Pj^esidential Address. 29
slipped in and got to work ; and it is plainly not in his heart any longer to wish it otherwise.
Going back for a moment to Gomme's ethnological work, we may note the same surreptitious ingress of a psychology that will not be denied. I take a couple of examples almost at hazard. Thus his theory of the origin of the village-community demands that the Aryan immigrants stand to the pre-Aryan aborigines in the relation of con- querors to conquered. Yet the former are assumed to have " adopted and adapted " certain beliefs of the indi- genous population. Why } Because for religious reasons the invaders are apt to borrow from the local folk so as to make themselves at home among the sacred powers of the land."^ Now such a principle is to a certain extent suscept- ible of proof, or disproof, by the collection and comparison of historical instances. But in essence it is a psychological cause that is invoked, and one which, if genuine, must have operated independently again and again. Once more, he puts forward a hypothesis which, though it is to be taken in close connexion with the other, rests on a psycho- logical principle of another order, nameh', one belonging to the psychology of sex. "It seems to me quite possible," he writes, "that the women of a conquered race, feared as they often were by their conquerors as the devotees of the local deities, miight use that fear under some conditions to establish a place of power which has left its mark on the history of marriage." -*' Now here we have just the sort of problem concerning the effect of culture-contact on marriage-organization that Dr. Rivers has constantly to face in his " History of Melanesian Society." It may or may not be necessary in such a context to speculate on what might happen in virtue of the tendency to regard women as the mysterious sex. But I fail to see how we
-'Compare /7V/(--Z();v, iv. (189,^), 13. In confirmation of such an explana- tion, see E. S. Haitland in follc-Lore, xxvii. (1916), 319. -Folk-Lore, ii. (1S91), 494.
