The most striking fact in the totemism of the Australian, which is most primitive in type, is that it has nothing to do with regulating mar riage, nor has the child necessarily either its mother's or its father's totem. Therefore, the totem is not a mark of blood-kinship. Accord ing to Australian belief, all the spirits live in certain restricted localities, and a man's totem is determined by the locality in which the spirit enters his mother's body, conception being thus explained. Further, far from exogamy being the result of totemism, exogamy is by no means a totemic law. There is in some tribes no restric tion on marriage within the tribal group, and in fact endogamy, or marriage within the tribal group, appears to have been the first rule.
Not less important are the facts in regard to the relation between the totemist and his food. Here, as in many other cases, it is clear that the animal or plant first venerated by a savage is the food that preserves his life. The redskin does not eat maize because he regards it as his sacred ancestor, but it is sacred and his parent because he is dependent upon it for sustenance. Among the Australians it is the totem animal which was eaten by predilection in the original form of the institution. Just as utility deified the Hindu's cow and the Indian's maize, so the opos sum was sanctified by the Australian who lived upon it, and as such it became his totem. For this reason the Australians have certain totem rites, the object of which was simply to increase the food supply. But, on the other band, as the in one clan refusing to eat one group of animals; while the American Indian, who takes as his private totem the animal seen in a dream the night before he is initiated into the tribe, may often have converted his own totem into that of his tribe. Honorific totems are merely assumed for politeness, so to speak. They are at least always of secondary origin. Color totems and odds and ends as totems are little more than fetishes and may be explained accordingly. Sex totems are not true totems at all, but deserve a specific name. Certainly in such cases there is neither blood-kinship nor any trait of real totem ism. Nor is it a. proof of totemism when a race holds the belief that its ancestors were divine and includes among its many ancestral gods divine animals, any more than it is totemic to derive a race from a tree. The Algonquin In
dians, for example, like the ancient Scandina vians, hold that all men are descended from an ash-tree; yet this tree is not their totem. It must be remembered also that when an institu tion is once established it moves forward of it self. Hence there may be clans which take to themselves totems unsuitable as food merely be cause other clans in their environment have totems, which are no longer interpreted as pro viders of food, but only as tutelary objects. In some cases totems may be assumed as coats of arms are assumed, simply to be in the fashion. Thus a tribe in Bengal a few years ago assumed as a tutelary divinity the dog, and the dog is to-day practically their totem. The reason why they chose any such divinity was simply that all their neighbors had totems and they wished to be fashionable, while the reason they selected the dog was that it was useless when dead' (i.e. it could not be eaten, and so might as well serve as a totem). Then, too, there are cases where what seems to be a genuine totem shows itself on in spection not to be of the same class with cases of genuine totemism. Thus an African selects be fore death the animal he will become after death, and this (species) then becomes sacred to his descendants.
Frazer divides totemism into what he calls the Egyptian kind and the Aino kind. Among the ancient Egyptians it was the cus tom, as it is to-day the custom among the Todas of India, to slay certain animals only as a sacrament; while among the Caucasian tribes and the Ainos of Japan it is customary to kill as a sacrament certain animals which are also killed regularly. But these two kinds are only stages of totemism. The sacramental slaying of an animal regularly killed is an expiation, based upon the same principle as that which induces a savage to apologize to a bear before killing it, but all sacrifice of totems is non-primitive.
The theory has recently been advanced that to temism was at first a kind of writing, indicating a clan; that those men called e.g. Beavers came to believe they were 'beaver sons;' and that wor ship of the beaver totem arose from this belief. In the most primitive totemism, however, the only support given to this version is that the totem is tattooed on the body of a clansman, which, obviously, might be due to a prior re spect for the totem.