1 84 The Origin of Exogamy and Totemism.
was none till, long afterwards, a subclass arrangement was devised to stereotype and express publicly the already existing bye-law against the union of father and daughter, motherand son. The poor children," rigidly excluded from the communities of both parents," are still in the paternal or maternal " communities " of the totem-kin and the phratry and in the family fire-circle. They have lost nothing. That exclusion is perfectly intelligible on the hypothesis that it was devised to prevent the marriage of parents with children, but it is difficult to see how it can be explained on any other: who is dreaming of explaining it on any other.-' People entered into the phratriac exogamy by the amalgamation which I and M. van Gennep suggest, and then, for conscientious reasons, "kept compounding it as they went on," as Byron says about people who " began with simple adultery."
Manifestly Mr. Frazer has not understood the theory of amalgamation. Beginning in the lower degree with the intermarrying pairs of the one totem to one totem pattern, the duality was necessarily preserved in the phratry con- federation, and was deliberately imitated in the formation of first four, and then eight subclasses.
Of course I cannot answer the question " Why did one totem pair off only with one other totem at first, why not with three, or five, or seven ? " I can only say that the Urabunna, (or certain of their tribes), the Karamundi, and the Itchumundi obviously did this very thing, and do it still. Apparently we find an intermediate stage of inter- marriage of only three or four totem-kins in each phratry. Rat (in Kararu phratry) marries Cormorant and Bull Frog ; Bull Frog in Matteri phratry marries Rat, (and probably Cormorant), while Cormorant (in Matteri phratry) marries Rat and Red Ochre, in Kararu phratry. This is the rule in the southern Urabunna tribe called Yendakarangu. It is said to be an Urabunna tribe, but has the Dieri, not the Urabunna phratry names. The facts in this strange case