TRANSFER OF RESPONSE IN THE WHITE RAT 17 



animal. A response was counted correct provided the ani- 

 mal had fully entered one or the other of the return alleys 

 and turned back, when the stimulus was given, before it had 

 reached a point where there was any possibility of its seeing 

 the closed doors of the alley. After the animal had begun 

 to learn it was given the stimulus when at various places 

 to prevent the formation of position habits. Thus some- 

 times the stimulus was given just when the animal had 

 entered the alley, sometimes when it was half way down the 

 alley, and sometimes when it had come almost to the point 

 where the closed doors could be seen. By such a variation 

 the animal would not learn to stop at any particular place 

 but only when the stimulus was given. The omission of the 

 stimulus five times out of twenty trials also prevented the 

 formation of the turning habit regardless of the presenta- 

 tion of stimulus and served as an effectual control. The 

 standard adopted for the negative series was 100 correct 

 responses out of 105 chances to respond to the stimulus 

 given. Since the stimulus was given but 15 times out of 

 each series of 20 trials it required the completion of seven 

 such series with but five errors all told to meet the standard. 



