THE DELAYED REACTION IN CATS 103 



hesitated often from one to three seconds on a single reaction, 

 but their occurrence was so irregular and their duration so 

 brief that their measurement and tabulation were very difficult. 

 Therefore, no period was recorded in seconds unless it was of 

 considerable duration. However, all hesitations were entered 

 in the notes. 



2. Two Compartment Experiments 



A. "Delayed" experiments. — The delay work was continued 

 in the two compartment tests by the usual method. The series 

 of presentations of boxes was changed from 



One of these three series of ten had been used each night. Each 

 one was used an equal number of times and at no time was 

 "one given twice in succession. In this way no one series was 

 given twice within three days. On the two compartment tests 

 the number of series was increased to four, as follows: 



These were taken in their order beginning with the first, and 

 no one was, therefore, given twice within four days. 



(a) Cats tested on light. — It will be remembered that in table 

 IV Jim is reported to have made 29 out of the last 30 trials 

 correct, after a four seconds delay. As his work progresses on 

 the two compartment experiments, the period of delay increases. 

 Since a very large proportion of his errors on the three com- 

 partment experiments were due to a tendency to drop out the 

 middle box, he would be expected to make a higher percentage 

 of correct reaction with this box omitted. Such is shown to 

 be the case in the data below. 



Of the 40 trials given Jim on six seconds delay 34 were cor- 

 rect. He escaped from the laboratory on the third day, after 

 his work, and after 36 hours absence was recovered and made 

 80% on 10 trials. Feeling sure that the cat was experiencing 

 no difficulty, the experimenter increased the period of delay to 



