REVIEWS 381 



the author at the school. Its object is "to demonstrate the dependence of the 

 terminology of relationship upon social conditions." There has hitherto been 

 so much misconception and confusion, arising from the premature formulation 

 by anthropologists of an earlier day of theories, for which the facts had as yet been 

 imperfectly collected and classified, that it was time for a sane and careful review 

 of the results of the inquiries of the last twenty years. During those years 

 British Anthropologists, led in the first instance by Dr. Haddon, and American 

 anthropologists, under the direction of the Bureau of Ethnology at Washington, 

 have by minute and scientific investigations very different from the haphazard 

 observations of previous generations accumulated a store of information on the 

 social conditions and organisation of peoples in the lower culture. Dr. Rivers's 

 wide experience in the field, his reputation for cautious and accurate reasoning, 

 and his authority on questions relating to social organisation, have rendered him 

 peculiarly fit for the task of exhibiting in a small compass the conclusions to 

 which our more recent and trustworthy information points. 



He has achieved the task ; and these lectures must for a long time be a 

 text-book of method and a model of exposition on the subject. It has always 

 seemed incredible to most students that the terms of kinship used by savages 

 and others could be mere terms of address. Even as polite formalities they 

 must have had some basis of social fact. Dr. Rivers, by a series of well-chosen 

 examples, drawn partly from his own personal inquiries in the field, has victori- 

 ously demonstrated this, and has shown that however strange, or even absurd, the 

 relationships that emerge may appear, they notwithstanding actually exist, or 

 at least existed at a period not very remote. 



His lucidity is admirable. His criticism of the terminology hitherto in scientific 

 use is excellent. " Classificatory " and " descriptive " are quite meaningless as 

 applied to the systems of kinship employed in the lower and higher cultures 

 respectively. It would have been well, however, not to take for granted that 

 all his readers would know what "cross-cousins " are or what are the distinguishing 

 characters of the Hawaiian, or as Morgan, the distinguished American anthro- 

 pologist called it, the Malayan system of kinship. 



Attention may be called to one interesting suggestion. Dr. Rivers refers (p. 90) 

 to Mr. Blunt's account, in the Report of the Indian Census of 191 1, of the terms 

 in use in the United Provinces. They present several remarkable features, 

 explained by Mr. Blunt as arising from the use by younger members of the 

 family of the same terms as are used by older members to describe their relationship 

 with the persons spoken of. The existence of the joint or extended family, where 

 it obtains, would cause close contact, which would render such an explanation 

 probable. But Dr. Rivers, while admitting this, points to the possibility of an 

 alternative explanation in the case of our term, baku, used alike for the son's wife, 

 for the wife, and for the mother. He suggests that it points to a form of polyandry 

 in which a man and his son have a wife in common. No such form is found now, 

 though of course it may once have existed. If we can rightly explain one of these 

 terms by itself without reference to the others (which perhaps is doubtful), it is 

 interesting to note that Caesar attributes exactly this form of polyandry to the 

 ancient Britons, and it would be worth while to inquire whether any terms 

 can be found in the Celtic or in the Teutonic languages explicable by such a 

 custom. Dr. Rivers finds traces of the extended family in these languages. 

 If he could trace also this special form of polyandry, his suggestion as to the 

 origin of the term ba/iu would begin to look somewhat more than possible. 



E. Sidney Hartland. 



