Ma’agan Mikhael, late 6th–early 5th century BC
Country: Israel
Place: Off Kibbutz Ma’agan Mikhael, 35 km south of Haifa, at a depth of 2.5 m
Coordinates (approximate): 32.56, 34.91
Type: Greek merchant ship, probably from the Ionian coast of Asia Minor
Identified: No
Dated: End of the 6th or beginning of the 5th century BC, based on the pottery.
Beam: Unknown; the remains are 3.11 m wide.
Keel Length: 8.26 m.
Length Overall: Unknown; the remains are 11.15 m long.
Number of Masts: One; the mast step and two mast wedges were found.
History of the wreck
The Ma’agan Mikhael wreck was discovered by chance in 1985 off Kibbutz Ma’agan Mikhael, 35 km south of Haifa (Israel), on a low sandy coast where no archaeological finds had ever been made (unlike the waterfronts of Dor or Caesarea Maritima). It was exposed by a series of changes in the coastal dynamics (the Aswan Dam on the Nile and coastal construction) that created a negative sediment balance along this stretch of coast. As in the Gela 1 wreck, a heap of stones, among which there was pottery and fragments of wood, was the first sign of the wreck. After the first archaeological inspections, the pottery recovered gave an initial date at the end of the 6th or the beginning of the 5th century BC. In 1986 Raban and Linder investigated the site, locating hull strakes with mortises and tenons; the following year more systematic work was carried out on the wreck with the assistance of the University of Haifa (Department of Maritime Civilizations), and in 1988, with a defined team, the excavation began. It was established that the ship probably came from the Ionian coast of Asia Minor (Pomey and Rieth 2005: 162).
Description of the site
The wreck lies at a depth of 2.5 m. Given the anaerobic environment in which it lay, the timber structure (11.15 × 3.11 m) was in very good condition (Kahanov 2003: 53–129), with the whole lower part of the hull preserved (Figure 1).
Figure 1. Section of the wreck, with starboard on the right (J. Rosloff, T. Levi; after Kahanov 2003: 54).
Cargo
The heap of stones that first revealed the wreck contained pottery and fragments of wood. A basket was found with several dozen unfinished tenons, treenails and pegs of various sizes, together with shipwright’s tools such as a drill, a mallet, a ruler and a square.1
Ballast
A heap of stones covered the wreck.
Anchors
Not reported.
Guns
Not applicable.
Iron concretions
Not applicable. The frames were fastened with copper nails.
Hull Remains
The whole bottom of the hull was preserved, including the false keel, the keel (from stem to sternpost), eleven strakes of planking on the starboard side with a wale and seven on the port side, as well as two knees, fourteen frames, the mast step, a keelson, four stanchions, two mast-step supports, two mast wedges and numerous pieces still to be identified. The hull section was of the pinched, wine-glass type (Figure 1); together with Gela 2, it is one of the first two cases documenting the transition from rounded to angular, pinched hull bottoms.
Keel, keelson and mast step
The single-piece keel, of pine (Pinus brutia), 8.26 m long, had a rectangular section (11 × 16 cm) with no rabbet of any kind to receive the garboard strake; instead, the garboard strake had a bevelled edge to obtain the characteristic rise of the wine-glass shape (Figure 1). The pegs of the keel–garboard joint were driven from the inside of the hull. Those on the starboard side were 11.5 mm in diameter on average and those on the port side 9.7 mm, which may be related to the work of two carpenters with different drills (Kahanov 2003: 55, fig. 8). The false keel was made of three lengths of oak (Quercus petraea/Q. pubescens) joined together. It was fastened to the keel by mortises and tenons, the locking pegs being driven through the side faces of the keel without passing right through it (Figure 2).
