one widow. She is wife to both, and any child she may have belongs to both. There are cases in which a husband connives at a connection between his wife and another man. This is not counted adultery, for it is an open transaction; and it is not polyandry, for the parties are not counted husband and wife. It is not considered respectable."
The existence of polyandry is not denied, but I venture to hazard the assertion that it is not the system of marriage in any tribe at the present day, and it seems to me impossible that it could have been the rule of marriage anywhere at any time in the past. The mere arithmetical difficulty in the way appears to me to be insurmountable.
Though such statistics as I have been able to get at among the lately heathen tribes in Feejee directly contradict the hypothesis, still I think we may suppose that the number of males generally exceeds that of females among the lower savages. But it does not seem to have occurred to Mr. McLennan to consider how great his theory of polyandry as a system of marriage requires that disparity to be. Under such a system it is evident that, whatever the average number of husbands to a wife maybe, at least so many times more numerous must the men be than the women. If y be the number of women, and x their average allowance of husbands; then, since we can not suppose that under such a system any marriageable girl would be allowed to roam in "maiden meditation fancy free," the number of men in the tribe must be x y, even supposing all the men to be absorbed in the "combinations of husbands."
Nor will marriage by capture help us here; because, for every woman stolen there must have been x husbands left lamenting, unless we suppose that a non-polyandrous tribe was kept in the neighborhood of each polyandrous tribe for its convenience, and that they never retaliated upon their aggressive neighbors.
To sum up: It has been shown that Mr. McLennan's postulate of female infanticide as the rule among the lower savages can not easily be granted; that his exogamous tribes are not exogamous in the sense which his theory requires; and that both marriage by capture and polyandry, as systems of marriage, involve something which has all the appearance of an absurdity. Without claiming too much, then, I think it may be said that the basis of Mr. McLennan's theory has been shown to be insecure. And this being so, it is all the greater pity that he allowed himself to treat with such contemptuous scorn the hypothesis advanced by Mr. Morgan in his work on "Systems of Consanguinity,"[1] which hypothesis is opposed to his own.
"This wild dream—not to say nightmare—of early institutions. . . . It seemed worth while to take the trouble to show its utterly un-
- ↑ "Systems of Consanguinity and Affinity of the Human Family," "Smithsonian Contributions to Knowledge," vol. xvii.
