Generally men and women of the same Totem cannot marry. Tribes, such as the Arunta, that do not observe this rule are so few that it is reasonable to consider them aberrant; so that in each case the non-coincidence of Totemism and Exogamy is a distinct problem. Nevertheless, there is reason to think that these institutions have not the same origin. For several primitive tribes, some of them inferior in culture to the average Australian, have customs of Exogamy, or (at least) forbidding marriage between individuals of some certain description, who have not, and are not known ever to have recognised Totemism. Many tribes, again, who by their ethnological position may be supposed once to have entertained Totemism, have abandoned it, but still maintain Exogamy by means of Marriage-classes (phratries) or otherwise. Further, many nations of advanced culture cannot be shown ever to have been totemists, but have always (as far back as can be traced) enforced Exogamy so far as to prohibit marriage within certain degrees of kindred. And that the connexion between Totemism and Exogamy is not original but acquired, is indicated by the consideration that the most obvious tendency of Totemism would be to favour Endogamy (the practice of not marrying outside certain limits), on the obvious principle that as an animal kangaroo mates with an animal kangaroo, so a human kangaroo should marry a human kangaroo. This is so obvious that the opposite rule that a human kangaroo must not marry another one, but (zoological paradox!) only an emu or a witchety-grub, seems perverse, and only to be explained by the influence of some superstition or incidental custom. According to Arunta tradition, their ancestors in the Alcheringa were endogamous. Whether Exogamy originated in the days of the hunting-pack—which is so far probable that the Boschmans observed it—or during the reconstitution of society after the breaking up of the packs, or even later, or in some regions at one period, in others at another, is, I fear, beyond our power of verifiable guessing. The best hypothesis as to the grounds of the custom is Westermarck’s, namely, that an instinct of mutual avoidance grew up between near kin. This requires that at the time of its origination promiscuity of sexual relations should not have prevailed in the human stock, since that would have destroyed the conditions necessary to the rise of such an instinct. If promiscuity was ever widely practised, it must have been after the breaking up of the hunting-packs. That at some stage of human life, and apparently with some Australian tribes not a remote one, promiscuity was established, has been argued (on the ground of some Australian customs) with much plausibility; but in his _History of Human Marriage_,[529] Westermarck has examined this opinion very carefully, and concludes that it is not tenable. His reasoning seems to me good throughout; and, having nothing important to add to it, I refer the reader to his work.