namely which constitute the true abstract idea of the botanist. In sum, the generic image comes half way between individual representation, and abstraction properly so called. It results almost exclusively from the faculty of apprehending resemblances. The rôle of dissociation is here extremely feeble. Everything takes place, as it were, in an automatic, mechanical fashion, in consequence of the unequal struggle set up in consciousness between the resemblances which are strengthened, and the differences, each of which remains isolated. 2. It has been said that the principal utility of abstraction is as an instrument in ratiocination. We may say the same of generic images. By their aid animals reason. This subject has given rise to extended discussion. Some writers resent the mere suggestion that ants, elephants, dogs, and monkeys, should be able to reason. Yet this resentment is based on nothing but the extremely broad and elastic signification of the word reasoning—an operation which admits of many degrees, from simple, empirical consecutiveness to the composite, quantitative reasoning of higher mathematics. It is forgotten that there are here, as for abstraction and for generalisation, _embryonic_ forms—those, i. e., which we are now studying. Taken in its broadest acceptation, reasoning is an operation of the mind which consists in passing from the known to the unknown; in passing from what is immediately given, to that which is simply suggested by association and experience. The logician will unquestionably find this formula too vague, but it must necessarily be so, in order to cover all cases. Without pretending to any rigorous enumeration, beyond all criticism, we can, in intellectual development, distinguish the following phases in the ascending order: perceptions and images (memories) as point of departure; association by contiguity, association by similarity; then the advance from known to unknown, by reasoning from particular to particular, by analogical reasoning, and finally by the perfect forms of induction and deduction, with their logical periods. Have all these forms of reasoning a common substrate, a unity of composition? In other words, can they be reduced to a single type—of induction according to some, of deduction according to others? Although the supposition is extremely probable, it would not be profitable to discuss the question here. We must confine ourselves to the elementary forms which the logicians omit, or despise, for the most part, but which, to the psychologist, are intellectual processes as interesting as any others.