1 82 The 07'igin of Exogamy and Totemism.
Mr. Frazer makes no reference to myself, or to M. van Gennep, on this matter, but (vol. i., pp. 284-5), argues against the theory of amalgamation, without noticing our replies to certain objections that had already been urged on us by others. " Why," he asks, " were these federal communities so regularly either two in number or multiples of two?" M. van Gennep had briefly said that our theory of con- vergence (amalgamation) "seule explique entre autres le fait du dualisme des elements de chaque groupe [p. xxxiv]." He added that the Australians generally "n'ont de noms de nombres que jusqu'adeux,"and,for an element of symbolism in this, refers to Mr. MacGee, " Primitive Numbers," in The NineteentJi Annual Rcpoj't of the Bureau of American Ethnology, vol. ii., pp. 821-51. These are not my own ideas, but those of M. van Gennep.
I would say that, if amalgamation began in the Urabunna " one totem to one totem " marriage, while such pairs finally federated into each phratry, Mr, Frazer's question is answered. I regard the later bisections of two classes into four, and of four into eight, as deliberate and intelligent imitations of the original model, — the sets of pairs, the " two class system." The natives, like Mr. Frazer, would think that it had been the result of bisection, not of amalgamation, and would imitate what they supposed to have been the wise method of their ancestors.
Mr. Frazer's argument ought to be given in his own words (vol. i., p. 285) : —
"While we may without much difficulty conceive that com- munities," (in this case totern-kins) " which in their independent state had been exogamous, should remain exogamous after they had united to form a confederacy ; it is far more difficult to understand why in uniting they should have adopted the com- plicated rules of descent which characterise the four-class and eight-class organisations of the Australian tribes."
Nobody has ever suggested (as far as I know) that " in
