70 



Retention. 



A retention test was male on five indiviiuala of the 65 

 day group v\ho were caused to relearn the problem after 90 days. 

 Durin/T this time they were fed daily in the maze except that at 

 the 85th day the food supply was cut down, and on the 89th day 

 no food at all was allowed. Probably a better plan would have 

 been to feed the rats in the food box of the mare for a week 

 preceding the retention test, usinp the same schedule employed 

 in preliminary feeding, and keeping the food box carefully parti- 

 tioiied off from the rest of the maze. 



Seventy-six percent of the original number of trials 

 were required to relearn, forty-eight percent of the time nec- 

 essary for learning was occupied in relearning, and fifty-two 

 percent of the original amount of distance was covered. The 

 absolute time when learning was 7.9 seconds, when relearning 

 9 seconds, this difference probably being due to the increas- 

 ed age, since the rats were approximately 200 days old at the 

 time of the retention test and absolute time increases with 

 age. 



Nothing more is shown by the test on retention than that 

 the interval between learning and relearning must be made very 

 much smaller if it is desired to begin a problem v/ith a view to 

 determining the curve of retention. 



The relation of time to distance in learning, and the 

 matter of elimination of alleys in the maze have been discussed 

 at length in papers already published. 



