Skip to content

“Ships of Ancient Greece” Concept Paper

“Ships of Ancient Greece” Concept Paper prepared by Nauticos and the Institute of Nautical Archeology


Nauticos discovered several interesting bottom contacts during the search for Dakar in the early spring of 1999. Over 300 sonar contacts were evaluated and 22 were deemed worthy of visual inspection.

Among the many targets analyzed, one in particular stood out as being of potentially significant historical importance. Video and sonar imagery was collected at this site to document the target for later examination. This imagery was provided to the Institute of Nautical Archeology at Texas A&M (INA) to determine the origins and significance of this ancient shipwreck.

In addition to this ancient shipwreck, Nauticos discovered five other sites nearby. These sites were not examined to the same extent as the ancient shipwreck site and include a modern freighter and an amphora site. Nauticos and the INA have joined forces to explore these sites. This summary provides an overview of what we currently know about the main ancient shipwreck site and our plans for future expeditions.


The Wreck Site

The shipwreck is Hellinistic in origin, most probably dating from the end of the third century BC or the beginning of the second century BC. The cargo was largely amphorae that contained wine. Two Rhodian amphorae are clearly present near the anchor stocks at what would have been the bow of the ship (Fig. 2). There are very few known wrecks from this time period and none of these are in such an excellent state of preservation.

Fig. 2: Bow area of the shipwreck with an anchor stock, collars and amphorae.

The shape of the wreck site is typical of ancient ships from this period. The amphorae form a more or less ovoid mound, having been stacked in the hold as many as three layers deep, and tapering longitudinally and vertically with the run of the hull of the ship. We estimate the vessel to be approximately 20 meters long.

Fig. 3: Ancient anchors similar to
the type used on
this vessel.

The bow is clearly identifiable given the presence of at least four anchors. Anchors of the period were comprised of wooden shafts and flukes, and lead stocks and fluke supports, or “collars” (Fig. 3). There are at least four collars and three anchor stocks all still oriented as if they were laying on the fore deck of the ship which was the usual position while underway. The most clearly visible stock protrudes from the sediment at a steep angle, and is judged to have a length of at least 2 meters and weight of several hundred pounds. It is not uncommon to find 6 or more anchors on a wreck site from this period, as the anchors required continuous repair and were often lost during deployment.

Fig. 4: Cooking cauldron amid the amphorae. The world’s longest deployed sediment trap.

The stern of the ship is demarcated by the paucity of ceramic material, and the hodge-podge of amphorae, some part of the cargo and others possibly to the crew provisions. There is also a scattering of cooking ware, or coarse pottery commonly found in the galley, which was located in the stern of these vessels (Fig. 4). There is at least one intact serving bowl and two intact pitchers typical of the period.

Fig. 5: Ballast stone located between the main cargo area and the crew area.

The presence of ballast stone just forward of the stern area indicates that the bulk of the cargo was loaded in the main hold forward of the galley on top of the ballast stone (Fig. 5). The galley was separated from the cargo area by both the cooking hearth brickwork and a strong bulkhead. As the wooden structure of the ship deteriorated the cargo would tend to spread and settle, in this case more toward the bow and sides than the stern, partially exposing the ballast stone.


The Cargo

Of the estimated 2-3,000 amphorae INA identified only 2 to 3 Rhodian amphorae on board stating that many more could lie within the pile of amphorae and other material. The predominant cargo seems to be Koan wine carried in amphorae like those pictured right in figure 6.

Fig. 6: Koan amphorae dating from the 3rd to 1st centuries B.C. The second amphora from the right is the most representative of the type found at this wreck site.

The wine of Kos was admired, but it was a relatively inexpensive grade, bought in larger quantities; like Rhodian and Knidian, it sometimes had seawater added as a preservative. Koan stamps occasionally include the letters KO, short for Koion, but the jars have been chiefly identified by the Koan coin symbols and the many Koan names in the stamps on the very distinctive double-barreled handles (Fig. 7).

Fig. 7: On the Left a Koan amphora handle stamped with the seal of Dorimachos
On the right a Koan amphora handle stamped with crab and club, symbols on the coins of Kos.

