Passing to the second question--Does all conscious imitation tend to reproduce the initiating stimulus?--we cannot unreservedly give an affirmative answer. It is true that when a child more or less successfully reproduces a sound which falls upon its ear, a like sound stimulus is afforded which may by a circular process incite to renewed effort, and lead to yet more successful reproduction. But when Professor Baldwin’s child, between nine and ten months old, imitated certain movements of the lips, there was no reproduction of the initiating visual stimulus. A chick seeing its companions run away or crouch will follow suit; and this would commonly be termed an imitative action; but there is here no reproduction of the initiating stimulus. Very much of the behaviour which is usually ascribed to imitation produces effects in consciousness quite different from that of the original stimulation. It is only by selecting one’s examples that one finds in them evidence in favour of Professor Baldwin’s “circular process.” Since, therefore, this circular mode of activity is neither a characteristic of all conscious imitation, nor a distinguishing mark of all adaptive organic action, the grounds on which Professor Baldwin bases his extended usage of the term appear to be fallacious. And in this usage we cannot follow him. Turning now to Professor Thorndike’s very different contention--that animals even so high as the cat and dog do not imitate in the sense of forming an association leading to an act from having seen another animal perform the act in a certain way--we may first describe some of his ingenious experiments designed to submit the matter to the test of observation under controlled conditions.[77]