x SOME EXPERIMENTAL RESULTS 225 



This was a case of sheer mechanical habit, gradually overcome. 

 The seizing of the wrong string became more and more mechanical, 

 so that after pulling it the dog would not wait to see whether 

 the biscuit had fallen down but would go to the new string at 

 once, even with the old string still in his mouth. 



It must be remembered, however, that the penalty of mistake 

 was a very slight one ; he always pulled the right string immedi- 

 ately after the wrong one therefore virtually lost nothing by it. 



So far the experiments were not decisive. The first 

 series showed that Jack could discriminate the strings, 

 and preferred that which he learnt to be connected with 

 the food. The second series, on the other hand, testifies 

 forcibly to the strength of habit. Later on therefore I 

 made a second series of experiments in which he saw the 

 biscuit put up, but could not see it after it was put up. 

 Three strings hung down side by side, at intervals of 

 about a foot ; and I put the food first on one card, and 

 then on another, varying them quite irregularly. The 

 result showed a rapid growth of discrimination, which 

 soon became nearly perfect. 



Jack, who had not had the string experiment for some time, 

 was very much excited, and at first kept pulling down the strings 

 before I had time to put the meat up. He then began pulling in 

 very random fashion, and only chose the right one first three 

 times in the first fifteen trials. Soon, however, he became more 

 attentive. In the next fifteen trials he was right seven times, and 

 seven times also in the next ten. The experiment was resumed 

 in the afternoon under stricter conditions. He was held back a 

 yard or two, and I counted three after the biscuit was placed, and 

 before he was let go. In the first four experiments he was right 

 twice, but had some difficulty at the beginning in seeing the 

 strings. In the next ten, he was right seven times, and one trial 

 was doubtful. 1 During the course of these trials he gave up 

 pulling down the remaining strings after getting the biscuit. 



I now made a further experiment to test the effect of 

 irrational imitation. Putting the biscuit on one of the 

 cards, I pulled it down, and let him take it. I then placed 

 a biscuit on another card, to see whether he would be 

 guided by what he now saw, or by what I had just done. 

 The case was closely analogous to the ordinary experi- 



1 He jumped between two strings, but caught the right one. 



Q 



