The sons would have to bring in mates from without--the result is Exogamy. But Mr. Atkinson does not observe the numerous tabus existing among savages, on ordinary (not sexual) intercourse between men and women; as if each individual, of each sex, was or might be dangerous to each individual of the other sex; _that_ is no idea of our speechless brute ancestors, of Mr. Atkinson’s hypothesis. These tabus do not amount to absolute avoidance, but they do amount to very marked restrictions; for example, on eating together, or sleeping under the same roof, even where husbands and wives are concerned. For the facts, to save repetition, it is enough to refer to Mr. Crawley’s book, _The Mystic Rose_. Now if these less rigid tabus between the sexes (which Mr. Atkinson noted in his observations on the life of New Caledonian natives) arose in the general savage superstitious dread of everything not a man’s or woman’s own self, they might become more rigid as propinquity increased. The most dangerous female would be the female who had most chance of being dangerous, by virtue of propinquity, namely the sister. She would therefore be the most strictly barred. The closest of all relations, that of lover and lover, and man and wife, would be most severely guarded, as most dangerous, by tabus. All this would happen (granting the verifiable condition of savage superstitious dread) even if Mr. Atkinson’s theory of our speechless beast ancestors’ way of life were wrong. We should probably find the effects (Avoidance and Exogamy) even if the primeval causes postulated by him never actually existed. Moreover any avoidance between mother and son that may exist (as in the case of the mothers of chiefs, in New Caledonia, and their sons) is perhaps no more than part of the general rule of restricted familiarity between the sexes, whether that rule arises from a superstition, or from the circumstance that men and women sometimes ‘disturb each other damnably,’ as Lord Byron remarked to his wife. It might be argued that the exogamous prohibition is only one aspect of the general totem tabu; and that, in the case of brothers and sisters, incest against the totem tabu needed to be guarded against (as most likely to occur) by precautions of avoidance peculiarly stringent. These precautions, then, would not necessarily come down from the time of our hypothetical speechless beast ancestors. They _might_ come down from that time, but the descent, it may be objected, is not necessary. The rules might have arisen among men as human as we are--Totemists. On the other hand, it might be argued for Mr. Atkinson that his hypothetical groups must be infinitely older than Totemism. When totem names were imposed on the earlier groups, the totem name and mark would only be a method of distinguishing group from group, probably becoming the germ of later superstitions by which everything connected with the totem was tabued, in each case, to the groups bearing its name. Either