TRAINING IN WHITE RATS UPON VARIOUS SERIES OF MAZES 43 



affirm with equal confidence at what part of the process 

 the overcoming of fear and the association between running 

 to the end of the maze and obtaining food begin to tell. 

 With regard to the transfer-effect according to the 

 time-criterion, it will be observed that the degree of transfer 

 did not increase regularly with the increase in the number 

 of trials in the first maze of the pair, except in the average 



2 Tr/als in E 



4 Tna/s 



/fffer 

 STnals 



/Ifter 

 /6 Tria/3 



/9fter Comp/efe 

 Leafn/na of £ 



PercenftPT^e^ of Transfer 

 of Tra/nin^ in Groups 

 Learning the f Maze 

 after the fhrtial Learning 

 of the O Maze • on the 

 Bas/s of the A^/era^e yiDta/ 

 Number of Errors ang^ the 

 Average Total Time per Rat. 



Tr/a's 



Errors 



Firne 



total time per rat of the group carried over from E to D. 

 But a regular increase of this kind was brought to light 

 when the transfer was reckoned on the basis of the average 

 surplus time required by the rats to run the mazes. This 

 surplus time was obtained by subtracting from the average 

 time per trial within a group the average time of the last 

 trial made by the rats of the group that completely learned 

 the same maze. 



