49 

 as to whether one, three or five trials were most "benefi- 



cial in retention. But the retention tests, when three 

 problems were given abreast, seem to show that they are 

 not as good as when one problem is given. This is the 

 most interesting data obtained in these tests, other than 

 the faot that relearning is much easier than learning. 



The results are not conclusive. What is needed 

 for retention tests are better systematic methods of 

 preparation after learning is complete. The length of 

 the period of preparation should be determined and more 

 attention paid to daily exercise, diet and disturbing 

 influences which might affect retention tests. The 

 principal disturbance noticed was that aroused by odors. 

 It is an old obstacle often appearing in the work of 

 learning, and how to control it is a difficult matter. 

 The first trial of retention is often prolonged because 

 of the detection of a new odor. Daily exercise and 



