100 JOSEPH U. YARBROUGH 



the cats were making the last half of the response in the absence 

 of the stimulus, no difficulty yet appeared. 



Delay III. — On this type of delay, where the stimulus was 

 cut off the moment the cat left the release box, Bess was given 

 100 trials. Of this number 84 were correct, with 54 of the last 

 CO correct. Phil made 56 correct reactions out of 70 trials, 

 making a percentage of 80. These results mean that, having 

 started correctly, the cats are able to retain their cue for cor- 

 rect reaction even though the remainder of the reaction must 

 be made in the absence of the stimulus. 



Delay IV. — In this delay, Bess was given 60 trials, 55 of which 

 were correct, and, of the last 40 trials, 39 were correct. Phil 

 was given 60 trials with only 60% correct. Of the last 30 pre- 

 sentations, 18 were reacted to correctly. Table V shows Phil's 

 tendency to drop out compartment b whenever the delays set in. 



TABLE V 

 Daily Record on Delay 4 With Sound 



Number Number Distribution of errors 



2 22 5 



Two seconds delay. — On this two seconds delay, Bess was 



given 130 trials and of this number 116 or 89% were correct. 



Of the last 80 trials, 74 were correct; with the percentage of 92, 



she was promoted to the four seconds delay. Phil was given 



* From January 23rd to February 3rd, Phil was being retrained on delays II and 

 III, receiving 10 trials each day. 



