74 BUREAU OF AMERICAN ETHNOLOGY [Bull. 192 



been argued that the former was the older type (de Laguna, 1947, 

 pp. 110 f). The shed-type house of the southern Nootka, Makah, 

 and Coast Salish, of course, interrupts the distribution of the gable 

 roof. Because the northern and central Nootka used only the single 

 ridgepole, and the southern Kwakiutl favored it over the double 

 ridgepole (Drucker, 1950, Traits 308 and 311), while the Heiltsuk 

 (northern Kwakiutl) of Roscoe Inlet used the single ridgepole in the 

 early historic period (Drucker, 1943, fig. 24), we may assume that the 

 double ridgepole was first used for plank houses by the Tlingit, Haida, 

 or Tsimshian. In fact, the multifamily house requiring or associated 

 with this type of construction seems to have been diffusing from this 

 center, and the evidence from Yakutat indicates that it had already 

 reached this northern area at least by late prehistoric times. It 

 would be natiural to relate this diffusion to the spread of Tlingit social 

 and ceremonial customs, which was still in progress in the protohis- 

 toric period (see pp. 9-10). 



Unfortunately we have no early archeological evidence from 

 Yakutat of the small house with single ridgepole. House 9 is the 

 only dwelluig that suggests this type and it is later than House 

 Pit 1. However, the bathhouse (or birth house). House 8, is so 

 similar in construction, that we may safely infer that the small 

 house with single ridgepole, and probably without bench or separate 

 sleeping rooms, was being built at the same time as the large lineage 

 house. The small house for one or two families was also known from 

 the Eyak, and may anciently have existed all over the northern 

 Northwest Coast before the development of the more elaborate 

 larger dwellings, just as it seems to have lingered on for poor people 

 even in regions where the latter multifamily house became common. 



Boxlike sleeping rooms with wooden walls would seem to be also a 

 relatively late development, and indeed in most Northwest Coast 

 houses only the chief and members of his immediate family usually 

 enjoyed such privacy. Ordinary people seem to have been content 

 with a low board or partition of mats, or piled up their belongings to 

 mark off their sleeping places (Drucker, 1950, Traits 345 and 346, 

 p. 251). These simpler contrivances, which also have a wider dis- 

 tribution, suggest earlier patterns. The use of movable partitions 

 was, however, advantageous when the benches were cleared to 

 accommodate crowds of spectators at a ceremony. 



Another probably ancient feature was the use of vertical wall 

 planks. The setting of these into the ground would seem to be 

 not only simpler but more widespread and presumably earlier than 

 fitting them into a grooved frame or sill. In fact, "insertion of the 

 planks in bottom and top logs has a far more limited distribution 

 than the use of vertical timber as a whole, being, as it seems, confined 



