Journal of Archaeology in the Low Countries 4-1 (October 2012)Thijs Maarleveld; Alice Overmeer: Aanloop Molengat – Maritime archaeology and intermediate trade during the Thirty Years’ War1
2 The Aanloop Molengat ship

2.1 Ship construction and internal organisation

The Aanloop Molengat investigation placed greater emphasis on the cargo than on the ship itself. Nevertheless, quite a few characteristics of the ship can be deduced. It sits more-or-less upright and the superstructure is missing. The composition of the find assemblage suggests that large parts broke away and were not deposited on site. Only heavy cannon were left behind, eleven in all. The lower hull is kept in place and hidden from view by the cargo mound, notably the concreted mass of iron. In consequence of the decision not to move or remove this, all observations relating to the ship’s construction have been made along the wreck mound’s sides. The lower hull extends over a length of 26.5 m and seems to be continuous and relatively straight. It is approximately 8 m wide. Observations at the northwestern end suggest that it has a slight list to the northeast, and that it may have broken lengthwise along the keel. This would explain the wide lengthwise crack in the mass of closely packed iron bars.

The keel is a solid beam of oak. At the northwestern end it protrudes some 3.4 m from the cargo, where it ends in a vertical scarf. At the other end, it could not be identified. It must, however, have extended over the full 26.5 m. At the broadest point observed, where it was cut for a dendrochronology sample, it is 44 cm wide. It is almost square, with a depth of 41.5 cm. It tapers towards the end where it still has a depth of 38 cm and is likely to widen in the other direction. To receive the garboards, a single rabbet has been cut on either side, at 5 to 6 cm below the top and rising forwards (fig. 10). A single square nail hole indicates that the garboard had been attached with square nails. Rows of two or three nail holes, spaced at about 13 to 17 cm intervals, occur on the sides, but have not been observed underneath. The nail holes are thought to reflect a layer of sheathing, probably in lightwood, as no metal has been observed. The end scarf is 33 cm long. One iron bolt and two iron spikes secured it to another construction element. In view of the tapering end and the rising rabbet, this could hardly have been another length of keel. Neither is it a plausible solution for joining sternpost and keel. The scarf, which faithfully matches Witsen’s (1671, 147-151) description of a stem scarf, almost certainly joined an apron or lower stem section. This implies that the northwest end of the wreck is the bow.


Figure 10 The raised fore-end of the keel. The sample is 2.6 m long (drawing: L. van Dijk (RCE)).

Following from the identification of the ship’s orientation, it can be inferred that the sternpost and deadwood broke away, taking some length of keel with them and breaking it loose in such a way that it is now missing from the otherwise fairly contiguous shell of bottom planking. Despite the list to the northeast, the southwestern (or port) side of the ship is best preserved. The sea bottom is highest here and amidships the ship’s side is continuous to include part of the turn of the bilge.

At the northwest or fore end, six (or seven) adjoining hull planks are observed, all ending in abrasion. They do not directly adjoin the keel. Either the garboard strake is missing or the keel has shifted at this end. The planks have widths of 25 to 44 cm. At the southeast or aft end, thirteen hull planks protrude from under the cargo. At the portside, two of them end in a scarf joint that is more than 60 cm long. The orientation of these scarves does not affect the interpretation that this is the stern. Plank widths vary from 28 to 38 cm. Their thickness has not been systematically recorded, but is at least 7 or 8 cm. It seems to be a bit more in the bilge along the side. The planks are flush; it is a carvel hull. Wood of a lighter colour has been observed on the outside of the longest hull plank aft. It is approximately 25 cm wide and may well be part of lightwood sheathing, such as is also inferred for the keel. Otherwise, all planks and timbers seem to be oak.

Transverse timbers are visible at the southwest edge (port side) where erosion has produced a smooth cut through the hull structure and where they stand together, almost without any spaces (fig. 11). Other frame timbers are visible at the northeast or starboard side, where alternating timbers have broken away. Room-and-space is irregular, but seems to average around 40 to 45 cm. Sided dimensions of the timbers vary from 14 to 25 cm, with an average of 19 cm, which is also approximately their moulded dimension. Some of the timbers have original endings, others continue in a sharp curvature. Although the alternating timbers are placed closely together in the bilge, no indications for interconnection have been observed.


Figure 11 Planking, frame timbers, ceiling and the rim of the concreted cargo along the southwest side of the wreck-mound (photo: P. Stassen (RCE)).

Internal structural planking was attached to the inside of the timbers. The widths of these ceiling planks vary from 33 to 50 cm. Their thickness is unknown, but apparently similar to the planking. They are flush-laid without any spaces, providing a closed ceiling, fastened with treenails of 3 to 3.5 cm in diameter.

In the sandwich of planking, timbers and ceiling, the timbers are relatively light. Their dimensions and spacing (17 to 18 elements at the bilge over 4 m of ship’s length) compare well with what we know of early modern merchant ships that are built shell-first in the Dutch-flush manner (Maarleveld 1984, 67; Maarleveld et al. 1994; Lemée 2006). Although this ship is larger than other examples, framing is not more substantial, and planks and ceiling may be slightly thicker (only B&W2 compares well in that respect; Lemée 2006, 200ff.).

The dendrochronology of the keel suggests that it came from Westphalia, a region in northwest Germany, roughly between the rivers Rhine and Weser (Jansma & Spoor 1991). This is consistent with a building spot in the Northern Netherlands, where practically all timber was imported, but it could of course point to northern Germany as well.

A striking issue is the absence of a clear indication for the presence and position of a mast. The mainmast cannot have been positioned where the wrought-iron bars lie. Close scrutiny of the site plan (see fig. 2) shows that the forward portion of concreted iron ends in a straight line at right angles amidships. This is most obvious between metres 14 and 15 on starboard. The straight line is likely to reflect the original presence of a bulkhead, behind which the mast was stepped, in an area that is characterised by a jumble of concreted cannon balls that have filled the empty space where the mast once stood. It might even be the case that the wrought-iron is not continuous, but leaves a cross-ship corridor of about one metre unoccupied. To what extent this would affect the indicative calculations of cargo weight remains to be seen. If a corridor is present it seems to be filled with cannon balls, perhaps in cases, and cast-iron is only fractionally lighter than wrought-iron or steel.

Apart from the inferred bulkhead amidships, which may have been a temporary structure, a bulkhead seems to have been present at the forward end of the iron bars. After all, here the bars end in a more or less straight line as well, possibly with some shifting forward in the middle. At the aft end, just below the barrels of tin, the very residual remains of a decayed bulkhead have actually been observed. The overall length of the hold between these two bulkheads was 19.5 m. The depth of hold cannot be inferred. Apart from the cannon, no remains or evidence on the orlop and upper structure was found on site.

Reasoning from the coherent structure as shown in fig. 2, the ship can hardly have been narrower than 8.5 to 9 m. To the preserved length of 26.5, one should add some 7 m or more for apron and stem. Moreover, a substantial part of the stern is missing. How much depends on the presence and form of a transom. The bottom is still wide at the aft end, but individual strakes are not, which might indicate that the hull narrows at this point. Nevertheless, it would be hard to fit the continuing hull, deadwood, sternpost and rudder in less than 3 m. The overall length cannot have been less than about 37 m. Another approximation of size is its cargo capacity. Even though the ship sank, it is not reasonable to estimate its capacity at less than the total weight of the goods it carried. The absolute minimum weight of the cargo as found is estimated at 550 tons, while more than 600 tons is more likely.