TRAINING IN WHITE RATS UPON VARIOUS SERIES OF MAZES 7 



in the first section of E, there are five runways in the true 

 pathway, and there is only one connecting runway, for 

 at T, U and V the animal passes directly from one runway 

 into another, although instead of turning at about a right 

 angle to do so it must turn through 180 degrees. In the first 

 section of A, there are three blind alleys, and in the cor- 

 responding of E there is but one. In the second section 

 of E, the true pathway turns abruptly to the left at the 

 entrance to the section, while in A a blind alley opens at 

 this place. In the third section of E, two blind alleys are 

 placed side by side; in the third section of A the blind 

 alleys alternate. In the fourth section of E, the entering 

 animal turns into the true pathway to the left; in the cor- 

 responding section of A, there is a blind alley to the left. 

 But there was positive transfer as between these two mazes, 

 judged by all three criteria. 



The Effect of Identical Parts 



In order to discover whether the presence of identical 

 parts would affect the kind or the degree of transfer, the 

 first sections of mazes A and B were made alike, and also 

 the third sections of mazes D and E. The savings in 

 trials and errors in the transfers between these mazes with 

 identical parts were notably greater than in the transfers 

 between the mazes without identical parts, as may be 

 seen by reference to table 1. The saving in time in the 

 transfer between A and B was the largest of all the time- 

 savings, but that between D and E was equalled by that 

 between E and A. 



If the greater positive transfet in the case of the mazes 

 with identical parts was due to the presence of the identi- 

 cal parts,— and it would seem too large to dismiss as merely 

 fortuitous,- — still it is not plain that this greater transfer 

 was brought about within the identical parts themselves. 

 It would seem reasonable to maintain that, if the presence 

 of the identical part in the second maze exerted a favor- 

 able influence upon the learning of the maze, the benefit 

 would be most likely to appear in this identical part, 

 although it is conceivable that the beneficial effect might 

 appear in some other part of the maze. For the purpose 



