We have thus reached totemism, and we trace its varying forms in the light of institutions which grew up in the evolution—under changing conditions—of the law of exogamy. The causes are demonstrably veræ causæ, conspicuously present in savage human nature, and the hypothesis appears to colligate all the known facts.
The eccentric and abnormal types of social organisation, as Mr. Howitt justly observes, are found in tribes which have adopted the reckoning of descent, or inheritance of names, in the male line. Phratry names lose their meanings or vanish, even phratries themselves decay, or are found with names that can hardly be original, names of cosmogonic anthropomorphic beings, as in New Britain. Totems, under male descent, become names of groups of locality, and local limits and local names (names of places, not totems) come to be the exogamous bounds, as among the isolated Kurnai.
In America, magical societies of animal names, and containing members of many totems, have been evolved. But we must not fall into the error of regarding such societies as "phratries." Nor must we confuse matters by regarding every animal now attached to any kind of association or individual as a totem. Each sex, in many Australian tribes, has an associated animal. Each dead man, in some communities, is classed under some name of an object of nature. Each individual may have a patron animal familiar revealed to him, in a dream, or by an accident, after a fast, or may have it selected for him by soothsayers. The totem kins may classify all things, in sets, each set of things under one totem. But the animal names which are not hereditary or exogamous are not judiciously to be spoken of as "Sex Totems," "Mortuary Totems," "Individual Totems," or "Sub-totems." They are a result of applying totemic ideas to the sexes, to dead men, or to living individuals, or to the universe. Perhaps totemic methods and style were even utilised and adapted when the institution of matrimonial classes was later devised.
[1] The Arunta exception has been explained. Cf. Chapter IV.
[2] Cf. Social Origins, pp. 55—57, in which the author fails to discover any mode by which the distribution could occur accidentally or automatically.
[3] J. A. I., August 1888, p. 40.
[4] Ibid., August 1888, p. 53.
[5] N. W. Thomas, Man, January 1904, No. 2.