STUDIES ON THE MENTAL LIFE OF ANIMALS. 557 



no utility; the association was formed in the same way as if each liad 

 been phiced alone in the box. 



Besides, one would expect that if the association between the sight 

 of the string- and the act which results from pulling it had been formed 

 by seeing the guide cat pull the string the subject w^ould pull at the 

 string as soon as he saw it; or, if there was a slight hesitation, a little 

 indecision, such a period would be extremely short, at least it woidd 

 be nearly constant for all subjects. Now, none of these expectations 

 was fulfilled. No subject pulled the string on being placed in the 

 compartment. The only exception w^as merely an apparent one — No. 

 6, in the midst of irregular struggles to get out, chanced to hit the 

 string with its paw. The association between pulling the cord and the 

 act of getting out was so slight that he remained in the box for 16 

 seconds before he noticed that the door was open. Cats No. 7 and No. 

 5 always failed, notwithstanding the fact that they fixed their atten- 

 tion upon the guide cat in an unmistakable manner 43 and 33 times, 

 respectively, and in a doubtful manner 111 times for No. 7 and 68 

 times for No. 5. 



No. 3 alone succeeded very promptly in getting out in three experi- 

 ments, and in 3 minutes and 30 seconds. He finally succeeded in 8 sec- 

 onds; but it should be noted that imitation was not necessarily a factor, 

 for this cat showed in the course of numerous experiments signs of a 

 much more lively intelligence than the others. He probably got out by 

 his own effort alone. Besides, when a subject gets out of the box, if 

 the success was due to imitation there ought to be a regular relation 

 between the time which he observed his guide and the time employed 

 in getting out; the latter should be in inverse proportion. Yet it is 

 the opposite that is observed — the longer the time of observation the 

 more time was lost in efforts to get out, or, indeed, the attempt ended 

 in complete failure. 



VI. 



If the experiments are varied the same results are reached. For 

 example, two cats were placed in the same cage, one of which, unac- 

 cjuainted with the means of exit, sees the door opened and goes out and 

 is fed with him. The experiment is repeated a number of times, then 

 the subject is placed in the box alone. No modification of his behavior 

 is noted which reveals the influence of imitation; the association is 

 formed no quicker; the method employed was different from that used 

 by his companion. Cat No. 1 opened the door ])y pulling at a loop 

 with his teeth; cat No. 7. the subject, pulled it with his paw. In 

 another 1)0X which could, at the will of the animal, be opened in two 

 ways cat No. 3 pulled at a loop at the back of the box, while cat No. 5 

 pulled a string at the front. 



