Although I have presented here only a few of the leading features which an analysis of the imitative processes reveals, enough has been said to show how complicated and difficult the problem is, and to render advisable a general summing up in more compact form of the results of these somewhat rambling observations. It will not do to call imitation instinct and leave it at that, since it is not a specific but quite an involved reaction. Moreover, the condition of imitation, namely, the tendency of movement ideas to produce corresponding movements, is not itself instinctive; but we have seen that this tendency alone does not explain all that we include under the name of imitation. This tendency of the movement ideas must have special grounds furnished by organic needs, and especially those which are instinctive; when the general idea of movement is coincident with one of these the impulse toward discharge becomes very strong. We cited in illustration of this the general movement-impulse, nursing, curiosity (how is it done?), belligerence (not only as regards distinctly hostile movements, but sensation as well), recognition, and illusion. If there is nothing else, then imitation taken alone is no instinct; it is only in very close connection to instinct, as our biological point of view has shown. It is, however, probable that these limits are not reached by the simplest imitation, such as coughing, gaping, etc., and use may be made of the hypothesis of transference (_loi de transfert_) from specific social instincts, which are themselves the result of a certain degree of imitativeness of the movement idea (agreement, answering, and the like) to movement itself in cases involving the movements belonging to a species. By this means natural selection of whatever developmental factor is employed acquires an essential impetus. Whoever regards such collaboration as probable will consider imitation as a phenomenon at least similar to instinct.