34 RUTLEDGE T. WILTBANK 



TABLE 30 



The Ranking of the Groups upon the Mazes, According to the Average 



Time per Trial 



Maze A Maze B Maze C Maze D Maze E Av. Rank 

 Group having learned 



/o?<r preceding mazes 1 3 4 1 2 2.2 



Group having learned 



/Aree preceding mazes 4 2 2 2 1 2.2 



Group having learned 



two preceding mazes 3 1 1 3 3 2.2 



Group having learned 



one preceding maze .2 4 3 4 3 3.2 



It is apparent from tables 26 to 30 that, as in the case 

 of trials so in that of errors and time, an increase in the 

 number of preceding maze-learnings does not necessarily 

 and invariably result in a decrease in errors or in time. 

 But that there is a tendency for the errors to decrease 

 with the increase in the number of maze-learnings may 

 be seen from the average rankings. 



The Theory oj Transfer 



The foregoing paragraphs show that the degree of trans- 

 fer is not a function of the serial position of the maze. 

 It appears to be a function of the maze to which the transfer 

 is made, irrespective of the position of this maze. Thus 

 maze E, whatever its position, tends to be the one in which 

 according to all the criteria the greatest savings are effected. 

 The degree of transfer may be a function also of the maze 

 or mazes from which the transfer is made. This became 

 apparent in our search for the causes underlying the in- 

 stances of negative transfer in trials on maze D. The 

 transfer is not a function of the difficulty of the maze, for 

 the E maze — to use it again as an example — was not the 

 least difficult maze, as measured by the records of original 

 mastery; and the B maze, which was the most difficult 

 with respect to original mastery does not appear as the 

 one in which the least savings were made. 



In attempting an explanation of the presence and per- 

 sistence of positive transfer, we may assume that the 

 animal having learned one or more mazes and being in- 

 troduced into a new one is to be looked upon as an organism 



