WHITE RAT AND THE MULTIPLE CHOICE METHOD 233 



on 10 new settings to check the possibility of memorizing the 

 particular settings rather than actually solving the problem. 

 One of the rats did the check series perfectly, two others with 

 only one mistake, and the other with two mistakes. It is evident 

 that all formed the habit of selecting the compartment at the 

 extreme right. 



A number of interesting individual tendencies were manifested 

 in the course of the experiment. A's method of reaction was 

 usually as follows: passing down the right wall of the reaction- 

 chamber to compartment No. 1, she turned leftward and followed 

 along close to the compartments until she reached the first open 

 door. In many instances she turned from the right wall before 

 reaching compartment No. 1, with the result that she entered 

 a door too far to the left. This factor was of course more opera- 

 tive with settings comprising doors at the right end. It is evi- 

 dent from table I that settings 1, 3, 5 and 10 yield a larger number 

 of incorrect first choices than do the others. With settings 

 farther toward the left the first approach to the doors brought 

 the subject more often in front of a closed door and hence the 

 first open one encountered was correct. On Sept. 9, over seven 

 weeks after the discontinuance of A's training, she was given 

 10 trials with the original settings and made 8 correct choices 

 out of 10. Evidently the habit was still strong. 



D manifested a tendency entirely different from that of A. 

 He ran along the left wall of the reaction-chamber and across 

 in front of the open doors until he reached a closed one, where- 

 upon he turned back into the last open one. This turning back 

 appeared to be caused by the touch of the vibrissae on the closed 

 door rather than by the sight of it. The turning was always 

 in the same direction, — leftward or toward the doors. D ex- 

 perienced more difficulty than the other rats with setting No. 2, 

 i. e., doors 8 and 9. Going down the left wall he often entered 

 compartment No. 9, whereas he normally turned at that point. 

 As he did poorly also on the other settings involving No. 9, it 

 would appear that he was reacting by tactual rather than visual 

 criteria, — habitually going along the wall until he encountered 

 an obstacle and then turning across, etc. The directness with 

 which No. 9 was entered in contrast with any other door was 

 noticeable. A month after the training ended, although sick 

 and able to move only slowly, he performed two trials correctly 



