III. TRANSFER OF TRAINING THROUGH SERIES OF MAZES 



The Persistence of the Transfer 



If we accept the statement that there is positive transfer 

 of training when rats are carried over from one maze com- 

 pletely learned to another which is then also completely- 

 learned, the question arises as to whether the transfer of 

 training will persist and retain its positive character if the 

 learning is continued through a series of mazes. It might 

 be presupposed with some plausibility that habits acquired 

 in situations which present marked peculiarities in the 

 midst of general resemblances, as do the situations in 

 the various mazes, would conflict with one another in such 

 a way that the transfer would cease to be positive. On the 

 other hand, it might be presupposed with perhaps equal 

 plausibility that in these situations practise tends toward 

 perfection, and the ability which an animal has to thread 

 the maze may consequently be expected to improve in 

 proportion to the number of mazes it learns to run. 



It was to throw light upon this question that the present 

 experiment was undertaken. The five groups of rats were 

 allowed to continue their learning through series of mazes, 

 each group learning one series. The series were composed 

 of the same five mazes, but the mazes were learned in a 

 different order by each group, according to the scheme 

 presented in table 6, in which the letters represent mazes 

 and the numbers groups of rats. 



TABLE 6 



The Order in Which Various Series of Mazes Were Learned by Various 



Groups 



Mazes ABCDEABCD 



Group 1 



13 



