MENTAL LIFE OF MONKEYS AND APES 73 



With the above reactive tendencies and modifications of 

 method in mind we may continue our description of results. On 

 June 9 there developed a tendency to increase the magnitude 

 of the original error by choosing nearer the left end of the groups. 

 This is odd, since one would naturally suppose that an animal 

 as intelligent as the orang utan would tend to avoid the general 

 region in which success was never obtained and to focus atten- 

 tion on the right, as contrasted with the wrong end of each group. 

 It obviously contradicts the law of the gradual elimination of use- 

 less activities. In other words, it is wholly at variance with the 

 principle of trial and error exhibited by many infrahuman organ- 

 isms. Julius, although making many mistakes, worked dili- 

 gently and, for the most part, fairly rapidly. The day's work 

 proved most important because of the change in method and 

 also because of the appearance of hesitation, the rejection of cer- 

 tain boxes, and the definite choice of others. My notes record 

 "this is a most important day for Julius in problem 2;" but 

 subsequent results do not clearly justify this prophecy. 



The method of choosing the first box at the left and then of 

 moving down the line until the right one was reached was so 

 consistently followed that during a number of days it was pos- 

 sible for me to predict almost every choice. Indeed, to satisfy 

 my curiosity in this matter during a number of series I guessed 

 in advance the box which would be chosen. The percentages 

 of correct guesses ranged from ninety to one hundred. June 10, 

 for example, yielded two series for which the ratio of right to 

 wrong first choices was to 10, and in which the method described 

 above was used consistently throughout. 



It was inevitable that punishment by confinement and the 

 discouragement resulting therefrom should interfere with the 

 regularity of work and make it extremely difficult to obtain 

 strictly comparable results from series to series and from day to 

 day. The data for this problem, as presented in table 9, have 

 values quite different from those for the monkeys, chiefly because 

 of the more variable conditions of observation. 



It was occasionally noted that the disintegration of a definite 

 method and the disappearance of the tendency on which it 

 depended occurred rather suddenly. Frequently it happened 

 that having used an inadequate method fairly persistently on 

 a given day, the animal would on the following day exhibit a 



