The most complete description of the Beothuk canoe was in the manuscript of Lt. John Cartwright, R.N., who was on the coast of Newfoundland in 1767-1768 as Lieutenant of H.B.M. Ship _Guernsey_. However, some portions are either in error or the description was over-simplified. For example, Cartwright says that the gunwales were formed with a distinct angle made by joining two lengths of the main gunwale members at the elevated middle of the sheer. This hardly seems correct since such a connection would not produce the rigidity that such structural parts require, given the methods used by Indians to build bark canoes. The three grave models show that the sheer was actually curved along its elevated middle. It is possible that Cartwright saw a damaged canoe in which the lashings of the scarf of the gunwales had slackened so that the line of sheer "broke" there. Cartwright is perhaps misleading in his description of the rocker of the keel as being "nearly, if not exactly, the half of an ellipse, longitudinally divided." The models show the keel to have been straight along the length of the canoe and turned up sharply at the ends to form bow and stern. Cartwright also states the keel piece was "about the size of the handle of a common hatchet" amidships, or perhaps 1 inch thick and 1½ inches wide, and tapered toward the ends, which were about ¾ inch wide and about equally thick. The height of the sheer amidships was perhaps two-thirds the height of the ends. [Illustration: Figure 86 A 15-FOOT BEOTHUK CANOE OF NEWFOUNDLAND with 42½-inch beam, inside measurement, turned on side for use as a camp. It gives headroom clearance of about 3 feet, double that of an 18-foot Malecite canoe with high ends. When the ends were not high enough to provide maximum clearance, small upright sticks were lashed to bow and stern. The shape of the gunwales would permit the canoe to be heeled to an angle (more than 35°) which would swamp a canoe of ordinary sheer and depth. (_Sketch by Adney._)]