The keel was joined to the stem and sternpost by a complex trait de Jupiter (rayo de Júpiter) scarf, with a large vertical locking key (Quercus sp.) secured by a peg. The stem was made of several pieces and, unlike the keel, had a notched rabbet for the garboard strake. It had a protective false stem (not preserved), which the sternpost did have (a skeg?). A sealing substance was found in the scarf joining the sternpost to the keel. The technique for joining the stem and sternpost with the curved piece of the stem was lashing or sewing, through regular tetrahedral holes like those documented in the Archaic Greek wrecks of Marseille (Jules Verne 9 and Jules Verne 7).
Over the framing lay a keelson fastened to the frames by treenails. The keelson preserved mortises for the stanchions, set at roughly equal intervals and not necessarily coinciding with the position of the frames. The mast step, similar to those known earlier, was set over the floor timbers by a series of notches perpendicular to its length.
Figure 2. False keel, keel without rabbet, and garboard strake with its side face bevelled to create the necessary rise (T. Levi; after Kahanov 2003: 60).
Frames
The frames were made from trunks of naturally curved growth (Pinus brutia) and were fastened to the hull by square-section copper nails (5.3 mm) with bent-over points, one per strake; in some cases they were driven from the outside using square wooden treenails as guides or plugs (Kahanov 2003: 51). The frames were widely spaced (75 cm on average at their base), some touching the keel and others not, but in no case physically fastened to it by nails. Their section was slightly trapezoidal, and the joint between floor timber and futtock was made by a simple scarf fastened by nails that used a treenail as a plug, unlike in the Jules Verne 7, Grand Ribaud F and Gela 2. The floor timbers have triangular limber holes in the central heel and rectangular ones at the ends of the floor timber. In the futtocks there are triangular limber holes placed very high.
Planking
The garboard strakes on both sides were single pieces, 10.4 m long and 4.4 cm thick (thinning to 2.3–2.5 cm at the stem and sternpost), without scarfs, which would have strengthened the backbone of the hull. Four copper nails were identified on the port side, perhaps to reinforce the joint with the garboard strake, without equivalents on the starboard side.
The hull strakes, covered with a yellowish resin, were of pine (Pinus brutia) and were joined by tenons (Quercus coccifera) in very closely spaced mortises. The thickness of the strakes is not regular, and their widths likewise vary according to the scarfs they have (Figure 3), but they are at least 11.3 cm (second starboard strake) and at most 32 cm (starboard strake 6). The scarfs joining two lengths of the same strake are simple bevels, diagonal (pico de flauta) or angled, similar to those of the Jules Verne 7.
In longitudinal section the tenons are thicker in the middle than at the ends, and their overall size (length, width and thickness) is slightly smaller than that of the mortises, for easier insertion. Since the Ma’agan Mikhael was dismantled into pieces underwater, the tenons had to be cut to separate the strakes, both because of the problems of underwater recovery and for subsequent conservation. For the statistical study of the tenons, 178 examples were available (28 of them complete), so the conclusions are clearly representative. The measurements (Figure 4) were based on Yovel’s working hypothesis, which was to analyse whether there was a precise correlation between the dimensions of the tenons, the parts of the hull they joined and the marked variability in the size of the strakes (various widths and thicknesses), that is, whether tenons of a particular size had been selected to join particular parts of the hull.
For this, the tenons were measured strake by strake, relating the maximum width and maximum thickness of the strake to the following measurements of the tenons: half-length, centre width, centre thickness and end thickness (Figure 5). The statistical analysis concluded that no precise relationship or pattern could be established between the dimensions of the tenons and the size of the strakes in which they were fitted; for example, starboard strake no. 6, the widest in the wreck (32 cm), did not have the largest tenons. Only in the garboard strakes were the tenons found to be larger. The longest tenons were not necessarily the widest or thickest, and vice versa. The average dimensions of the tenons in the Ma’agan Mikhael are 15.4 × 3.9 × 0.7 cm, measured on treated or conserved wood, which may involve a deviation of 1–1.5%. The most constant measurement of the tenons is the width (4 cm), while the others vary more.
