1 82 MIND IN EVOLUTION CHAP. 



impulse theory. Mr. Thorndike himself obtained nega- 

 tive results from his experiments, which he generalises. 

 But here again his experience is surely at variance with 

 common knowledge. I have seen dogs in process of learn- 

 ing tricks by being held in position. Wundt * explains 

 that he taught his poodle to shut the door in that fashion 

 in four or five lessons moreover ; and my wife, who had 

 a pug which shut the door to admiration, tells me that she 

 taught it in the same way. The regular method of teach- 

 ing elephants is, as I am informed by the experienced 

 elephant trainer at the Belle Vue Gardens, Manchester, for 

 the keeper to guide the trunk from point to point. 2 To 

 test Mr. Thorndike's assertion, I myself made a little ex- 

 periment with a kitten, which seems clear enough. I 

 showed the kitten some food, fish or milk, and then, setting 

 the dish for a moment on another table near, I lifted the 

 kitten on to a desk, and handed it the food. I did this 

 six times without result, the kitten merely mewing and 

 begging in its own fashion. The seventh time the kittei 

 jumped on to the desk of itself. 3 After this I left hom< 

 and it had no more lessons for a fortnight. On my retun 

 I found that it jumped on to the desk at the first trial, an< 

 after that the act became habitual. In this case the kittei 

 did the thing regularly after doing it once ; and it nevei 

 had any inducement to get on to the table, such as might 

 produce a nascent impulse, in seeing the food there whil( 

 it was below. It saw the food in my hand, near but not 

 on the desk, or else, as I bent to pick it up, on the tabl< 

 near. 



No doubt cats and dogs differ, and so do their trainers, 

 but even a few positive instances are sufficient to destroy 

 Mr. Thorndike's generalisation. 4 



1 Lectures on Human and Animal Psychology, p. 360. 



2 See below, p. 198. 



3 On the sixth trial, the kitten got on to a chair as if to get up, but 

 went down again. Except on one day, when I gave it two lessons, the 

 experiments were, roughly, at intervals of twenty-four hours. 



* Mr. Small, whose remarks (American Journal of Psychology, Jan. 

 1900) go to the root of the matter, points out that any "Attempt ... to 

 restrain or constrain the animal would defeat the desired end attention 

 would assuredly be diverted from the objective point to the restraining 

 conditions would be lost in the affective absorption induced by them " 

 (p. 163). And commenting on Mr. Thorndike's experiments, he says : 



