16 

 for retention after an interval of thirty days ^ and others 



after sixty days. The averages of the rats which had 

 an int rval of sixty days are alone given. Besides this 

 data, Tables I-b and I-c give in addition, for each rat, 

 the number of trials as .Tell as the number of days in 

 h the problem was learned. 

 The most important faet to be notised when examining 

 these tables in succession, is that rats which were given 

 one trial a day, required fewer trials than those given 

 three or five trials, and that those given three trials 

 fewer than those given five trials. It is just the re- 

 verse, seemingly, when taking into account the number of 

 days as three and five trials required fewer days than 

 one trial daily. On closer examination it will be seen 

 that a few rats were trained s beyond the time 

 when they had reached the norm of one second or two sec- 

 onds. This was done to obtain a satisfactory norm. 

 Such a prolongation of trials was scarcely harmful to 



