Reviews. 391
from a rigorous examination of the Central Australian evidence. But there is much room for difference of opinion as to the Banks Islands evidence ; for Dr. Rivers appears to state explicitly that what is believed to enter the woman is not a real animal or plant but some incorporeal phantasm of one ; in fact, we do not know that the belief is not the same as that of the Arunta, and that what is incarnate is not a human spirit.
Although Dr. Frazer does not mention it, there is one point on which the totemism of the Central tribes of Australia differs markedly from that of the other totemic peoples, and it suggests that totemism elsewhere must have originated differently if the totemism of Central Australia has not been modified. Pre- cisely how many totem kins there are among the Arunta is probably unknown ; Strehlow gives a list of fifty-nine ; Spencer and Gillen enumerate sixty-six. Now, in the south-eastern tribes, so far from finding a large number of totems, we find a very small one; eight or ten is the ordinary number, if we exclude multiplex totems. It is prima facie highly improbable that the number of objects should be so small, if Dr. Frazer's theory of a conceptional origin is the correct one ; if conceptional totemism ever existed there, it must have been much modified. But this is not the only difference ; plant totems are common in the centre and north, but almost unknown in the south-east. Why is this ? If the eating of food or sight of an object was held to produce pregnancy, and from this belief arose totemism, plants,, which women rather than men would collect for food, should surely provide as many totems as the animal kingdom !
Admitting, however. Dr. Frazer's premises, is he right in tracing hereditary totemism to this source ? The crux of the situation is evidently to explain how the hereditary principle was introduced ; and here Dr. Frazer has little guidance to give us.
The American view of the origin of totemism is that it was developed from the personal totem. Dr. Frazer objects to this that (i) personal totems are rare in Australia, (but on this point see Mrs. Langloh Parker), and (2) many totem kins reckon descent in the female line, and that the personal totems of women are unimportant. Admitting the latter fact, the answer is obvious : inheritance from the mother's brother will produce