Rhodian amphorae, like those in figure 8, had a creamy appearance with a peg toe and acutely angled handles. A pair of stamps which may contain the “rose” or rayed head of the sun god Helios often marks them. This same image of Helios appeared on Rhodian coins. On Rhodian amphorae the stamps contain two names, one an endorsement, perhaps by a licensed manufacturer, the other a date, “in the term of so-and-so” usually an annually appointed official, the common way of expressing date in antiquity. Why amphorae were dated is not fully understood. The chief purpose may have been to verify the amphora as a container of a standard capacity.

Fig. 8: Rhodian amphorae from the 3rd Century B.C.

Another purpose may have been to date the contents, identifying for instance the age or special vintage of finer wines and the freshness of the cheaper wines that were generally not worth drinking after a year. Traders and tax collectors had to recognize the make of a jar to know the capacity within a given tolerance and verify the stated value of the contents. Amphorae varied in size but generally ranged between 2 and 3 feet tall and held a little less than 7 US gallons when filled to the brim.The shaping and marking for easy recognition of these commercial containers provides important evidence about the history of ancient trade. Ancient trade routes can be extrapolated from their known point of origin, revealed through the markings and morphology, to the sites were they have been found. Large collections of intact amphorae are relatively rare finds on terrestrial sites. (Fig. 9).

Fig. 9: Pair of rectangular stamps from the Rhodian manufacturer Agoranax. Dated in the term of Sostratos and in the Month of Artamitios.


The Hellenistic Period

Fig. 10: The ancient Greek world around 300 BC.

Following the death of Alexander the Great in 323 BC, his kingdom was divided among his generals. The Antigonid dynasty maintained control of mainland Greece. The Seleucids governed the entire Eastern Empire. And the Ptolemies ruled ancient Egypt.

The Hellenistic period was an international, cosmopolitan age. Commercial contacts were widespread and peoples of many ethnic and religious backgrounds merged in populous urban centers. Advances were made in various fields of scientific inquiry, including engineering, physics, astronomy and mathematics. Great libraries were founded in Alexandria, Athens and the independent kingdom of Pergamum. Because old beliefs in Olympian gods were infused with foreign elements, especially from the east, Oriental cults became popular in the Hellenized world.

 The 3rd century BC saw the rise of ancient Rome. After securing most of the Italian peninsula, Rome entered into a protracted conflict with the Carthaginians for control of Sicily, Spain and the other regions of Punic domination in the Punic Wars. The former empire of Alexander was taken steadily and methodically into Roman hands. The great city of Corinth was destroyed (146 BC), Athens captured (86 BC), and Cleopatra and Mark Antony defeated at the Battle of Actium (31 BC). Their defeat marks the end of the Hellenistic Age.

Between the 3rd and 1st century BC, the Island of Rhodes emerged as the preeminent island in maritime commerce. Rhodes controlled most of the sea trade in the eastern Mediterranean until the Romans eventually rose to prominence. The bulk of the cargo discovered by Nauticos is Koan amphorae from the isle of Kos, which is a tiny island close to Rhodes. Kos was famous in ancient times for the excellent wine that it produced and exported throughout the region.


The Importance of this Discovery

This wreck is valuable to historians and archeologists because of the evidence and information it provides about trade in the region and about open-water trade routes in general. The fact that there are several other similar wrecks in the same region is extremely interesting for several reasons:

1. If they are all from the same general period or time frame, they may provide detailed information about long distance trade over open water at a specific moment in history. This is significant because conventional archaeological wisdom believes that ancient sailors navigated by hugging the coast. Additionally, if the wrecks are from a single fleet that was lost all at once, it is a fascinating mystery in itself.

2. If the wrecks span many generations, then it may provide new and important evidence about trade between Crete, Cyprus, Turkey, and Egypt, over a broad span of time. This would be the first evidence of sustained open-water traffic in the ancient world.

3. More exciting is the possibility that one of the other targets that Nauticos discovered is a Minoan shipwreck. The Minoans ruled Crete and most of the Aegean in the Early-Middle Bronze Age, establishing a thallassocracy throughout the ancient near east, but no trace of a shipwreck has ever been located. The oldest known shipwrecks discovered date to the Late Bronze Age, at Cape Gelidonya and Ulu Burun in Turkey. Both have been excavated during the past 35 years by INA. Equally exciting would be a ship from Egypt or the Near East from the Bronze Age, or a Mycenaean or Phoenician shipwreck from the Iron Age.

