In fall 1985, while diving along the shore at Kibbutz Ma῾agan Mikha῾el, 32 km (20 mi.) south of Haifa, Israel, Ami Eshel noticed, in water less than 2 m deep, a pile of stones not native to the Levantine coast. Pieces of wood and pottery sherds protruded from the stones. Shifting local sand and a diver's keen sense of observation resulted in the discovery of the remains of an ancient merchantman.

Preliminary investigations were carried out soon after the discovery by Shelley Wachsmann, under auspices of the Israel Antiquities Authority. Based on the examination of an oil lamp and storage jar handles, a tentative date of the late fifth century BCE was assigned to the finds. In the following year, during further exploration by Elisha Linder and Avner Raban, several hull planks were observed while a test trench was being cut. The timbers were in an excellent state of preservation because they had been buried in the sand, an anaerobic environment, and protected under a thick layer of stones that turned out to be part of the ship's ballast. Excavation was undertaken by the Center for Maritime Studies of the University of Haifa, under the direction of Elisha Linder.

Survey and Excavation Techniques.

To determine the extent of the remains, define the ballast area, and measure the depth of the sand that had accumulated over the site, a probe pipe was inserted into the sea bottom by means of a water jet, which disclosed a clay layer about 4 m below the present sea level. Test trenches aided in locating the vessel's endposts, which were nearly 8 m apart. For recording purposes, datum poles were inserted to define the plan of the site; they remained in place throughout the excavation to facilitate triangulation measurements in recording artifacts.

The ship lay in shallow water at a walking distance of less than 60 m from the shoreline. The surge made working conditions difficult, as sand continually drifted back into the excavation areas, affecting visibility. To prevent interference, a horseshoe trench was dredged around the site, its outside wall lined with some fifteen hundred sandbags, but that failed to eliminate the surge and sand migration; even waves less than a meter high made it impossible to work safely and to avoid movements that would endanger the delicate procedures. In three seasons of excavation (autumn 1988, spring and autumn 1989), out of a total of 160 days at the site, sea conditions allowed only thirty-two days of actual excavation of the hull and its contents.

The first task was to remove the ballast stones, whose total weight was more than 12 tons; some of the stones weighed more than 50 kg (100 lbs.) each. They consisted of three lithological groups: metamorphic, magmatic, and sedimentary rock. The first group, comprising 65 percent of the total, was mostly blue schist. The origin of the stones pointed first to the central Mediterranean seacoast, near Corsica and Calabria. Closer analysis determined a more likely location, however, farther east, in the Greek islands. Beginning at the offshore end of the ship, which the finds indicated was the stern, where the galley would normally be located, the excavation proceeded slowly toward the bow, using a dredger operated by a powerful water pump. Because damage by storm waves was feared, only one section of the inner hull was excavated at a time. Its contents were recorded, mapped, and photographed before they were removed.

Hull Structure.

The vessel's original measurements were approximately 12 m long and 4 m wide. More than 11 m of its bottom were recovered, including its keel and the lower portions of its stem- and sternposts. The starboard side was found in a better state of preservation than the port side, rising at its highest to more than a meter above the bottom of the keel. The ship was constructed in the shell-first method: the hull was completely built before the internal frames were installed. The planks were attached to one another by a series of pegged mortise-and-tenon joints. Surprisingly, however, the ends of the planks were also lashed to the end posts by cords. Other structural components included full frames secured to the planks by copper nails, a longitudinal stringer supporting vertical stanchions, and a massive mast step, forward of the middle of the ship, that carried a single mast for a large square sail. The lower end posts were attached to the keel and consisted of a single timber nearly 8 m long with a false keel, or shoe. All of the hull's structural components were of pine, except the tenons, the treenails, and the false keel, which was of oak. The wood was in an excellent state of preservation, the external timbers unblemished by erosion or teredo worms. The ship appeared almost new and may therefore have been on its maiden voyage when it sank.

Ship's Contents.

In addition to the heavy load of ballast and a great deal of dunnage (twigs and branches that buffered the ship's hull from the ballast stones), an impressive array of artifacts was intact in the hull or scattered around it. A complete one-arm anchor, made of oak, with a lead-filled wooden stock was discovered off the ship's starboard bow. The anchor's shank and arm were carved from a single timber. Remains of rope were found around its crown and lifting loop. The find is unique and thus essential to understanding the evolution of the anchor in antiquity.

