£3 

 ing than when one, three or five trials were given. 



Tables I-d and I-e,when compared with Table I-a,show that 

 still fewer trials were required to learn the problem, and 

 of the two, I-d and I-e,the latter table indicates that a 

 trial- every t is the more economical in the way 

 of trials. In both cases, of course, a longer time elapsed 

 before the problem was learned. The number of trials 

 found necessary for the rats to learn the problem, when 

 trials were given every other day and every two days, are 

 also indicated in the tables and on the distribution curve, 

 Plate II. 



To the tables just described, some additional data 

 has been made. About the time the rats were being experi- 

 mented with for the additional problem on the latch box, 

 Professor Donaldson, of .Vistar Institute j suggested the 

 value of having the spinal co?-d and brain weights of the 

 rats in connection with each rat's record. The rats were 

 sent, art or the retention tests, to \7istar Institute and 



