The Byzantine-Aksumite Period Shipwreck at Black Assarca Island, Eritrea

Lying in the Massawa Channel, between the mainland of Eritrea and the Dahlak archipelago, the two islands known as the Assarcas became the focus of archaeological investigation beginning in 1995.  Both are ancient coral outcrops, with just enough accumulated topsoil to support euphorbia cacti, scrub grass, and the associated insects and arachnids.  The southern island is known as White Assarca for its surrounding of white sand beaches.  The northern island is called Black Assarca as it is encompassed on all sides but one by coral cliffs stained dark by the sea.  The cliffs reach in places approximately two meters above the sea.  As their uppermost parts are exposed at low tide, reefs surrounding Black Assarca under the water present a danger to watercraft even when the tide is high.  The reefs extend outward from the island for ten to twenty meters, if not more in places before gradually dropping off to a sandy bottom.  Here, at the edge of the reefs on the northern side of Black Assarca Island a ship wrecked some 15 centuries ago and lay undiscovered until 1995.

Excavation of the site off Black Assarca began in early February 1997 and ran for 55 days into March.   Frequent storms and rough seas hampered operations and reduced excavation efforts significantly.  During the course of the excavation, conical and lentoid/globular amphoras and their sherds were found, as were two small non-descript iron pieces, one glass shard, and one lead steelyard counterweight.

Most artifacts were located in the sandy area at the base of the reef.  This sector was excavated down to approximately one meter below the original seabed.   Ceramics extended even deeper into the sea floor, but time did not permit deeper excavation.  Based on one amphora at this level, at least another sixty centimeters of wreck level could be present.  Indications from the area immediately down slope of this amphora and under a large coral head in the center of the site support this view.  As the coral head sat atop the amphoras, having grown there over the centuries, the adjacent area to the north was excavated horizontally.  Thus it was determined that one amphora lay piled upon another amphora, in a scrambled fashion to a depth of approximately a meter and a half.  As excavation began to undermine the coral head, threatening to cause it to tumble down, efforts in this area were suspended.

West of the central part of the wreck the mix of sherds continued in much the same manner, however east of the wreck’s center, the team encountered a different scenario in the area where the main pile of exposed amphoras lay.  This yielded some intriguing data for wreck formation processes.  Under the “main pile,” which was actually only an area where sea-disturbed amphoras had accumulated, was a layer of sand with a number of sherds mixed in.  Under the sand was what was initially believed to be bedrock.  Some amphora sherds were concreted into this and had to be chiseled out of the matrix.  Further examination revealed that this layer was not rock but dead coral lying in large pieces.  These were movable with some effort.  Under the coral was fine silt in which lay amphoras side-by-side in a somewhat orderly fashion.  All of the amphoras exhibited hairline cracks but otherwise retained their shape, held together by the overburden of silt.   

The most common ceramic form was the conical amphora type-- the Aqaba or Ayla-Axum amphora-- found in all excavated sections of the site. Although all but one was broken, they formed a body of material cohesive in both form and style. The corpus of material indicates an origin for the ceramics in Byzantine Aila. The dating can only be estimated to the 5th or 6th centuries based on the other finds of similar ceramics at Red Sea, Indian Ocean, and Mediterranean sites. It is possible that the wreck could date to the early 7th century as well.