320 Correspondence.
pared to deny that in certain cases — not to claim undue considera" tion for my somewhat ' scratch ' individual examples — abuse and the throwing of filth may, at the lower levels of ' cult,' cloke a passionate interest as towards 'powers' which in essence is wholly one with the feeling that at another moment may issue forth in prayer, sacrifice, or other form of reverential address ?
As to " Baiame and Co.," I would begin by stating that I follow Mr. Lang in believing them to be a genuine achievement of savagery and no mere rechauffe of the teachings of missionaries. On the other hand, has Mr. Lang given to the world any general explanation for me to follow of the fact that his " Superior Persons," from Baiame and Co. to Tui Laga, possess in common a certain ethical quality ? Failing to get a decisive ' lead ' from him, I have to fall back on the well-tried, if unromantic, working hypothesis that moral deities derive their character from associa- tion with the moral institutions of society. Such a generalisation does not exclude Plurality of Causes when it comes to connecting particular deities or specific attributes with special institutions. Thus I do not suppose initiation ceremonies to have generated Mr. Lang's group of deities as a whole. It is on the contrary most likely that many of their number " came otherwise." (I can- not pretend, however, to have tried to work out the separate family histories of the worthies that figure in Mr. I-ang's long Hst.) Moreover, it would clearly be putting the principle of the Uniformity of Nature to sad misuse to argue that the Bull Roarer must, if ever, then always, attain to apotheosis ; or that a moral institution, if ever, then always, must associate with itself, and reflect itself in, the attributes of whatever deities the instruments and accompaniments of its ritual are capable of suggesting to the human imagination. Thus it may be usual for savages to feel ' awe ' at a storm ; yet some tribes are said positively to enjoy a hurricane. Or again primitive moral institutions do certainly tend to rest on some sort of ' supernaturalistic ' support ; but there are, I fancy, religious initiations and civil initiations, religious marriages and civil marriages, at every stage of man's advance. Hence, if it comes to a question of "logic," I would venture to remind Mr. Lang of the shortcomings peculiar to the unsupported Method of Agreement. Anthropology is still at a very empirical stage ; and the search for viedia axiotnata — for limited connec- tions of cause and effect — must therefore be suffered to coexist
