Page 224 - Nicolaes Witsen & Shipbuilding in the Dutch Golden Age

Chapter Three
Figure 3.1. Plate LXXI. A pleasure yacht 42 feet long. In Witsen’s drawing it is striking that no frames are depicted; the shape of the hull was probably supposed to follow from the data in the contract.
but the tors ion in the bil ge planks in the bow and stern was extremely difficult. If they came out too level, they would form an ugly angle w ith the s ide planks at a l ater stage. If ma de too vertical, the resul t was even worse: a nasty hollow would remain visible. This stage had to be tried several times. It soon became clear that only experi- ence could lead to a satisfactory result, and each attempt was a lot better than the previous one.
Measuring proved to be t edious. There are no str aight lines in suc h a vessel. Witsen describes how a line was stretched from stem to stern, from whic h measurements were taken. I tried the same approach in this experiment, but it yielded but few hard results because of the impossi- bility of using a plumb line and a w ater level to go with it.
The problems ended once the bilges were planked and the master ribband was fitted at the right positions (be- cause the master ribband in this case coincided with the wale, the w ale itself could be used a s a m aster ribband that would not have to be remo ved later on). The shape of the top timber s was pr actically the s ame all over, dif- fering only in the bow and the stern; their sh ape had to be established experimentally. The ceiling on the in side then provided enough strength for the hull to be planked. Common sense and careful study of prints showing simi-
Figure 3.2. Preliminary drawings of the pleasure yacht. Prior to building the model, I made these drawings to elaborate on the contract data. The resulting model was 58 centimeters in length. (Drawing by A. J. Hoving)
lar craft supplied the rest of the information necessary to complete the construction.
The most striking thing about the experiment was the speed with whic h a model was built with the shell-first method. It was of c ourse a small, s imple craft with rela- tively few frames, but even then it was surprising that only a couple of hours of work yielded a structure that revealed the eventual shape of the ship . Clearly Witsen’s state- ment about the speed of Dutch shipbuilding (four months to build the 134-foot pinas with twenty-two workmen; five months to build a 175-foot man-of-war with fifty workmen) was no idle boast.
All other fe atures of the shell-first method were pres- ent—the nearly flat bottom, the angle in the bil ges, and the relatively arbitrary length of the frame timbers, which remained unconnected to one another and were fastened only to the ceiling and planking; there were also traces of where the floor planks and bilge planks had been tempo- rarily joined with chocks spiked in place.
Subsequent checks revealed that the hull was not per- fectly symmetrical, the resul t of faulty measuring. But in this respect the model followed the example of the seven- teenth century, a time when almost all ships were asym- metrical in one way or another.
206