Ship Mainz 3

See this text in

Discovery and state of preservation
Constructional details

Discovery and state of preservation

Ship 3 was found together with the other four ships in the winter of 1981/82 in the construction trench of the Hilton Hotel II in the Löhrstraße in Mainz. It was broken into two pieces, whereby the port had survived from the sternpost to the mast-frame. The uppermost rows of planks at the stern, however, were absent.

Amidships the three lowest rows of planks of the starboard lay connected with the keel and the port planks. The right side of the stern was completely detached and lay next to the hull. As a result, the 2nd, 3rd and 4th rows of planks were destroyed so much, that the two pieces of the wreck were no longer connected to one another. However, with the aid of the steering system the correct position of the starboard stern could be determined: in the two side-frames of the stern there is still the recess for the rudder-beam. This originally ran vertical to the keel and penetrated the hull at the same height. Thus, the position of the detached part of the stern can be absolutely determined in its height and length. The reassembled wreck today measures 13×80 m long and approx. 3×40 m wide. The height at the stern is 1×30 m, amidships approx. 1 m.


The dendrochronological analysis of the oak yielded fell dates from the end of the 3rd century. Thus, Ship 3 seems to have been constructed earlier than ships 1,2,4 and 5. According to the archaeological situation, however, it sank at the same time as the other late Roman ships from Mainz, i.e. at the outset of the 5th century.


Wreck 3 belongs to a different type of boat than ships 1,2,4 und 5. This Type B is shorter, wider and higher than Type A and appears generally more compact. Nevertheless, we are still concerned with a military ship, not a cargo boat. This is indicated by the very narrow stern and the ratio of length to width of 5:1, which is not typical of cargo boats. However, this craft only allowed room for seven rowers on each side and in addition to its greater weight was consequently slower than Type A. When necessary it could be sailed, as proven by the surviving mast-frame. Ship 3 served very probably as a patrol ship for surveillance of the Rhine border during the 4th century. An impression of the original appearance of this type is given by reconstruction Mainz A in the Museum of Ancient Shipping in Mainz.


Constructional details


Wreck 3 displays not only constructive elements common to the other late Roman ships from Mainz, but also peculiarities. For instance, like the other ships (1,2,4 and 5) it has 2 - 2×5 cm thick, 20 - 30 cm wide carvel-built planks lacking any mortise-and-tenon joints. Of the 9 rows of planking in all, the 4th and 9th rows can be termed stealers, i.e. both rows of planks do not reach the ship’s ends. A series of scarfs could be seen on Ship 3. For instance, Z-shaped scarfs in the rows of planks port 3 and port 4; butted joints are discernible once on row port 3 and twice on row port 6. A series of repairs in the rows of planks port 3 (twice), starboard 1 and starboard 3 indicate that Ship 3 was in use for a long time. A long crack in the row port 5 was sealed in ancient times with caulking material.



The 28 cm wide keel with a ‘bilgerinne’ along the middle displays marks at intervals of approx. 33×5 cm, with which the ancient shipwright indicated the position of the frames. This measurement can also be found in the proportions of the ship and roughly corresponds to 1 pes Drusianus, the usual measure of length in the Gallo-Germanic region during the Roman period. At the aft end of the keel was attached the sternpost, comprising a bent piece of timber. Although broken, it has survived to the height of the bulwark. The scarf at its upper end indicates that a protrusion was fastened here, which terminated in a decorative element.



A special feature is represented by the 10 - 12 cm wide and 5 - 6 cm thick frames of Ship 3. In contrast to ships 1, 4 and 5 with their side-frames and floor-timbers and Ship 2 with its half-frames, here almost without exception connected, continuous frames were constructed. These are jointed in the area of the bilge or keel and display three limber-holes.
Only in the narrow stern were tripartite frames of two side-frames with pommel-ends and a floor-timber employed. What is more, the final frame consists of only two side-frames. The frames resting upon the sternpost have a recess on the underside, so that the frames from above can clasp the sternpost.

On its rearside the mast-frame displays a rabbet as support for the bottom planking. It is interrupted at the mast-step, the area of the keel, therefore, seems to have lain open. As in all late Roman ships from Mainz, iron nails provided the bonding between frames and planks on Ship 3, too.


Indications of a deck

Although no interior fittings survive, except for a few remains of a ceiling of coniferous wood, there are some indications for them. Thus, remains of interior nails found as rows each side of the keel , i.e. where according to the mast-frame there were no bottom planks, as well as a row of nails set close to one another just above the bilge , suggest a substructure for supporting a deck. For precisely the ceiling-board above the bilge, proven by the row of nails, cannot have been a thwart clamp, such as we know from ships 1 and 4; for that it sits too low in the hull. A further important indication for a deck is provided by the starboard frames at the stern. Just below the top of the bulwark they displayrecesses and remains of interior nails .

Here, too, a ceiling-board was attached, which, however, so far astern cannot be connected with the rowers’ benches (thwarts). It probably supported the transverse timbers for a deck completely covering the stern. However, amidships this must have been open at the sides for the rowers, since, if they had rowed beneath deck, the gunwales must have been provided with oar-ports for the oars. This is, however, not the case with Ship 3. The rowers, therefore, must have sat so high, that they could manoeuvre the oars over the sides.


Indications of an outrigger

The surviving starboard side at the stern lies considerably higher than at port amidships. There are 9 planks at the stern, but only 8 sets of planks amidships, whereby the uppermost plank with a partly surviving original surface is extremely narrow. This piece of plank lies in the axis of the frame recesses and interior nails in the stern mentioned above and which can be connected with the deck. This means that the hull’s side amidships only reached the height of the deck.

Beneath this piece of plank there is no indication of a thwart-ceiling, as proven for Type A of the Mainz ships in wrecks 1 and 4, nor for a covering-board, recorded in Ship 1, nor for separate rowlocks, as assumed for Ship 2 and 4. Since Roman military vessels, in contrast to merchant ships, were generally rowed, in the case of Ship 3 one has to presume a rowing apparatus which has left no traces on the bulwark. This is only possible with an outrigger, with the aid of which the tholes were set outboard beyond the hull. A series of ancient depictions of warships shows such outriggers. Since both the construction of the outrigger, as well as the rowers’ benches could rest on the joists, a reinforcement of the hull and the provision of a thwart clamp, such as is necessary for boats without outriggers, is superfluous. Reconstruction 2 allows more precise insights into this construction.


Ship 3, which is preserved with artificial resin, is exhibited in the Museum of Ancient Shipping in its state afloat.

Barbara Pferdehirt

Translation by Clive Bridger, Xanten

To see all images of this ship:

Back to the homepage of NAVIS