84 ROYAL SOCIETY OF CANADA 



beginning his first element is based on an assumption which a close study 

 of the subject makes it 'difficult to justify. It is assumed as an accepted 

 and incontrovertible fact that the social organization of savage peoples 

 into clan groups in the matriarchal stage has its foundation in totemism. 

 But no proof has been, or can be, given for this statement and such evi- 

 dence as we can gather on the point leads to the opposite conclusion. All 

 we certainly know of the earlier stages of human society is that hordes 

 cr bands lived together under an organization which we call matriarchy 

 or "mother-right;" that is kinship was traced through the mother only, 

 the most obvious and the most certain form of relationship. Now, it is 

 clear that the recognition of uterine ties must bind the mother to her 

 ■offspring and them to her in closer bond than any other. Again, uterine 

 brothers and sisters are a naturally defensive and co-operative group and 

 spontaneously aid each other to avenge insults and redress wrongs. Here 

 then, we probably have the pristine unit of social organization. But 

 the mother of this " family " is also uterine sister to other sisters and 

 brothers; therefore her "family" is connected by ties of blood to other 

 "families." N"ow, the aggregation of these blood-related "families" con- 

 stitutes a wider group, and this is the clan of matriarchy. Clans are 

 confessedly blood-related groups, and this bond or union is everywhere 

 seen to be based on this kinship of blood. The formation of clans, th^n, 

 has nothing to do with totems, and it is not the common totem, which is 

 inherited from the founder of the clan, that makes the members of the 

 clan kinsmen. Clans, then, are purely social groups held together by the 

 common] tie of blood; and may, and most certainly do, exist as such, 

 apart from any totem concept. The totem is obviously a later feature, 

 and is in no sense an essential part of the clan structure. So much is 

 this seen to be the case that Dr. F. Boas,^ a most cautious and experi- 

 enced investigator, has remarked that the earlier social grouping of the 

 Haida and Tlingit appears to have been on lines similar to the com- 

 munal organization of the more southern tribes, as the clans so fre- 

 quently bear territorial names instead of totem names. Wemiaminow 

 and Krause also noted that certain Tlingit clans were called after the 

 localities where their communal houses stood. Indeed, it is a common 

 practice with the Haida and Tlingit to call their clans after the names 

 of their houses or the places where they are erected. And yet these 

 tribes have a strictly matriarchal organization witJi group totems. It is 

 not safe, then, to affirm that totemism implies the division of a people 

 into totem-kms; the kinship is not totemic but always consanguineous. 

 Totemism per se has no>thing to do with clan structure. 



Another feature of element No. I is the rule of exogamy. "Totem- 

 ism," says Dr. Haddon, " involves the rule of exogamy, forbidding mar- 

 ^ See Fifth Report on N.W. Tribes of Canada, B.A.A.S., 18S9. 



