6 JOHN LINCK ULRICH 



having had one trial were permitted to feed a short time in the 

 problem box, and were then removed and fed in their cages. 

 Those given three and five trials were allowed, after each trial, 

 to nibble a little food. At the end of all trials, these rats, hke 

 those given one trial, were fed in their cages. No harm seemed 

 to have resulted from not feeding the rats entirely in the problem 

 box after experimentation was over; on the other hand, there 

 were material advantages gained by not doing so. Time was 

 saved in experimentation, and a variety of food was given which 

 would have been impossible if the rats were fed only in the 

 problem box. 



The time for each trial was taken by a stop watch in minutes, 

 seconds and half seconds. Later the seconds were converted 

 into decimals of a minute. Time spent by the rats nosing about 

 the entrance box when it was opened was not counted, but 

 only from the moment the rat entered the hood covering the 

 latch box until the problem was solved. 



Daily experimentation was continued until a rat reached a 

 norm, or an average of one second per trial and held it for two 

 successive days. This norm was required of all rats regardless 

 of the number of trials which they were given each day. In 

 other words, in the last two days of experimentation, a rat had 

 to attain an average of one second in two trials with one trial 

 daily, an average of one second in six trials with three trials 

 daily, and an average of one second in ten trials with five trials 

 daily. When an animal lifted the latch of the latch box in an 

 unusual way, or habitually went around the box before attack- 

 ing the latch, the time was increased considerably, and in these 

 rare cases a norm of two seconds was taken. 



The usual norm of one second was not a difficult one to reach, 

 for the more active rats learned how to lift the latch in one- 

 half of a second. Such quick time as one-half of a second was 

 more often attained by a rat given one trial than those given 

 three or five trials daily. When either norm was reached, learn- 

 ing was assumed to be complete. 



After completion of learning the rats were held for retention. 

 The retention tests were taken after an interval of sixty days. 

 In the very beginning of this work, the rats were tested after 

 thirty days, but retention was so perfect at the end of this time 

 that the interval was doubled. During the sixty days of no 