Among the ceramic items were basket-handle storage jars; utensils the sailors used daily, such as jugs, plates, lamps, and a water “pithos”; and miniature juglets and cups or painted pottery that were probably the sailors' personal belongings. Also in the latter category are three beautifully carved wood artifacts: a heart- or leaf-shaped “jewel box” with a swivel top and two violin-shaped “cosmetic palettes.” A collection of carpenter's tools consisting of wood handles belonging to cutting and boring tools, parts of bow drills, a carpenter's square, scores of treenails and tenons, and a whetstone, was found scattered around a woven basket that was probably the shipwright's tool kit. A few meters of rope, in widths varying from 2-40 mm and in a three-strand left-handed twist, were found intact, in undisturbed bundles, with some knots still detectable. Grape and fig seeds and olive pits were identified, as well as barley. A preliminary study of the pollen samples indicates an eastern Mediterranean coastal variety of plants that grow during the summer. The metal finds included what seemed to be a small copper incense scoop and copper and iron nails. A single lead ingot was recovered that was probably a source of raw material for use on board.

Shipbuilding Technology.

Only a very few hull remains have been discovered in the Mediterranean in a reasonable good state of preservation from the general same period as the Ma῾agan Mikha'el ship. Two are of particular interest: the late fourth-century BCE merchantman excavated and raised from the sea at Kyrenia, Cyprus, and the early sixth-century BCE shipwreck found near the coast of Gela, Sicily, which is still being explored. Considering the many similarities and some of the basic differences among these ships, it now appears that the Ma῾agan Mikha'el ship is a missing link in hull construction between the other two vessels. One example points to the extensive use of lashing the timbers by means of cords in the Gela ship, a method that was only partially applied in the Ma῾agan Mikha'el ship and disappeared altogether in the Kyrenia vessel. The conservation process of the Ma῾agan Mikha'el hull timbers and perishable artifacts lasted three and a half years. Plans are underway for the ship to be reconstructed and displayed at the University of Haifa, in a newly built museum hall.


The Ma῾agan Mikha'el ship plied the Mediterranean in about 400 BCE, a peak period of maritime activity and trade by the Phoenicians and Greeks. In its mode of construction it follows two Mediterranean traditions: mortise-and-tenon joints and cord lashing. More than half of the ship's displacement was comprised of ballast stones not native to the Levantine coast. Whether these were loaded at the port of departure or picked up en route, in a secondary use, is not known. The pottery remains one of the main clues by which the ship is dated and its sailing route traced. Presently, the leads point to Cyprus and Eastern Greece. Who carved the ship's unique wood anchor and where parallels for the carved wooden “boxes” are to be found are not yet known. Because it seems to have been on its maiden voyage, it is possible that the vessel itself was the merchandise. Further investigation and the results of laboratory testing may be able to fill in the existing information gaps.


  • Eiseman, Cynthia J. E., and Brunilde Sismondo Ridgway. The Porticello Shipwreck: A Mediterranean Merchant Vessel of 415–385 B.C. College Station, Texas, 1987.
  • Frost, Honor. “The Punic Ship: Final Excavation Report.” In Atti della Accademia Nazionaledei Linecei, supp. 30. Rome, 1976.
  • Linder, Elisha. “The Ma῾agen Michael Shipwreck Excavation: First Season.” Center for Maritime Studies News (Haifa University) 16 (1989).
  • Linder, Elisha. “Excavating an Ancient Merchantman.” Biblical Archaeology Review 18.6 (1992): 24–35.
  • Rosloff, Jay. “The Ma῾agan Michael Shipwreck Excavation, 1989 Season.” Center for Maritime Studies News (Haifa University) 17 (1990).
  • Rosloff, Jay. “A One-Armed Anchor of c. 400 B.C.E. from Ma῾agan Michael Vessel, Israel: A Preliminary Report.” International Journal of Nautical Archaeology 20.3 (1991): 223–226.
  • Steffy, J. Richard. “The Kyrenia Ship: An Interim Report on Its Hull Construction.” American Journal of Archaeology 89.1 (1985): 71–101.

Elisha Linder