Such are the beliefs relative to the soul and its destiny, in their most primitive form, and reduced to their most essential traits. We must now attempt to explain them. What is it that has been able to lead men into thinking that there are two beings in them, one of which possesses these very special characteristics which we have just enumerated? To find the reply to this question, let us begin by seeking the origin attributed to this spiritual principle by the primitive himself: if it is well analysed, his own conception will put us on the way towards the solution. Following out the method which we have set before ourselves, we shall study these ideas in a determined group of societies where they have been observed with an especial precision; these are the tribes of Central Australia. Though not narrow, the area of our observations will be limited. But there is good reason for believing that these same ideas are quite generally held, in various forms, even outside of Australia. It is also to be noted that the idea of the soul, as it is found among these central tribes, does not differ specifically from the one found in other tribes; it has the same essential characteristics everywhere. As one effect always has the same cause, we may well think that this idea, which is everywhere the same, does not result from one cause here and another there. So the origin which we shall be led to attribute to it as a result of our study of these particular tribes with which we are going to deal, ought to be equally true for the others. These tribes will give us a chance to make an experiment, as it were, whose results, like those of every well-made experiment, are susceptible of generalization. The homogeneity of the Australian civilization would of itself be enough to justify this generalization; but we shall be careful to verify it afterwards with facts taken from other peoples, both in Australia and America. As the conceptions which are going to furnish us with the basis of our demonstration have been reported in different terms by Spencer and Gillen on the one hand and Strehlow on the other, we must give these two versions one after the other. We shall see that when they are well understood, they differ in form more than in matter, and that they both have the same sociological significance.