Against the hypothesis of Hill Tout, Miss Fletcher, Boas and Frazer there is such an array of decisive facts that one is surprised that it has been so readily and so generally accepted. In the first place, we know that a man frequently has the greatest interest not only in respecting, but also in making his companions respect the species serving him as personal totem; his own life is connected with it. Then if collective totemism were only a generalized form of individual totemism, it too should repose upon this same principle. Not only should the men of a clan abstain from killing and eating their totem-animal themselves, but they should also do all in their power to force this same abstention upon others. But as a matter of fact, far from imposing such a renunciation upon the whole tribe, each clan, by rites which we shall describe below, takes care that the plant or animal whose name it bears shall increase and prosper, so as to assure an abundant supply of food for the other clans. So we must at least admit that in becoming collective, individual totemism was transformed profoundly, and we must therefore account for this transformation. In the second place, how is it possible to explain, from this point of view, the fact that except where totemism is in full decay, two clans of a single tribe always have different totems? It seems that nothing prevents two or several members of a single tribe, even when there is no kinship between them, from choosing their personal totem in the same animal species and passing it on to their descendants. Does it not happen to-day that two distinct families have the same name? The carefully regulated way in which the totems and sub-totems are divided up, first between the two phratries and then among the various clans of the phratry, obviously presupposes a social agreement and a collective organization. This is as much as to say that totemism is something more than an individual practice spontaneously generalized.