Mestre Walter ‘Waltinho’ Assis de Santana’s saveiro
John Patrick Sarsfield described the construction of one saveiro by mestre Walter “Waltinho” Assis de Santana in 1984. The ship size was defined by the keel length, which sometimes depended on the size of the tree available and the will of the client, and in his case was around 15 m. After shaping the keel, which had no rabbets, mestre Waltinho assembled the stem and sternposts, shaped by eye, generally following the input of the ship’s future owner. Sarsfield did not detail the manner in which the posts were fastened to the keel, but mentioned the existence of corais (inner knees), which are curved timbers that reinforce the connections between the keel and the posts, already mentioned in Lavanha and Fernandez early 17th century manuscripts (Lavanha 1996, Fernandez 1989). Although there were no rabbets on the keel, the corais were rabbeted, and Sarsfield mentioned the existence of an inner stem post in mestre Waltinho’s boat, which enlarged the area which received the hoods of the hull planks.
Once the keel and posts were set, the master frame was designed. To achieve a fair shape, the shipwright defined the longitudinal curve of the deck (sheer line) by eye, and set its height at the middle of the keel and the posts. The added height of the sheer at the stem post was called tosamento. The sheer line at the stem post was higher than at the sternpost, and both were higher than the depth in hold measured on the middle of the keel, where the master frame was placed. The maximum beam was then established as one-third of the length on deck, which is loosely measured, generally from the inner face of the stem to the inner face of the stern panel. Sarsfield writes that the height of the deck above the center of the keel was about two-fifths of the maximum beam. Once the dimensions of the rectangle that defined the master frame were set, the frame was traced in a way similar to those recorded in late 16th century shipbuilding treatises, such as the circa 1580 Livro da fabrica das naus by Fernando Oliveira (1991).
The beam AB was divided in two by a center line CD and each side subdivided in two: AE, ED, DF and FB. The flat of the floor was set at one-half of the beam EF and the frame was then traced out with a batten, tangent to the flat line EF, and slightly extending into the flat zone. The turn of the bilge points were therefore located on the curve, slightly above points E and F.
As in the Mediterranean and Iberian methods, two molds are cut from the master frame shape, one for the floor timbers and another for the futtocks. With the help of a sliding scale, the graminho, engraved on a board – named tábua – these molds were then used to define the gradually changing shapes of the ship’s frames.
The frames that defined the narrowing portions of the vessel’s bottom, near the bow and the stern, were cut afterwards, in accordance with the shapes of the entries and runs, which were defined with the help of ribbands. In Portugal these end frames were called enchimentos. In Valença we also heard the word enchimentos, although we were told that in the past these frames were called cavernas de pé in the stern and muganzas in the bow, these last sometimes composite, assembled from three timbers (mestre Francisco ‘Chico’ de Assis Belém pers. comm.)
The first step of the process that defined the central frames was to establish their total number and fix the room and space value. Sarsfield states that this number was established through “what seems to be an iterative mental trial and error process” (Sarsfield 1985a: 5). In 2013 mestre Waltinho told us that he divided the keel in four parts and determined how many frames he could fit in the two central portions.
The following step was to determine the total narrowing and rising of the turn-of-the-bilge curve, measured on the last predesigned fore and aft frames, which Sarsfield refers to as almogamas, the Arab name mentioned in 16th century Portuguese shipbuilding treatises. In our conversation with mestre Walter, however, he was adamant that he had never heard the word almogama in his life. This was corroborated by another four shipwrights interviewed in Valença and Camamu.
In this particular case, the total narrowing of the turn of the bilge point was equal to its total rising, and this single value was used to narrow and rise this point both fore and aft, as described in c. 1600 Lavanha’s manuscript, Livro primeiro de arquitectura naval (1996). Also as in Lavanha’s manuscript, in mestre Waltinho’s case the narrowing applied to each side was one-sixth of the flat of the floor. This value is high when compared with those used on river craft, and suggests that mestre Waltinho was referring to a hull intended to sail in the sea.
The third step was to distribute the total rising and narrowing of the turn of the bilge curve along the frames, and this was done with the graminho. As in Portuguese documents, this graminho was made using a geometric construction known in the 16th century as método da besta (crossbow), the equivalent of the 15th century Italian mezzaluna, described in the Trombetta manuscript (Anderson 1925; Oliveira 1991, Castro 2007) and presented in the figure.
The fourth step was to mark the shape of the floors and futtocks on the timbers, a process that was done by sliding the molds with the help of the graminho, which is marked on a tábua as wide as the keel, in order to obtain increasingly narrower and higher frames as we move away from the master frame. Once the central frames were traced, cut, assembled, and set in place, the hull shape was defined with the already mentioned ribbands, and the remaining frames cut and fastened to the keel.
The fifth step was beveling the frames’ outer faces so that they presented a fair and smooth surface upon which to fasten the hull planking. This was a complex operation, which shipwrights controlled in a masterful way with yet another graphic scale – marked on the same tábua as the graminho – and a bevel gage: the suta, or açuta, as Sarsfield writes. According to him “no dubbing or correcting of these bevels was ever observed once the frames were in place on the keel” (Sarsfield 1985a: 7). The frames were beveled both on their inner and outer faces, to receive the hull planking on the outside, and the longitudinal structure on the inside – stringers and clamps.
A sixth step consisted of beveling the bottom of the predesigned fore frames in order to tilt them forward – caimento – and reduce the narrowing of the deck determined by the narrowing to the turn-of-the-bilge line. In Sarsfield this beveling was done using the beveling scale, but using the beveling angle of the previous frame, a measure that reduced the inclination of the frames forward. In 2013 we observed the use of separate scales for the beveling of the frames’ surfaces and bases. Both mestres José do Nascimento and Edir showed us tábuas with separate scales.
John Patrick Sarsfield explained the basic construction process clearly and in detail, and published a number of drawings that further clarify the text and served as a basis for some of the illustrations presented here. After all the frames were fastened to the keel, the structure was reinforced with a keelson (sobrequilha), stringers (escoas), clamps (dormentes), a main wale (cintado), after which the deck was built and the structure planked. He died tragically, however, before completing his work, and detailing the types of timbers used and their dimensions, the scarves and fastening patterns, and the variations on the conception and construction sequences. We should mention here that although Sarsfield died more than 20 years ago, in 1990, he is still remembered by the entire community with warmth and admiration. This paper is intended as the first of a series aiming at completing his work.