Fig. 11: Artistic rendering of the Kyrenia at a Busy Ancient Seaport.
Image courtesy of


Preliminary Research Design

The investigation of these shipwrecks has the potential to be one of the most significant studies of shipwrecks to date. The investigations have the possibility of adding to the archaeological record information that has previously been inaccessible. Because these shipwrecks lie in deep water, they have not been subjected to the effects of strong tidal currents, surface wave action, high dissolved oxygen levels, or diver exploration. The structure of the vessels as well as the artifacts associated with them, have only been subjected to natural bacterial and chemical processes and artifact shifting as the vulnerable portions of the shipwreck disintegrated. As a result, it appears that the wreck sites are far more complete than any others that have been extensively investigated. While many shipwrecks have been thoroughly and scientifically investigated, all of them lie in shallow water and are exposed to the environmental characteristics associated with shallow water. In deep water, only limited investigations and salvage projects have been carried out to date. The primary purpose of our investigations will be to confirm the sonar targets believed to be additional vessels and to conduct detailed surveys of each of the known shipwrecks. All of the investigations will recover as much data about the sites as is technically feasible with the least destruction of the archaeological fabric of these shipwrecks.

A second purpose of these investigations will be to collect data that will support the theory that deepwater shipwrecks are archaeologically more complete than their shallow water counterparts. This data will be critical for continuing to make the argument that deepwater shipwrecks are scientifically worth the added cost of their investigation. This case has yet to be proven through scientifically collected data. Almost all investigations in deepwater to date have been documentary in nature.

Detailed Non-Intrusive Mapping

In order to accurately interpret each of the sites, and the possible association that they may have with each other, it will be essential to document the precise location and orientation of all objects associated with each of the them before any artifacts are disturbed. This applies to the wrecks themselves as well as artifacts lying on the ocean floor in the vicinity. For each wreck site, a survey area 500 feet by 500 feet will be established to encompass all artifacts related to the wrecks and any geologic features that have been created by the deposition of the shipwreck in the bottom.

To accomplish this task, a medium sized, heave compensated ROV system will need to be specially outfitted. That ROV will need to have a variety of remote sensing, navigation, and speciality tools and equipment installed, integrated, and tested prior to executing the expedition. The ROV will need a side scan sonar, multi-beam bathymeter, stereo cameras with lighting, and a sub-bottom profiler. The ROV will also need to have a complex navigation suite consisting of a deep-ocean transponder system with an integrated Doppler Velocity Log (DVL) and an IMU to accurately measure heading, pitch, and roll.

All sensor data will need to be precisely geo-located. To accomplish this, a high frequency bottom mounted positioning system such as Sonardyne’s Extra High Frequency (EHF) LBL system will be necessary. Sonardyne’s (EHF) system operates at 50-110 kHz with an accuracy of approximately 2-15cm. In addition, precise and high frequency platform telemetry information will be needed to accurately reduce and synthesize the sensor data. Platform telemetry information should include heading, pitch, roll, and speed over ground.

The telemetry information can be gathered using a combination of an Inertial Measurement Unit and a Doppler Velocity Log. None of these navigation sensors are usually found on an ROV, however this type of navigation system while complex has been successfully deployed on ROVs and many AUVs and other systems previously. The topside control of the ROV will need to have autopilot functions integrated into it, so that a series of planned track lines can be executed with the various sensors over the wrecks sites. Those track-lines will need to be navigated parallel, level, and straight. Initially, a high-resolution side scan sonar survey of each shipwreck will be conducted. Lane spacing will be approximately 50 feet with range scale set to 50 meters to allow for approximately 300 percent coverage of the bottom. This lane spacing and coverage will support the generation of a detailed sonar mosaic of each wreck site. That mosaic will map each of the shipwrecks and all objects external to the hull remains. The sonar survey will also map any scouring and sedimentation changes adjacent to the wrecks. A sonar frequency of approximately 500 kHz with a resolution of approximately 3” in both the along track and across track orientations will be sufficient to complete the job. The sonar mosaic will serve as an initial base map for the investigation of the shipwrecks. Second, generate a precise hydrographic map of each of the sites and the areas surrounding each of them, approximately 500 feet by 500 feet. To accomplish this, a narrow multi-beam echo sounder such a Reson 8125 should be employed. That sonar has a resolution of 1-2cm vertical with a beam width of .5 degrees. The multi-beam echo sounder should be run over each of the sites on appropriate lane spacing to accomplish a minimum of 150% coverage. The bathymetric information should be corrected to account for tidal variation in the area; this can be accomplished mathematically using one of several algorithms. This narrow multi-beam echo sounder will provide initial three-dimensional spatial mapping of the sites.

