TRAINING IN WHITE RATS UPON VARIOUS SERIES OF MAZES 19 



tive transfer. This fact comes out upon a comparison of 

 the records of the control-groups and trial-groups concerned. 

 Out of 420 errors made by the control-group in learning 

 the maze, 102 were made in the first blind alley, and the 

 rest were distributed among the other blind alleys. Out 

 of a total of 223 errors made by group 2 in this maze, 

 166 were made in the first blind alley; and out of a total 

 of 212 made by group 1, 126 were made in the first blind 

 alley. It was the large number of trials necessary to elim- 

 inate these errors made in the first blind alley which resulted 

 in the negative transfer, as may be seen from an examination 

 of tables 17 to 20, and figure 9, in which respectively the 

 number of errors and the percentage of errors made in 

 each blind alley by each group are shown. 



