TRAINING IN WHITE RATS UPON VARIOUS SERIES OF MAZES 47 



of the habit completely formed in the second maze with 

 the habit nearly formed in the first maze. Where the 

 transfer was made after two, four or eight trials, the forma- 

 tion of the maze-habit had not progressed nearly so far, 

 and a larger number of errors were being made than were 

 being made by the sixteen-trial group when it was trans- 

 ferred. The consequence was that with these groups 

 any interference with the imperfect maze-habit would not 

 render futile, to all intents and purposes, many of the 

 trials which had already been made, and that furthermore 

 any interference with the imperfect habit, including a 

 relatively large number of errors, would be not unlikely 

 to correct an error instead of diverting the animal from 

 the true pathway. 



Table 41 presents the number of trials necessary to 

 learn the maze — the trials before the interruption plus 

 the trials after the interruption, and also the percentages 

 of savings and losses. The savings and losses were com- 

 puted by comparing the records thus obtained with the 

 records of the control-group for the particular maze, and 

 besides being shown in the table just named, they are 

 presented graphically in figs. 15 and 16. Accordingly, 

 the total number of trials in D before and after the learning 

 of E were compared with the number of trials in D before 



TABLE 41 

 The Average Number of Trials Required in a Maze Partially Learned 



BEFORE THE COMPLETE LEARNING OF A SECOND MaZE AND COMPLETELY 



Learned after the Learning of the Second Maze; and the Per- 

 centages OF Transfer of Training 



Maze D Maze E 



Trials Percentages of Transfer Trials Percentages of Transfer 



(Two Trials in Partial Learning) 

 17. 9± 7.8 27.36 8.43±2.20 63.51 



(Four Trials in Partial Learning) 

 16.43± 4.61 33.39 15.70± .42 32.06 



(Eight Trials in Partial Learning) 

 23.00± 5.23 6.77 14.40±2.24 37.66 



(Sixteen Trials in Partial Learning) 

 29 .1± 4.70 —17.91 24.00±4.00 —3.90 



(Control-groups: Group 4 for D, Group 5 for E) 

 24.67±12.67 23.10±7.10 



