202 MIND IN EVOLUTION CHAP. 



bolt, and the dog never clearly saw the difference between 

 pushing it back all the way and part of the way. 



A special source of confusion on this head would be that if the 

 bolt were nearly but not quite back, an extra strong pull and a 

 vigorous shake of the box would shake it back and open the box. 

 Hence Jack was tempted to trust to strength rather than art. 

 Hence, also, he never quite perfectly learnt that the bolt must be 

 pushed before the lever was pulled, 1 though he seemed to get 

 nearly right on this point in the first ten trials. 



Upon the whole, whether by his own attempts or by 

 watching me, or by both methods combined, Jack learnt 

 that he had to push the bolt and then paw the lever, but 

 exactly how, or how far the bolt was to be pushed, he did 

 not understand. His perceptions, as indicated by his 

 method of dealing with the bolt, were, to use Mr. Small's 

 expression, " crass. " He knew that a lot of pulling with 

 the teeth or pushing with the paw at the bolt was the 

 thing wanted, but he had no "idea " of the object of the 

 pushing and pulling i.e. y he did not seem to direct it 

 definitely to the effecting of a particular physical change as a 

 step towards his desire. In this respect his action con- 

 trasted with, for example, the behaviour of the cat in 

 pulling out the other bolt, and with his own behaviour in 

 pulling up the string, and in other experiments to be 

 mentioned later. 



A further point of interest in this experiment was the 

 impotence of a casual success. 



Jack actually succeeded at the third trial, but clearly without 

 understanding how, for at the next attempt he tried to pull the 

 bolt the wrong way. A second sporadic success occurred much 

 later, and at the forty-fourth trial 2 came a third success, after 

 which there were few complete failures, 3 but the work remained 



1 In the last fifty-eight trials he tried the lever first seven times, three at 

 the beginning of one day's series, two at the beginning of another, and 

 the remaining two quite sporadically. 



2 There was a gap of two or three days after the first fourteen trials, as 

 Jack in his efforts with the lever broke it. 



3 Three or four times I showed him without waiting to see if he would 

 manage it himself the main interest being to see whether he would get 

 really perfect. His average time during the last three days was, roughly, 

 37", 29", and 32*. The time frequently fell to 5", and sometimes rose 

 to 70 or So' 7 . 



