6 RUTLEDGE T. WILTBANK 



average number of trials required to learn the maze, of 

 the total number of errors and the total time per rat. 

 The percentages are given in table 1. 



TABLE 1 



Percentages in Savings in Trials, Errors and Time in Transfers Between 



Two Mazes 



Trials 



A to B 59 



B to C 33 



C to D 27 



D to E 69 



E to A 20 



E to D 15 



Although the present experimenter did not try, as did 

 Webb, to devise mazes which would differ so much from 

 one another that the learning of one would exert a detri- 

 mental influence upon the learning of the other, neverthe- 

 less an examination of plate I, figs. 1 to 5, will disclose 

 that there were marked differences between certain of 

 them. Consider for example mazes B and C, and note 

 the differences in the number, length and arrangement of 

 the blind-alleys. In the first section of maze B, the animal 

 at the end of the first runway turns to the right, but in 

 the corresponding section of C it turns to the left. In 

 B immediately after this there is a blind alley to the right, 

 while in C there is none either to the right or left. After 

 the second turn in C, the animal encounters a blind alley 

 to the left; after the second turn in B, it follows, the true 

 pathway. At the fifth turn in C, the animal faces a situa- 

 tion different from any to be met in B, for not only is it 

 at this point in C far more likely to meet the true pathway 

 at a right angle, but there is a blind alley at its left im- 

 mediately as it turns into the true pathway, a similarly 

 placed one being lacking in B. An examination of the 

 other three sections of the mazes will multiply discrepan- 

 cies of this kind. 



Again, consider the differences between A and E. In 

 the first section of the former, the true pathway consists 

 of three runways and the connecting runways, Q and R; 



