THE DELAYED REACTION IN CATS 



95 



TABLE III— Continued 



Rats- 

 No. 2. 

 No. 4. 

 No. 5. 

 No. 6. 

 No. 7. 



Number of 

 trials on 

 learning 



176 

 175 

 505 

 800 

 361 



Cats- 



Bobby. 

 Jim . . . 

 Tom. . 

 Fay . . . 

 Bess . . . 

 Phil . . . 

 Kitty . . 



Number of 

 trials on 

 learning 



130 

 110 

 170 



50 

 180 



70 

 110 



(c) Controls used. — In the construction of the apparatus, every 

 effort was made to eliminate all possible differences in the com- 

 partments which could be used as guides to correct reactions. 

 The backgrounds surrounding the entrances to the compart- 

 ments were all alike painted black. Since the backgrounds were 

 all of the same brightness, and, since everything remained con- 

 stant with the single exception of the exit doors to the com- 

 partments, controls were put in to determine their possible effect. 

 In order to test this, the three doors were all opened and the 

 tests were given by the usual method under conditions in all 

 other respects normal. The results were entirely negative. In 

 no case did an animal make use of the doors as cues to its reactions. 



Again, control tests were introduced to determine whether or 

 not the animals were really depending upon the applied stimuli 

 (light or sound) for cues for guiding their reactions. To test 

 this, experiments were made under normal conditions except 

 that each time the stimulus (sound or light) was withheld .30% 

 correct reactions was the highest made by any subject under 

 these conditions. It is clear, therefore, that normally the reac- 

 tions were made either to sound or to light. 



Not being able to secure the same pitch and intensity in each 

 of the three buzzers, control tests were made to determine whether 

 the animals had formed associations between them on the basis 

 of quality. The buzzers were all interchanged — buzzer a took 

 the place of b, b the place of c, and c the place of a. No case 

 was found where the differences in pitc'i and intensity were used 

 as cues for reaction. These qualitative differences could well 

 have been effective during the period of learning the association ; 

 but, on the delayed experiments, they could be of little or no 

 value. The essential cues in handling delays must be factors 



