372 Totemism in the Evolution of Religion
that I do not hold that the mere anointing of an altar-post^ proves the deity thus worshipped to have been originally a totem, or that the sacramental meal and the social institu- tions of totemism are so " mutually dependent " " that the one cannot exist without the other. On the contrary, I have argued that the sacrificial feast is used by all sorts of societies as a means of binding themselves to their divinity — voluntary associations as well as blood-relations, mem- bers of a nation or a tribe as well as members of a totem- clan. And, once more, so long as the totem-clan is the earliest social organisation known in the evolution of society, so long will those who believe in the correlation of social and religious evolution look to the totem-clan as the earliest society of which the members could habitually worship a common deity.
But though I am ready to admit, and even to insist, that very few of the deities known to us are transfigured totems, still my argument does require me to maintain that a totem- plant or animal may under stress of circumstances develop into a non-totem deity. Whether M. Marillier does or does not consider this process of evolution possible, I can- not quite make out. If he argues that a clan-totem is not a tribal deity,^ I quite agree. If he argues that a divine animal by the very fact that it becomes a tribal deity ceases to be merely the totem of one particular clan of the tribe,"* I agree again. And if he admits, as he seems to do,^ that an animal which originally was a clan-totem may become the deity of a whole tribe, then that is all I want. It is possible for a clan-totem to evolve into a tribal deity ; and the fact that in so evolving it ceases to be a totem is not, as M. Marillier strangely seems to imagine, an argument
' M. Marillier, III., p. 226. - Ibid., p. 230. ' I., p. 229.
- I., p. 231.
" I., p. 247.
