Secondly, this selection of the intelligent involves the survival of those in whose higher brain-centres there is room for a greater range and variety of interconnection by means of associating fibres. It involves a selective survival of the larger and more finely organized brains. It is probable, as Professor Ray Lankester has recently indicated, that the ridiculously small-brained mammals and reptiles of the past were creatures of instinct with little capacity for intelligent control. Their lives were simple, and their enemies and competitors no better provided with higher brain-centres than themselves. Stereotyped instinctive behaviour sufficed to enable them to hold their own, and meet the requirements of a life of dull and unprogressive monotony. Strength without cunning made these big-framed animals for a while masters of the situation. But among those existing animals whose skeletons indicate an analogous zoological position, there is none which exhibits a cerebral development so poor. And we may fairly conclude that the fact that these huge creatures have left no lineal descendants may be taken as evidence of the importance and value, in evolution, of that cerebral tissue which is the organic basis of intelligence. The higher brain contains the potentiality of that experience without which the evolution of intelligent behaviour in any race of vertebrate animals is impossible. V.--THE INFLUENCE OF INTELLIGENCE ON INSTINCT We have seen that the relation of instinct to intelligence is essentially that of congenital to acquired behaviour. We have seen, too, that in the Lamarckian interpretation what is acquired in the course of life may be transmitted through inheritance, and thus the intelligent behaviour of one generation may become instinctive and congenital in the next. But serious biological difficulties stand in the way of the acceptance of this interpretation; there is, moreover, little or no evidence of the assumed transmission to offspring of any acquired modifications of structure or behaviour. We have, therefore, been led to infer that instinctive behaviour has been evolved through the selection of adaptive variations of germinal origin, the influence of intelligence being restricted to the fosterage of co-incident variations, that is to say, of those congenital variations which coincide in direction with the acquired modifications of behaviour due to intelligence. It is clear that on this interpretation the influence of intelligence on instinct is more indirect and less simple than that implied by the Lamarckian hypothesis. Intelligence and instinct are in large degree independent, though there is continual interaction between them. We have now to consider the nature of this interaction, and to this end we must indicate the relation of acquired modifications to the hereditary groundwork of the animal constitution.