Professor Thorndike is of opinion that monkeys are probably imitative in ways beyond the capacity of dogs and cats; but, at the time of writing, he had not substantiated his opinion, by analogous experiments. If so, it will perhaps prove that they are rational beings in the narrower sense defined in a previous chapter of this work. For it appears that the kind of imitation which Mr. Thorndike’s experiments go far to disprove, is what we may term reflective imitation. A cat with no experience of the means of escape, one that has tried to get out of the box by chance efforts in many directions and has failed, sees another cat perform an act acquired in this way, and learns nothing from the sight. This, no doubt, proves that the cat had not in any sense grasped the nature of the problem before it, had no notion of just where the difficulty lay, had not the wit to see that the performance of the other cat supplied the missing links in its own previous behaviour. It is questionable whether such missing links could be supplied in this way in the absence of some powers of reflection. The cat is unable to form an association, leading to an appropriate act, from having seen another animal perform the act in a certain way, partly because it cannot perceive the reason of its previous failure, and see that the other’s performance affords the requisite clue. The whole gist of the chance experience interpretation of animal behaviour is that there must be chance experience to build on. The cat cannot gain this by looking on never so intently, unless it be provided with a rational as well as a sensory eye. The act of pulling the string has been reached by the successful cat through the gradual elimination of many failures; it is a differentiated act, having no place in the previous experience of the kitten. It has never entered into the conscious situation, and cannot be supplied at will by a non-rational being.