If an answer must be given to the question whether instinct or intelligence has priority in the course of the evolution of behaviour, it may be urged that, on theoretical grounds, the claims of instinct are the stronger. Taking animals as we actually find them, however, they afford numberless examples of behaviour at first instinctive but subsequently modified, in greater or less degree, in accordance with the teachings of experience. Let us, first, assume that the environment is slowly changing, or has changed, in some definite manner. Such change would, of course, be relative, and might be due, either to new conditions brought to bear on the animal, or to the animal being itself brought, in the expansion of its life, within their influence. The old instinct is no longer quite adapted to the changed circumstances. If the change were sufficient in amount, and occurred somewhat suddenly, variations of instinct might not occur soon enough to enable the animal to reach adaptation by the gradual process of natural selection. If dependent on instinct alone the animal would, under these circumstances, be eliminated. But if intelligence were able to modify the behaviour to meet the new conditions this elimination would be prevented. In successive generations intelligence would constantly modify behaviour in the same manner and in a definite direction. Meanwhile congenital variations in different directions would occur. Those which were in directions antagonistic to that dictated by intelligence would tend to thwart accommodation and render it less effectual; but those which were coincident in direction would conspire with accommodation and render it more effectual. The individuals in which variations of instinct tended to thwart intelligence would be eliminated; while those in which coincident variations assisted and aided intelligent modification would survive. Thus intelligence would lead the way along lines which congenital variations would follow. And in the course of a number of generations the new instinct would reach the fully adaptive level, and further modification by intelligence would become unnecessary unless the environment continued to change yet more. Individual accommodation of behaviour would in this way determine the direction of instinctive variation; and yet throughout the process there would be, strictly speaking, no transmission of the intelligently acquired characters of the behaviour.