India, Greece, or Rome, that we possess much historical material. What we know of the peoples of Australia, Melanesia, New Guinea, or Borneo represents only the condition of their inhabitants as revealed to modern travellers, and their history is a blank. In the case of the former group we may expect to find "survivals" in the present which can be analysed and explained by our knowledge of their historical literature. Of the latter we know nothing save what we can pick up at the present day. It may be true that, as in the case of India, the historical record may have been manipulated to justify the pretensions of a priestly body, or to support some scheme of dogmatic theology. Still, with all its imperfections, the historical record does help to explain much which would otherwise be obscure, while in the case of savage races we can know nothing of the stages through which any single belief or usage may have passed.
If, for example, we examine the social system of the Arunta, we find no record which assists us in interpreting it, save some vague tribal legends projected into the Alcheringa, the age of the mythical tribal ancestors. Hence we are at a loss to explain the origin of a regulation, such as that which divides the tribes into two moieties; and we are left to infer, from our ideas of the probabilities of the case, aided by a comparison of facts drawn from other groups in a similar stage of culture, that it results from the coalition of two distinct tribes, or that it was established by some primitive statesman, or council of the tribal greybeards, who were forced to take action in view of the obvious physical dangers resulting from the intermarriage of the members of a small community. I confess that I find some difficulty in accepting the theory that a reorganisation such as this was the work of some savage Lycurgus and his assessors. Such reforms are, I am inclined to believe, seldom introduced per saltum; it is more probable that they represent the final stage of a long
