Thus there are rites without gods, and even rites from which gods are derived. All religious powers do not emanate from divine personalities, and there are relations of cult which have other objects than uniting man to a deity. Religion is more than the idea of gods or spirits, and consequently cannot be defined exclusively in relation to these latter. III These definitions set aside, let us set ourselves before the problem. First of all, let us remark that in all these formulæ it is the nature of religion as a whole that they seek to express. They proceed as if it were a sort of indivisible entity, while, as a matter of fact, it is made up of parts; it is a more or less complex system of myths, dogmas, rites and ceremonies. Now a whole cannot be defined except in relation to its parts. It will be more methodical, then, to try to characterize the various elementary phenomena of which all religions are made up, before we attack the system produced by their union. This method is imposed still more forcibly by the fact that there are religious phenomena which belong to no determined religion. Such are those phenomena which constitute the matter of folk-lore. In general, they are the debris of passed religions, inorganized survivals; but there are some which have been formed spontaneously under the influence of local causes. In our European countries Christianity has forced itself to absorb and assimilate them; it has given them a Christian colouring. Nevertheless, there are many which have persisted up until a recent date, or which still exist with a relative autonomy: celebrations of May Day, the summer solstice or the carnival, beliefs relative to genii, local demons, etc., are cases in point. If the religious character of these facts is now diminishing, their religious importance is nevertheless so great that they have enabled Mannhardt and his school to revive the science of religions. A definition which did not take account of them would not cover all that is religious. Religious phenomena are naturally arranged in two fundamental categories: beliefs and rites. The first are states of opinion, and consist in representations; the second are determined modes of action. Between these two classes of facts there is all the difference which separates thought from action. The rites can be defined and distinguished from other human practices, moral practices, for example, only by the special nature of their object. A moral rule prescribes certain manners of acting to us, just as a rite does, but which are addressed to a different class of objects. So it is the object of the rite which must be characterized, if we are to characterize the rite itself. Now it is in the beliefs that the special nature of this object is expressed. It is possible to define the rite only after we have defined the belief.