In the author’s conclusion, the different sizes of the tenons in the wreck therefore seem to be determined by craft production by groups of carpenters, who made them in series to a more or less set, but not exact, format. They were then used without a precise order in the construction of the ship, except in the joint between the keel and the garboard strakes, where the tenons are larger and more robust than the rest, being intended to join larger timbers. As for the mortises and their distribution along the edges of the strakes, Kahanov (2004: 132), after studying a thousand of them, established that a common pattern was used for cutting the mortises, with an average depth of 8.2 cm, width of 4.1 cm and thickness of 7.9 mm, spaced 13 cm apart, except at the ends of the ship, where the spacing ranged from 8.5 to 27 cm. This means that there was a standard for mortises housing tenons of various formats. He also detected differences in the direction and size of the mortises at the scarfs joining lengths of strake (Figure 3), for which no technical reason could be established.
Figure 3. Patterns of scarfs between two lengths of planking strake (J. Rosloff; after Kahanov 2003: 79). In the author’s terminology, S11: simple bevel; S4: three-angled (Z) scarf; P4 and others: diagonal (pico de flauta) scarf.
Figure 4. Measurements taken on the 178 tenons studied from the Ma’agan Mikhael wreck (after Yovel 2004: 106).
Figure 5. Measurements to be taken for the statistical study (by the author, after Kahanov 2004: 126).
Ceiling, thick stuff, and wales
One wale was preserved on the starboard side.
Caulking
The hull strakes were covered with a yellowish resin. A sealing substance was found in the scarf between the sternpost and the keel.
Size and scantlings
The preserved structure measures 11.15 × 3.11 m. The keel (11 × 16 cm, 8.26 m long) has no rabbet, the garboard strakes are 4.4 cm thick, and the frames are widely spaced (75 cm).
Table 1. Scantling of the timber remains of the Ma’agan Mikhael Shipwreck
| Timber | Sided [cm] | Molded [cm] |
| Keel | 11 | 16 |
| Keelson | – | – |
| Maststep Mortise | – – | – – |
| Buttresses | – | – |
| Floor timbers | – | – |
| Futtocks | – | – |
| Room-and-space At central frames | 75 (avg., at base) | |
| Planking Garboard | 11.3–32 – | – 4.4 (2.3–2.5 at ends) |
| Ceiling | – | – |
| Main mast Ø Chock | – – | – – |
Average tenon dimensions: 15.4 × 3.9 × 0.7 cm; average mortise: 8.2 cm deep, 4.1 cm wide, 7.9 mm thick, spaced 13 cm apart.
Wood
Keel, planking and frames: pine (Pinus brutia). False keel: oak (Quercus petraea/Q. pubescens). Tenons: kermes oak (Quercus coccifera). Scarf locking key: oak (Quercus sp.).
Bibliography
Kahanov, Y. 2003. The hull. In E. Linder and Y. Kahanov, The Ma’agan Michael Ship. The Recovery of a 2400-Year-Old Merchantman. Final Report, vol. 1, 53–129. Jerusalem/Haifa: Israel Exploration Society and University of Haifa.
Kahanov, Y. 2004. Mortise analysis. In J. Tresman (ed.), The Ma’agan Michael Ship. The Recovery of a 2400-Year-Old Merchantman. Final Report, vol. 2, 125–132. Jerusalem/Haifa: Israel Exploration Society and University of Haifa.
Pomey, P. and Rieth, E. 2005. L’archéologie navale. Paris: Errance.
Yovel, I. 2004. Tenons. In J. Tresman (ed.), The Ma’agan Michael Ship. The Recovery of a 2400-Year-Old Merchantman. Final Report, vol. 2, 105–111. Jerusalem/Haifa: Israel Exploration Society and University of Haifa.
Notes
1 According to Kahanov, this would indicate that work on the ship had not yet been completed (Kahanov 2003: 53).