14 WALTER S. HUNTER 



alley, thus preventing retracing; but once the rat went through 

 this alley, its choices were in accordance with its one kinaesthetic 

 tendency, to turn to the right. In several cases some evidence 

 was found indicating the r61e of the end-stop. Let us take the 

 case of a rat who ran twice to the right meeting failure at the first 

 and success at the second choice. This success, before the posi- 

 tion habit was thoroughly established, would lead the animal 

 to run to the left. It was found that if success (a free pathway) 

 and not failure (a blocked pathway) were offered the rat on the 

 first trial, that he would at the next trial go to the left. 



We may now turn to a brief description of the behavior of rat 

 110 who learned to run the simple alternation temporal maze. 

 This rat, as stated above, also developed position habits which 

 required breaking by all three of the methods there described. 

 The process of learning was long and tortuous. At the close of 

 the 59 trials, rat 110 could run the continuous path of Irlrlrlrlr 

 without error, save that it usually retraced at least once per 

 trial. This retracing was usually running from some point in 

 the side alleys back along the top of the T to the end-stop of the 

 opposite side. In a space maze, this would probably be counted 

 as two errors. The fastest time made by this animal was 52 

 seconds. Without apparent cause, the time fluctuated between 

 this and 105 seconds. The response never became so automatic 

 and stable as is customary in a space maze. After the habit was 

 mastered, only the entrance-stop was used regularly. The end- 

 stop was quickly and quietly inserted only when the rat made 

 a wrong choice. It does not seem probable that the lack of 

 stability in the habit was due to distracting influences. The 

 cause of the instability was rather the difficultSUes of the task. 



The following results were secured with controls used with rat 

 110: (1) When twenty continuous choices per day were given as 

 opposed to ten, no disturbance was apparent. This control 

 doubled the length of the maze. The animal practically con- 

 tinued to run successive figure 8's until removed from the maze. 

 (2) When the entrance-stop was taken out of the maze and not 

 used after the first trial, the rat failed to turn into the central 

 alley. (3) The maze was rotated 180 degrees and the rat was 



