blee est dissoute, quand, nous retrouvant seul avec nous-memes,
nous retombons a notre niveau ordinaire, nous pouvons niesurer
alors toute la hauteur dont nous avions ete souleve au-dessus de
nous-nienie " (p. 299).
This answer is somewhat disappointing. First of all, we feel a little suspicious of a theory which sees the origins of religion in crowd phenomena. Again, from the point of view of method, we are at a loss. Above we had been dealing (with some difficulties) with a transcendental collective subject, with a "society which was the creator of religious ideas " : " Au reste, tant dans le present que dans Thistoire, nous voyons sans cesse la societe creer de toutes pieces des choses sacrees " (p. 304). Then society was the divinity itself, i.e. it was not only creator, but the object of its creation, or at least reflected in this object. But here society is no more the logical and grammatical subject of the metaphysical assertions, but not even the object of these assertions. It only furnishes the external conditions, in which ideas about the divine may and must originate. Thus Prof. Durkheim's views present fundamental inconsistencies. Society is the source of religion, the origin of the divine ; but is it " origin " in the sense that " the collective subject . . . thinks and creates the religious ideas > This would be a metaphysical conception deprived of any empirical meaning; or is society itself the "god," as is implied in the state- ment that the " totemic principle is the clan," thought under the aspect of a totem ? That reminds one somewhat of Hegel's Absolute, "thinking itself" under one aspect or another. Or, finally, is society, in its crowd-aspect, nothing more than the atmosphere in which iiidividtials create religious ideas ? The last is the only scientifically admissible interpretation of the obscure manner in which M. Durkheim expounds the essence of his theories.
Let us see how our author grapples with actual and concrete problems, and which of the three versions of "origins" just men- tioned he applies to the actual facts of Australian totemism. He starts with the remark already quoted about the double form of the social life of the Central Australian tribesman. The natives go through two periodically changing phases of dispersion and agglomeration. The latter consist chiefly, indeed, almost ex-
