The Sociological Significance of Myth. 325
Panunga, Bukhara, Purula, and Kumara, which were con- ferred by the visitors, and the narrative does not say that these people instituted the classes, much less the dual organisation which underlies them.
The narratives recorded by Strehlow ^"^ give a somewhat different account. In one place ^^ it is said that, when the rella mancrinja or intariiija (the Inapertiva of Spencer and Gillen) were first visited by the altjirangamitjina (the Alcheringa ancestors of Spencer and Gillen), they were already divided into eight classes. Later 2*' it is stated that, when Mangarkunjerkunja, whose totem was a fly-catching lizard, came from the north, he instituted the rules of marriage between the classes, which, it is again stated, had been already distinguished from the beginning.
Taking the accounts of Strehlow and of Spencer and Gillen together, it seems clear that the narratives do not give an account of the .formation of the dual system or of the matrimonial classes, but of some change in the functions of this social grouping in the regulation of marriage.
The position to which we have now been led is that, when Australian myth deals with the origin of social institutions, it is usually the totemic system which forms the special topic of the narrative, and not the dual system and matrimonial classes which seem to form the essential basis of the social structure. This suggests that Australian totemism has become the subject of myth, not through its social importance but for some other reason, and for this other reason we have not far to seek. I suggest that it is the magico-religious importance of totemism, and not its social functions, which have made it so exceptionally and prominently the subject of Australian myth. If so, it will follow that the preoccupation of Australian narrative with social forms is largely apparent rather than real, an appear-
- Mythen, Sagen unci Alarchen des Aranda-Stdmmes in Zentral-Australien^
Frankfurt-am- Main, 1907.
" Op. Cit., p. 3. 26 (9^_ ^j/^^ p. 6.
