The umiak shown is 20 feet 8½ inches over the headboards, 4 feet 9½ inches extreme beam and 17⅜ inches depth--apparently an average-sized boat of her class. The width of the bottom over the chine members is 2 feet 7 inches. The keelson is rectangular in section and in two pieces, hooked-scarphed together; each piece is shaped out of the trunk of a small tree with the root knees employed to form the bow and stern posts. The floor timbers are quite heavy and support the chine members by having the floor ends tenoned into the chine pieces. At bow and stern the chines are joined to the keelson in a notched scarph; at these places the keelson is sided rather wide to give good bearing. It is evident that this portion of the boat's structure is the first built and forms a rigid bottom to the hull. The floor timbers are lashed to the keelson by lacings of sinew, whalebone, or hide, passed through holes bored in both, as indicated in the plan. The ends of the floors are pegged where they tenon into the chines and the ends of the chines are pegged to the keelson, but this was evidently not a universal practice, as there are models showing lashings at floor ends and at chine ends. The headboards are carved out of blocks in a ~T~-shape and are stepped on top of the stem and stern posts and lashed. The fit is extremely accurate. The bow headboard is narrower athwartship than the stern headboard. The detail of the hook scarph in the drawing shows a method of lashing that is widely used. [Illustration: Figure 167 SMALL UMIAK FOR WALRUS HUNTING, west coast of Alaska, 1888-89. Reconstructed from damaged umiak formerly in U.S. National Museum, and from models.] Because of the manner in which the keelson is cambered and the floor fitted, the bottom of the covered hull shows in cross section a slight ~V~, reducing toward the bow and stern, that is typical of the Alaskan umiak. The amount of deadrise seems to have been determined by the manner of fitting the floor timbers and it helps the boat to run straight under paddle and oars. In present day umiaks the amount of ~V~ in the bottom is slight; too much would make the boat difficult to sledge overland without employing chocks to steady the hull. Perhaps in the past, where sledging was not required, the deadrise was greater, as indicated by some old models.