TRANSFER FACTORS IN MAZE LEARNING 



341 



entrance side; and (c) those involving turns toward the other two 

 sides and hence neither toward nor away from the food box. The 

 third (c) type were discarded as irrelevant or at least equivocal 

 for this problem. The blind alleys for the first (a) and second 

 (b) types were listed in two different columns for each successive 

 maze, and count was then taken for the number of times each 

 had been entered by all the rats of the group. Averaged totals 

 show that a blind alley turned in the food box direction was entered 

 on an average of 4.58 times by the four rats, while one turned away 



9- 



<3- 

 7- 

 6- 



z- 



3- 



/""/ G U /? <5 



I I T I I I I 



from food box was entered on an average of 1.59 times. This is a 

 substantiation of the rough observation first made : that the animals 

 came to show a fairly definite tendency in the food box direction 

 that was transferred from maze to maze. Evidence that this 

 tendency was gradually developed in successive runs, being ab- 

 sent at first and strong toward the end, is presented graphically 

 in figure 8. It shows the total number of entrances into mazes 

 of the (a) type and again of the (b) type for the whole group, each 

 divided by the number of mazes of each type offered. It is based 

 upon table 3. In considering this figure the reader should keep 

 in mind the fact that all errors tended to be reduced somewhat, 



