THE DELAYED REACTION IN CATS 101 



60 trials with 51 or 85% correct. Of his last 40 trials, 35 were 

 correct. 



Four seconds delay. — One hundred and seventy trials were 

 given Bess with 121 correct responses. Of the last 30, only 18 

 were correct. With this low record, it was thought best to 

 return to shorter delays before trying to advance. From 

 January 24th, 1916, until February 14th, she was given 10 trials 

 daily on delay IV and on two seconds delays. Of the 200 trials 

 given during this period 120 were given on the two seconds 

 delay, the last 30 of which netted 28 correct reactions. This 

 high percentage of correct reaction on the last 30 is due to the 

 dropping out of the middle box ; so, also, may the low percentage 

 of correct reaction immediately preceding be explained by the 

 tendency to drop out the middle boxes the delays were length- 

 ened. The percentage of correct responses in the last 30 trials 

 immediately preceding the dropping out of the middle box was 

 66, while the percentage of the first 30 after its being dropped 

 out was 94. 



(c) Maximal interval of delay with three boxes. — 



In table VI the maximal delays attained by my cats are 

 given, and for comparative purposes similar data on Hunter's 

 animals and Walton's dogs are included. The reader should 

 remember that these tests were made with a choice of three 

 boxes and that training stopped here because of a well developed 

 tendency to drop out the middle box. In the case of Phil, the 

 last cat reported in the table, this tendency was not well devel- 

 oped. As is shown in the table, he was making a good record 

 on the two seconds delay, and there are no indications that 

 he could not have bridged a longer period of delay with three 



boxes. 



TABLE VI 



