4 WALTER S. HUNTER 



these abortive attempts at learning and the conditions under 

 which they were made are as follows. 



Rats 30, 33, 34 and 35 of the previous problem, simple alter- 

 nation, were each tested on double alternation for 550 trials, ten 

 trials daily. An interval of twelve seconds between each trial 

 was consumed in feeding and re-inserting the rat. None of the 

 rats made any progress toward the mastery of the problem during 

 these fifty-five days. 



Rats 5, 6, 7, 8, 16, and 18, untrained rats two months old, 

 were trained on the problem of double alternation, under the 

 above conditions. Rat 18 received 464 trials. The other re- 

 ceived between 600 and 650 trials. At the conclusion of the 

 training period, no rat had made significant progress toward 

 mastery of the problem. 



Rats 5, 6 and 16 were now tested on the same problem under 

 slightly different conditions. The end-stop was placed, not in 

 the side alley as indicated in the figure, but just beside the point 

 of choice (c) so that the animals were forced to run in double 

 alternation without the possibility of error. Each rat was given 

 400 trials under these conditions. At the close of this period, 

 the end-stop was placed as usual in the side alley. Twenty 

 trials were now given each rat with the possibility of choice 

 again present. No rat made better than 60 per cent of correct 

 reactions. 



In the tests so far described in this section, failure may have 

 been caused either by the length of time interval between trials 

 or by the fact that in the series of ten trials six were to the left 

 and only four to the right. In order to test this, three new rats 

 (51, 52, 53) were tested on double alternation giving eight trials 

 daily, llrrllrr. Punishment was used, but food was only given 

 at the close of the day's work. As soon as a rat ran to the exit, 

 he was immediately re-inserted for another trial. There was 

 thus practically no interval between trials. Rats 51 and 53 re- 

 ceived 512 trials. Rat 52 received 488 trials. At the close of 

 these tests, no rat had made significant progress toward mastery. 



Rats 2, 3, 4, 100, and 101, untrained and two months old, were 

 tested on this problem with four trials daily, llrr. There was an 



