boas] TSIMSHIAN SOCIETY 523 



The Lax-se'°la division of the Ganha'da would correspond to the 

 KiksA'di of Sanya. The two divisions have the sculpin crest. 



The subdivision Ganha'da may perhaps correspond to the Tongass 

 GanAXA'di. 



It is also interesting -to note that the Haida Raven crests grizzly 

 bear, mountain goat, raven in sea, and moon are said to be of Tsim- 

 shian origin. Thns a native origin is assigned to all the important 

 creslsof the Haida Raven side, so far as they are not based <>n sea ani- 

 mals. The Tsimshian GispawadwE'dafall clearly into twogroups, — 

 an inland group with land-animal crests, and a coast group with sea- 

 animal crests. It would seem likely, therefore, that the Haida Ravens 

 corresponded, first of all. to the seacoasl group, and that the identifi- 

 cation with the inland groups developed later. The fact that a 

 definite group of the Haida Ravens lack the grizzly-bear crest is also 

 in favor of this theory. If the Tsimshian are right in their opinion 

 that their ancestors, more particularly the GispawadwE'da. lived 

 inland in the ancient town of T !Ein-lax-a'm, then the sea-animal 

 crests must necessarily lie more recent than the land-animal crests, 

 and the G*it-na-gun-a'ks group would represent a comparatively 

 speaking late development among the Tsimshian. 



Swanton has suggested that the Haida Ravens may be the ancierit 

 Haida, and that the exogamic groups may have developed owing to 

 the immigration of a foreign element, who formed the Eagle group. 1 

 In regard to the Tlingit he suggests a similar theory, and thinks that 

 the Ravens may be a foreign element among the Tlingit.- It might 

 seem that the Tsimshian tales give countenance to the view that one 

 of the exogamic groups represents the central body around which 

 the newcomers clustered as new units. 



In our discussion of the division of the exogamic groups (p. 486) 

 it has been pointed out, that, according to tradition, part of the 

 G'ispawadwE'da are apparently the only division of the tribe that 

 constituted the ancient Tsimshian. 



The question arises whether we shall adopt a euhemeristic explana- 

 tion, and consider the legendary history of the clans as reflecting 

 historical facts. I believe the data do represent actual occurrences. 

 I am veiy doubtful, however, whether the facts would justify us in 

 asstiming that the coming-together of these different groups led to the 

 formation of the exogamic groups. The stories themselves do not 



i Swanton 2, p. 104. 



a Swanton 4. p. 407; see also John R. Swanton. "The Development of the Clan System and of secret 

 Societies among the Northwestern Tribes" (American Anthropologist, N. s.. vi, 1904, pp. 477 it seq.). 

 Incidentally I would say that I do not consider Swanton's hypothesis of the splitting-off of the 

 Bellacoola from an interior Salish tribe as 1 ikely. The Bellacoola dialect is closely associated with the coast 

 dialects of Salish stock. To mention only two points, it shares with them the development of grammatical 

 gender, which is absent in the dialects of the interior, and it has in common with them certain terms relating 

 to the sea. Linguistically it is quite necessary to assume a close relation between Bellacoola and the 

 other coast dialects. 



