55 

 Ho observable difference in the responses evoked 



by each trial, when several trials were given in succes- 

 sion, was recognized. It cannot be said that the rat was 

 less active in the last of the succeeding trials. Fre- 

 quently the last of three and give trials was accomplished 

 in a shorter time than the ones preceding. The changes 

 which must take place cannot be detected by observation 

 of the animal's behavior; they are evidently physiological. 

 Probably when the time between trials is of such a length 

 that the physiological state created by the preceding 

 one is permitted to "set", so to say, before the next one 

 is given, every trial is effective. learning is dependent 

 upon every successful and effective trial. If several 

 trials are given in succession, only the physiological 

 state created by the first trial is of advantage in 

 learning, the others do not, to any great extent , reinstate 

 the conditions of the first trial; time must intervene 

 between trials so that the physiological state of the 



