The study which we are undertaking is therefore a way of taking up again, _but under new conditions_, the old problem of the origin of religion. To be sure, if by origin we are to understand the very first beginning, the question has nothing scientific about it, and should be resolutely discarded. There was no given moment when religion began to exist, and there is consequently no need of finding a means of transporting ourselves thither in thought. Like every human institution, religion did not commence anywhere. Therefore, all speculations of this sort are justly discredited; they can only consist in subjective and arbitrary constructions which are subject to no sort of control. But the problem which we raise is quite another one. What we want to do is to find a means of discerning the ever-present causes upon which the most essential forms of religious thought and practice depend. Now for the reasons which were just set forth, these causes are proportionately more easily observable as the societies where they are observed are less complicated. That is why we try to get as near as possible to the origins.[3] It is not that we ascribe particular virtues to the lower religions. On the contrary, they are rudimentary and gross; we cannot make of them a sort of model which later religions only have to reproduce. But even their grossness makes them instructive, for they thus become convenient for experiments, as in them, the facts and their relations are easily seen. In order to discover the laws of the phenomena which he studies, the physicist tries to simplify these latter and rid them of their secondary characteristics. For that which concerns institutions, nature spontaneously makes the same sort of simplifications at the beginning of history. We merely wish to put these to profit. Undoubtedly we can only touch very elementary facts by this method. When we shall have accounted for them as far as possible, the novelties of every sort which have been produced in the course of evolution will not yet be explained. But while we do not dream of denying the importance of the problems thus raised, we think that they will profit by being treated in their turn, and that it is important to take them up only after those of which we are going to undertake the study at present. II But our study is not of interest merely for the science of religion. In fact, every religion has one side by which it overlaps the circle of properly religious ideas, and there, the study of religious phenomena gives a means of renewing the problems which, up to the present, have only been discussed among philosophers.