528 TSIMSHIAN MYTHOLOGY [hth. ann. 31 



When we disregard the isolated crests of the Haida, there remain 

 a very few that are common to most families of each side. These 

 may be considered the oldest crests. 1 



Among the Tsimshian the most highly specialized crests — such as 

 the scalp with fins, abalone bow, etc. — may be considered as of recent 

 origin. Their name3 indicate that they are ornaments, and show that 

 they can be nothing but crests. 



In the tales accounting for the acquisition of crests the modern 

 rules of hereditary transmission are not regarded. Leonhard Adam 

 has already called attention to this fact. 2 I am inclined to see the 

 explanation of this condition in the circumstance that crests are new, 

 and may therefore be used as presents to be given by the chief of one 

 exogamic group to another. In the same way as in historical times 

 Chief Dzeba'sa of G'it-qxa'la presented his Haida friends with new 

 crests, thus showing his greatness, so the supernatural beings gave 

 their crests to their sons or to their daughters' sons. I believe we 

 should see in this simply an expression of the interest of the father in 

 the welfare of his son or grandson, not an indication of the develop- 

 ment of paternal succession. 



Combining the whole evidence here presented, I think we must say 

 that no safe proof of the origin of exogamy by the junction of two 

 tribes can be given; that the probabilities are rather in favor of a 

 reduction of the number of exogamic units. Historical events have 

 led to a twofold subdivision: on the one hand, the exogamic groups 

 developed by the accession of bodies of outsiders which continued 

 to retain their names; on the other hand, the exogamic groups, in- 

 cluding these subdivisions, tended to split up into several village 

 communities when the number of inhabitants of the old village 

 became too large, so that the same group with its subdivision's was 

 found in several villages. The sets of names for the two kinds of 

 subdivisions are different among the Tsimshian, so that the two prin- 

 ciples of division can readily be recognized. Among the Tlingit and 

 Haida, on the other hand, both groups seem to have been designated 

 by terms of the same kind, with the result that the sets of names de- 

 veloping by accretion can not readily be recognized. The character 

 of the groups, however, is revealed in part at least by then crests and 

 by the crest myths. 



Dr. Swanton, in discussing his theory of the native origin of 

 the Haven group of the Haida as opposed to the foreign origin of the 

 Eagle group, calls attention to the fact that the supernatural beings 

 of the Haida belonged almost exclusively to the Raven side. I 

 think this may be explained in a different manner. Most of them 



' L. Adam (p. 188) misunderstands me if he thinks that my remark that these reflect the ancient organiza- 

 tion means that they represent each one exogamic unit. What I mean is that each side had these as its 

 oldest crests,— the G-it'i'ns, eagle and beaver; the Raven, killer whale and grizzly bear. 