Third, collect digital sub-bottom data throughout the same 500 foot area to accurately map the geologic environment on which the airplane rests. Lane spacing should be set at intervals appropriate to map small sub-seafloor geologic features in the area. The sub-bottom profiler survey will provide information related to sedimentation in the area and the overall site formation processes.

Fourth, a stereo photo-mosaic should be created of the entire site with a minimum resolution of 1/16th of an inch. This mosaic will serve to document the overall condition of each of the shipwrecks and the location of all objects external to their hulls. The stereo imagery will also provide very accurate three-dimensional spatial information of the hulls and related artifacts. The photo-mosaics generated from this data will serve as the overall plan view of each of the site.

Environmental Information

Core samples should be collected throughout the 500 foot survey area centered on each shipwreck in order to document the geomorphology of the area and for later analysis of the site formation processes. Chemical analysis of the sediments collected during the coring, will provide information relative to the formation of the shipwreck sites and the historical interaction of the shipwreck and artifacts with the seawater and the bottom sediments. Distribution maps generated from these samples will indicate historical current patterns and deviations caused by the deposition of the shipwrecks and associated remains. Additionally, the chemical analysis of the sediments will provide information that will be significant during the conservation of any recovered artifacts.

Water samples should also be collected on, in, and around each of the sites. A sampling plan should be devised that will capture chemical and biological information for each of the shipwreck sites as well as the area around each of them. Analysis of this information will be significant during the conservation of any artifacts recovered from the sites. Diagnostic Archaeological SamplingA detailed conservation plan will need to be drafted prior to the recovery of any objects. That conservation plan should address methods to be employed, funding sources, the conservation laboratory to be used, and the ultimate disposition and storage of the artifacts. The services of a qualified conservator, familiar with the conservation of artifacts from a submerged ancient shipwreck, will be sought out to draft the conservation plan.

Diagnostic artifacts should be recovered from each of the shipwrecks to determine their age and provide preliminary information related to the ships’ origin and last port of call. Those artifacts should not be recovered until their precise geographic position and location within the wreck are mapped in detail. A properly prepared recovery basket with an emergency acoustic locating device that would provide a method to relocate the objects should they be lost during recovery, should be used to recover all artifacts from the seafloor.

In order to build the case for continuing to go to the trouble and expense of investigating deepwater shipwrecks, it will be essential to recover data from one of the shipwrecks that captures a cross section of the information that is contained in these vessel remains. To capture this data, it will be necessary to completely excavate a portion of one of the vessels, recover all the artifacts from the excavated area, and collect a series of biologic and chemical samples from the area designated for excavation. It will be necessary to record locational data related to the position of all artifacts in the designated area to a precision where the area could theoretically be reconstructed.

The test excavation will consist of a meter wide trench that will be excavated across one of the shipwrecks. The trench will be excavated in approximately 10 centimeter vertical levels and divided into 1 meter long subsections. At the beginning and end of the excavation of each level, stereo photographs will be taken of each 1 meter square. As artifacts are completely uncovered during the excavation, they will be tagged with a unique identifier indicating both their horizontally and vertically locations, a stereo photograph taken, and then they will be placed in the recovery basket. Once the entire trench is excavated a photo-mosaic run will be made along the entire trench and all artifacts remaining in the trench recovered. It is not anticipated at this time that the ship’s structure in the test trench will be recovered. Once the excavation is complete, the test excavation will be back filled with sediments from the surrounding area.

In terms of excavating the test trench, it is envisioned that in order for the ROV system to gain the stability that will be required to accomplish the complex tasks that will conducted, a light weight scaffold or bridge will need to be constructed that will span the width of the wreck. This bridge will provide a stable platform that the ROV can attach itself to while it is excavating the test trench and recovering the artifacts. A variable suction water pump that will drive a venturi dredge of 3-4 inch diameter and a highly dexterous set of 7 function manipulators will be necessary to conduct the excavation. The dredge will need to have an outflow hose long enough for the outflow end to be secured a short distance down current of the excavation site. At the end of the dredge outflow hose will be attached a mesh net to capture any small artifacts that are accidentally ingested.