64 HELEN B. HUBBERT 



the longest, 14.2 If a rat remained in the maze for 15 minutes 

 without reaching the food box he was taken out and replaced 

 in the entrance box for a second attempt. Distance and time 

 were recorded in the same way as for a successful run, i.e., when 

 an animal worked for 15 minutes on his first trial, and failed, and 

 then attained success after eight minutes on his second attempt, 

 the total time of his first trial would be 23 minutes. Distance 

 was treated in a similar way. 



In the Table 1 given above, the time record and distance 

 record of all individuals at work upon the problem at any given 

 trial were separately averaged, e.g., in trial 1, 27 animals were 

 used, the time and distance records were taken and then averaged 

 separately, giving one point on the time curve and one on the 

 distance curve respectively. Once the problem had been learned 

 by an animal it was taken from the group. As may be seen in 

 Curve 1, plotted from Table 1, the number of animals at work 

 is steadily decreasing. 



The striking similarity between the time and distance curves 

 bears out the contention of Watson and of Hicks that the time 

 record is a good index of progress in learning. It will be recalled 

 that this position has been severely criticised by Washburn 2 and 

 by Yerkes 3 . The similarity of the curve contour, and the close 

 correspondence of the maxima and minima is apparent. As 

 Hicks has pointed out, certain differences between T. and D. 

 appear in the early trials. The drop in time in the first nine 

 trials is 89.5 per cent, while that in distance is 78.5 per cent. A 

 partial explanation for the increased percentage of the time drop 

 may be found in the rat's behavior during the first few runs in 

 the maze, when he often crouches against some partition and 

 refuses to run for three, six, ten and sometimes even fifteen 

 minutes. In such cases the time average increases enormously, 

 while the distance average remains practically unchanged. 

 Perhaps a better method of procedure would have been to deduct 

 from the time record the time spent in absolute quiet, but this 

 would lead into difficulties of standardization, making necessary 

 an arbitrary decision as to how much time shall elapse without 

 movement on the part of the rat before deduction is justifiable— 

 a pitfall similar to that encountered in computing errors, and one 



« Washburn, M. F., Jour. Comp. New. and Psy., 1907, Vol. 17, p. 532. 

 3 Yerkes, Robert M., Jour Phil, Psy. and Sci. Methods, Vol. IV, p. 585.. 



