Recent Biology 381 



these views be true, as there seems reason to believe, then 

 imitation holds a remarkable position in relation to intelligence 

 and instinct. It stands midway between them and aids them 

 both. In some functions it keeps the performance going, and 

 so allows of its perfection as an instinct ; in others it puts a 

 stress on intelligence, and so allows the instinct to fall away 

 if it have no independent utility in addition to that served by 

 intelligence. 1 In other words, it is through imitation that 

 instincts both arise and decay that is, some instincts are 

 furthered and some suppressed, by imitation. And all this is 

 accomplished with no appeal to the inheritance of acquired 

 characters, Professor Groos agreeing with Weismann that the 

 operation of natural selection as generally recognized is sufficient. 

 The difficulty which I see to this conception of play as a 

 pure instinct is that which is sometimes urged also against 

 considering imitation an instinct, *>., that it has no definite 

 motor coordinations, but has all the variety which the different 

 play forms show. If the definite congenital plays are considered 

 each for itself, then we have a great many instincts, instead of 

 a general play instinct. But that will not do, for it is one 

 of Professor Groos' main contentions, in the chapter on the 

 psychology of animal plays, that they have a common general 

 character which distinguishes them from other specialized 

 instinctive actions. They are distinguished as play actions, 

 not simply as actions. This difficulty really touches the kernel 

 of the matter, and serves to raise the question of the relation 

 of imitation to play ; for imitation presents exactly the same 

 conditions a general instinct to imitate, which is not exhausted 

 in the particular actions which are performed by the imitation. 



1 In a private communication Professor Groos suggests that the two views 

 may well be held to supplement each other. The case is very much like 

 that of early intelligence, in the form of association ; where it fully accom- 

 plishes the utility also subserved by an instinct, it tends to supersede the 

 instinct ; otherwise, it tends to the development of the instinct (Groos, 

 p. 64). (See p. 140 above and cf. the same writer's Play of Man, translated 

 by the same hand, in which the principal suggestions of this notice have 

 been taken account of by Professor Groos.) 



