1. During the first period, this extension is the work of the imagination, rather than of generalisation properly so called. By an instinctive tendency, well-known, though not explained, man concludes for intentions, a will, and a causality analogous to his own, in the medium that acts and reacts around him: his fellows, all living things, and whatever else by their movements simulate life (clouds, rivers, etc.). This is the period of primitive fetishism that is fixed in mythologies and languages. It may actually be observed in children, in savage races, in brutes (as in the dog that bites the stone by which it is hit), even in rational man, when—becoming again for the moment a creature of instinct—he falls into a passion at the table that has hurt him. This period corresponds fairly well with that of generic images, because the idea of cause thus generalised results from gross, external, partial, accidental resemblances, which the mind perceives almost passively. It cannot be doubted that the higher brutes have a generic image of causality; i. e., they are capable—given an antecedent—of invariably representing to themselves the consequence. This mental state, which has been termed “empirical consecution,” and which is not infrequent even among men who may never rise beyond it, resolves itself into a permanent association of ideas, the result of repetition and of habit.[123] All this, however, is merely an external conception of causality, of its form, and not its nature; it is an outside view, an approximation. The proper characteristic of this period is that it remains subjective, anthropomorphic, representing cause as an intentional activity, which produces movements only in view of an end. 2. The second period begins with philosophic reflection, and proceeds by the slow constitution of the sciences. Its development may thus be summarised: little by little it deprives the notion of cause of its subjective, human character, without however completely attaining this ideal end; it reduces the essentials of the concept to a fixed, constant, and invariable _relation_ between a determined antecedent and consequent; hence it sees in cause and effect only the two moments, or aspects of one and the same process, which is fundamentally the affirmation of an identity. Here imagination recedes, to make way for abstraction and generalisation,—for abstraction since it is less a question of terms than of a certain relation between the terms, for generalisation because the natural tendency of the mind is to extend causality to the whole of experience.