But there is nothing in this patriarchal organisation of the family and of the tribe which compels us to assume that it was evolved out of some earlier non-patriarchal form of family. The warrant for such an assumption, if to be found, must be sought elsewhere. Let us seek. Analogy will not help us. The patriarchal system may, elsewhere in the world, have been evolved out of the matriarchate; but, as the late Mr. M‘Lennan warned us, we may not assume that marriage has everywhere had the same history. The widest survey of the various forms of human marriage (Westermarck's) that has yet been made warrants no presumption in favour of the priority of the matriarchate. If the matriarchate was a pro-ethnic Aryan institution, it is on Aryan ground that traces of it must be discovered. Such traces are said to be discernible. There are traces amongst some Aryan peoples of the levirate. The levirate is said to indicate polyandry, and polyandry to presuppose the matriarchate. This is a perfectly legitimate line of argument, but before resorting to polyandry for an explanation of the Aryan levirate, it is worth while to inquire whether there is anything in known Aryan customs capable of supplying an explanation. According to Aryan custom, the estate of a man who leaves no son passes to the next of kin, _i.e._, his brother, or it may be a more distant relative. If the deceased leaves no son, but a daughter, then according to Athenian law, according to the Gortyn Code, and probably also according to Aryan custom, the next of kin (whether brother or not) must not only take the estate, but also marry the heiress, if any (whether wife or daughter of the deceased). According to the Gortyn Code, if the next of kin is married, he must put away his wife; if the heiress is already married, she must leave her husband. Now, if the obligation to raise up seed to the deceased extended only to his brothers, the Tibetan form of polyandry would afford an explanation which, whether correct or not, would, at any rate, account for all the facts. But inasmuch as the obligation is binding on all the near kin, and extends to the daughter as well as the wife of the deceased, it cannot be explained by the hypothesis of the Tibetan form of polyandry or any other form short of incest in every degree possible, not only amongst the members of the same joint undivided family, but also with the women who have married out of that family into some other. In truth, so far from _mutterrecht_ being the source of the Aryan custom, that custom bears on its face the marks of the rudest and most savage application of the agnatic theory. The provisions of the Gortyn Code which require that the next of kin shall marry the heiress, even if the marriage necessitate divorce on both sides, show that the mother was held absolutely incapable of transmitting rights—only a kinsman could do that. A devotion to the principle of agnation so strong as to over-ride the