Nair polyandry resolves itself into cohabitation between permitted groups. Mr. McLennan asserts that "the Nair husbands are usually not brothers—usually not relatives." But in what sense does he use "brothers" and "relatives" here? If by "brothers" he means only "children of the same parents," and by "relatives" only "persons who are related according to our own notions of relationship," then his statement is of little weight, for a group of "tribal" brothers may include many persons other than these.
The Thibetan instance quoted from Turner, where "five brothers were living very happily under the same connubial compact"[1] seems to be a clearer case. But there is no proof that this is an instance of true polyandry, and not of polyandry combined with polygynia, like the Nair custom, the custom of the Britons as described by Cæsar[2] and all the other instances given by Mr. McLennan where tribal brothers hold their wives in common. And, considering how easy it is to mistake instances of group-marriage for polyandry, such proof may be reasonably demanded from one who represents polyandry as an extensive system of marriage.
The law of the Levirate, which Mr. McLennan considers "it is impossible not to regard as. . . derived from the practice of polyandry,"[3] does not appear to me to have anything at all to do with polyandry. It was a regulation to prevent the elder branch of a stock from becoming extinct. Its underlying motive is found in the preferential claim to the birthright vested in the elder branch; and this preferential claim is found only in tribes who have descent through males, or at least who, having settled down to agriculture, are well on their way to that line. The lower savages know nothing of that motive. Mr. McLennan lays stress upon the fact that the widow was the Levir's "wife without any form of marriage." But there is no proof that this is a survival of polyandry; for, in the first place there is no need for us to look upon it as a survival of anything at all, and, in the second place, it would serve very well as a survival of group-marriage. In many tribes the brother's widow is the Levir's wife "without any form of marriage." He does not even wait until she becomes a widow. He is of the same group with her husband, and her group is "wife" to his group.
It is not denied that cases of polyandry occur. A few instances of it have come under my own observation. But in every case the men were of a clan which intermarried with that of the woman, the circumstances were exceptional, and the custom was not the general practice—not even the frequent practice—of the tribe. In full accordance with this is the following account of polyandry at Mota, written to me by the Rev. R. H. Codrington before mentioned:
"Polyandry exists, but rarely—never with young people, but mostly as a matter of convenience, as when two widowers live with
