Two whole totem kins, say Wolf and Raven, or Eagle Hawk and Crow, were, in the new conditions, plus the old legal survival, cut off from marriage. If they died celibate, their disappearance needs no further explanation. But they do not disappear. If they changed their totems their descendants are lost under new totem names; but, if totems were now fully-blown entities, they could not change their totems. They could, however, desert their local tribe, which has no tribal "religion" (it sometimes, however, has an animal name), and join another set of local groups (as Urabunna and Arunta do constantly naturalise themselves among each other, to-day), or, they could simply change their phratries (late their local groups). Eagle Hawk totem kin, by going into Eagle Hawk phratry, could marry into Crow phratry; and Crow totem kin, by going into Crow phratry, could marry into Eagle Hawk phratry. This, I suggest, was what they did.
This would entail a shock to tender consciences, as each kin is now marrying into the very phratry which had been forbidden to it. But, if totems were now full blown, anything, however desperate, was better than to change your totem; and after all, Eagle Hawk and Crow were only returning each into the new phratry which represented their old local group by maternal descent. Thus in America we do find Wolf totem kin, among the Thlinkets, in Wolf phratry, and Raven in Raven phratry; with Eagle Hawk in Eagle Hawk, Crow in Crow phratries, Cockatoo and Bee in Cockatoo and Bee phratries, Black Duck in Black Duck phratry, in Australia.
The difficulty, that Crow and Eagle Hawk were now marrying precisely where they had been forbidden to marry when phratry law first was sketched out, has been brought to my notice. But the weakest must go to the wall, and, as soon as the totem became (as Mr. Howitt assures us that it has become) nearer, dearer, more intimately a man's own than the phratry animal, to the wall, under pressure of circumstances, went attachment to the phratry. Il faut se marier, and marriage could only be achieved, for totem kins of the phratry names; by a change of phratry.
But is the process of totem kins changing their local groups (now become phratries) a possible process? Under the new régime of fully developed totemism it was possible; more, it was certainly done, in the remote past, by individuals, as I proceed to demonstrate.
[1] Totemism, p. 62. Cf. McLennan, Studies, Series II. pp. 369-371.
[2] L'Année Sociologique, i. pp. 5-7.
[3] It is not plain what Mr. Frazer meant when he wrote (Totemism, p. 63). "Clearly split totems might readily arise from single families separating from the clan and expanding into new clans." Thus a male of "clan" Pelican has the personal name "Pouch of a Pelican." But, under female descent, he could not possibly leave the Pelican totem kin, and set up a clan named "Pelican's Pouch." His wife, of course, would be of another "clan," say Turtle, his children would be Turtles; they could not inherit their father's personal name, "Pouch of a Pelican," and set up a Pelican's Pouch clan. The thing is unthinkable. "A single family separating from the clan" of female descent, would inevitably possess at least (with monogamy) two totem names, those of the father and mother, among its members. The event might occur with male descent, if the names of individuals ever became hereditary exogamous totems, but not otherwise. And we have no evidence that the personal name of an individual ever became a hereditary totem name of an exogamous clan or kin.
[4] It was first put to me by Mr. N. W. Thomas, in Man, January 1904, No. 2.
[5] Mr. Howitt affirms that the relative lateness of these classes, as sub-divisions of the phratries, is "now positively ascertained." (J. A. I., p. 143, Note. 1885.)