I. INTRODUCTION 



The object of these experiments, stated in general terms, 

 was to study the kind and the degree of transfer of training 

 in white rats which were allowed to learn a maze after 

 they had learned, either completely or partially, one or 

 more other mazes. Specifically, the object was to find 

 out: 



1. Whether there was transfer of training between a 

 pair of mazes which were learned completely one after 

 the other, and if so whether it was positive or negative. 

 These expressions, " positive transfer " and " negative 

 transfer," are familiar to those interested in this phase of 

 psychological research; but it may not be useless to re- 

 mark, for the sake of some chance reader, that by the 

 former is meant that the mastery of one or more mazes 

 renders easier the mastery of a subsequent one, while by 

 the latter is meant that the mastery of one or more mazes 

 renders more difficult the mastery of a subsequent one. 



2. Whether, if either positive or negative transfer is 

 present, it will persist through a series of mazes; and 

 whether, if both present and persistent, it will also be 

 cumulative. 



3. The effect upon the learning of a maze of having 

 partially learned a preceding maze. 



4. The effect of the complete learning of a maze upon 

 the mastery of another maze already partially learned. 



5. Whether it is more advantageous to learn two mazes 

 in succession; or to partially learn one, then completely 

 learn the other, and finally perfect the learning of the 

 former. 



6. The effect upon the relearning of a maze of having 

 learned four intervening mazes between the original learn- 

 ing of the maze and its relearning. 



Diagrams of the mazes are presented in Plate I, figs. 1-5. 

 In all the mazes except C the blind alleys were twenty 



1 



