The traditions of the Arunta represent the exogamous ‘phratries’ as later than the totemic (but not yet exogamous) division. Dr. Durkheim thinks this improbable or impossible. It is true that the ‘phratries’ or ‘classes’ are now much more important, among the Arunta, than the totems, on which Dr. Durkheim insists. They need not, therefore, be earlier. VIEWS OF DR. DURKHEIM The theory of Dr. Durkheim is not, perhaps, expressed with his usual lucidity; at least I have found some difficulty in understanding it. The following summary, however, seems to be correct. ‘The phratry,’ he says, ‘began by being a clan’ (in my terminology an exogamous local totem group). ‘There is no reason why this general idea should not apply to the Arunta. Consequently, since there are actually two exogamous phratries, we have reason to admit that this society was originally formed by two primary clans, or, if any one prefers the phrase, by two elementary totem groups, both exogamous (_également exogames_), for under this form the two phratries must have begun to exist. Now in that case there was at least a moment when marriage was forbidden between members of the same totem,’ though now among the Arunta this rule no longer obtains.[1] So far Dr. Durkheim and I hold identical views; we differ on a point of detail. What are, and whence came, the totems within the phratries? Dr. Durkheim conceives the case thus: Originally there was a ‘clan’ (local totem group) which was exogamous, and married out into one other equally exogamous clan. The members of each such exogamous totem group (‘clan’) then multiplied and ‘swarmed off,’ in colonies, and all such colonies took a new totem, while retaining ‘the sentiment of their primary solidarity’ with the original totem group. These are the ‘secondary’ totem kins. But why should they take new totem names and new totems?[2] I know not, but the original group from which they swarmed off now became their ‘phratry.’ This phratry, in many cases, still has a totem name, ‘which is the proof that it is, or has been a clan,’ that is an exogamous totem group.[3] Therefore exogamous totem groups were ‘primary,’ the existing totem kins are ‘secondary,’ they have split off from the original groups. As far as I am able to follow Dr. Durkheim’s reasoning, he and I differ on this one point. We both regard the two ‘phratries’ as having been originally local exogamous totem groups, which united in connubium. But in each ‘phratria’ there exist several totem kinships. Dr. Durkheim regards these as ‘secondary’ branches which split off from the two original local totem groups, and which, in each case, took new totem names, while retaining membership in their original totem groups, now ‘phratries.’ They are totemic colonies of a totemic metropolis. I, on the other hand, as has been explained, conceive that each of the two local totem groups which became phratries (say Emu and Kangaroo) already, by the action of