474 Reviews.
the phratries and classes as given by Mathews with that given by Messrs. Spencer and Gillen,^ we shall find that the classes having interconnubial relations and the descent agree in every particular, — in short, that the arrangement is precisely the same, except that two of the classes {A2 and A\ in Mathews' list) have changed places with ^4 and B2 of the same list. In one word, while Mathews' list represents the change as in process, Messrs. Spencer and Gillen's Hst shows the change as completed, and the descent in both totem and phratry on the male side. To effect this the two classes referred to in each phratry have changed places.
The consideration of the evidence produced by Professor Durkheim thus affords a presumption of the truth of his hypothesis. But it cannot be regarded as absolutely proven. No tribe has yet, to my knowledge, been found com- bining eight matrimonial classes with female descent ; but there are certainly a few tribes having only four classes which yet reckon descent through the father only. Dr. Howitt, from whom we learn of their existence {Native Tribes ofS.E. Australia, p. 114), is able to give us too little information to found very definite ideas as to the details of their organisation. The classes, however, are found in intimate relations with totems. Some of the tribes, the Kaiabara and Muruburra, for instance, definitely allot certain totems to each of the sub-classes. In the Annan River tribe, near Cooktown, Queensland, information of which was forwarded by Dr. Roth, both the classes and sub-classes are named from animals : an arrangement pointing to their totemic origin. Further research will be necessary to ascertain the mode of development of the organisation of all these tribes, and to obtain more direct and satisfactory evidence of the correctness of M. Durkheim's hypothesis.
Meanwhile that hypothesis would appear to account for the formation of the eight matrimonial classes. It is supported by the facts which Professor Durkheim alleges ; and if these facts
1 Northern Tribes, p. lOO. The names are not difficult to identify. M. Durkheim has not discussed the changes I have pointed out here, but they seem to me to confirm his hypothesis.
