sorters — now joined by half-a-dozen local glass menders, a professional glass conservator and two full-time illustrators — each study a specific shape; one has classified nearly 250 glass lamps, and another has identified and partially mended the remains of 800 tumblers and beakers, 100 of them engraved with intricate designs. Other shapes noted include bowls, bottles, pitchers, plates, jars, cupping vessels for draining blood — even hospital urinals.

Some of the Serce Limani wreck's glittering cargo is already on show in Bodrum Castle, along with the remains, found nearby, of the Tomb of Halikarnassos — one of the Seven Wonders of the World (See Aramco World, May-June, 1980). Eventually a new museum at Bodrum will house the restored hull of the glass wreck. But meanwhile, at least, the scientists have been able to answer some of its riddles.

In Corning, New York, for example, chemical analyses of 80 specimens of glass from the wreck, made by the Corning Museum of Glass, indicate that it is of Islamic manufacture, according to Dr Robert Brill, a research scientist at the museum. "With few exceptions, the vessel glass and cullet have a uniform composition and could well have been made at a single factory," says Dr. Brill.

One surprise was the discovery that "four fragments of rare examples of emerald-green glass from the wreck have an extremely high lead content," says Dr. Brill. "Discounting certain Chinese and Japanese glasses — which are many centuries earlier than the Serce Limani glass — only four parallels for this unusual composition are known among the more than 1,600 samples of ancient glass analyzed previously in Corning. The parallels are all Islamic."

The rare emerald-green glass and other lead-containing objects on board were also subjected to lead-isotope analysis — to try and learn where they might have been made. "The objects contain lead from four or five different mining regions, and probably were brought on board at different ports of call," says Dr. Brill. "At the present it appears that these leads came from mines in Turkey or Iran."

The emerald-green glasses, some ceramic glazes, and a lead net sinker all contain a particularly rare type of lead. "Among nearly 1,500 ancient objects and ores studied previously, only two other leads have ever been found with the same isotope ratios," says Dr. Brill. "One is an ore from a mine in northwestern Iran, which might have supplied lead to cities along the Black Sea coast. The other is from an emerald-green flask from Heshbon, an archeological site not far from Amman in Jordan. That glass dates from the 9th to 11th centuries and might have been made in the same factory as the Serce Limani emerald-green glasses."

Dr. Brill's research, of course, is just a start. Despite this summer's discoveries, the wreck remains what one nautical archeologist describes as "the most intriguing ancient ship yet discovered."

John Lawton is a contributing editor of Aramco World Magazine.

This article appeared on pages 4-11 of the July/August 1984 print edition of Saudi Aramco World.

See Also: ARCHEOLOGY AND ARCHEOLOGISTS,  GLASS,  SHIPWRECKS

Check the Public Affairs Digital Image Archive for July/August 1984 images.