128 MIND IN EVOLUTION CHAP. 



perience, an element in a state of excitement similar in its 

 tendency and character. 



c. But it is also possible, and in some cases probable, 

 that more or less random movements of the animal itself 

 play a part. The fish that first happens to come near the 

 surface will get the pick of the crumbs, and this fact will 

 tend to confirm or a stamp in " the random mode of re- 

 action. In the instance quoted above from Preyer, 1 irrita- 

 tion of the skin set up random reflexes, of which some 

 gave relief, and were accordingly preferred. Apart, how- 

 ever, from the simplest cases, in which the originally 

 random movement meets immediate u confirmation," there 

 remains for this explanation a certain difficulty which has 

 been very clearly pointed out by Mr. Thorndike, 2 who has 

 made much more liberal use of the explanation than any 

 other writer. Except in these cases, the " confirmatory " 

 wave must come after the initial movements of the action 

 leading up to it are well over. How then can it help to 

 establish them ? To this the first reply is, Can it do so ? 

 That is to say, apart from intelligence, is any complicated 

 or lengthy series of actions learnt in this way ? If so 

 which remains to be proved I imagine that the "confirm- 

 atory wave " must be conceived 3 as gradually spreading, as 

 it were, backward. We can imagine the delight of feeding 

 and the impulse to come to the best place for it attached, 

 first to the sight of the food itself, then to the keeper who 

 brings it, then to the sound of his approaching footsteps, 

 and finally, perhaps, to the click of a gate which he opens 

 on his way. 4 Remoter associations of this kind are un- 



1 P. 42. 2 Animal Intelligence, pp. 103, 104. 



3 That is to say, if (i) we assume (as above, p. 122) that the primitive 

 form of assimilation rests on fusion of excitements through the medium of 

 a unitary conation, and (2) we treat these cases as mere extensions of the 

 primitive form. 



4 Thus, in a succession of experiments on the nymph of the mayfly, 

 Mr. J. E. Wodsedalek first of all by repeated efforts induced the nymph 

 to follow bits of food. This took several weeks. Gradually the insect 

 was brought to the part of the dish nearest to the operator, when it was 

 allowed to feed, and after about four weeks most of the specimens would 

 frequently swim after the food when brought near them, and would often 

 swim towards him when he made his experiments (J. A. B., 1912, pp. 12, 

 13). In organisms of very low type we can observe the spreading process 

 even in cases of reaction to simple sense stimuli. Messrs. Fleure and 

 Walton (Zoologischer Anzeiger, Band 31, 1907, p. 215) gave a sea 

 anemone scraps of filter paper, one every twenty-four hours, placing it on 



