THE BEHAVIOR OF THE CROW 193 



standard stimulus plates in the stimulus shifter. Difference in 

 the illumination of the stimulus areas was eliminated. 



A 5 centimeter » circle versus a 2 centimeter » circle was 

 chosen for the beginning of this study. The correct exit was 

 indicated by the larger circle. 



This change in the conditions of discrimination naturally 

 threw the crows into confusion. They refused to enter the dis- 

 crimination-chamber unless forced to do so by the darkening 

 of the entrance-box. If this were done, they would pass to and 

 fro before the two stimulus-chambers, but they would not enter 

 far enough into either of them for the exit doors to be opened. 

 The series of the first two days had to be interrupted on account 

 of the crows' fright. On the third day no attempt was made to 

 work. During the day the crows were fed somewhat less than 

 the usual amount of food. The next morning (August 5th), 

 they were tried with a 9 centimeter circle versus a 5 centimeter 

 circle. By this change the illumination of the stimulus-cham- 

 bers was made to approximate that to which the crows had 

 become accustomed in the experiments on the discrimination 

 of intensity. Their hunger, on this day, was great enough to 

 overcome in large measure their fright. The results of this, 

 and the remaining series on size discrimination are given in 

 Table 2. 



After one series with the 9 centimeter versus the 5 centimeter 

 circle, a 2 centimeter circle was substituted for the 5 centimeter 

 circle. The crows' behavior now became practically normal. 

 The only significant difference from previous reactions was a 

 greater hesitation in choosing. Before finally entering a cham- 

 ber, they would often pass to and fro several times before the 

 two stimulus-chambers, again and again starting to enter one 

 chamber only to back out and go to the other. As appears in 

 the table, crow no. 1 made twenty correct choices in succes- 

 sion, while crow no.. 2 succeeded in choosing correctly eighteen 

 times in twenty. This sudden return of calm and controlled 

 reaction and the high percentage of correct choices, were due 

 probably to the fact that the illumination of the stimulus- 

 chambers through the 9 centimeter and the 2 centimeter circles 

 was closely similar to that in the experiments on intensity 

 discrimination. 



' Stimulus plates will be designated by the diameter or the side. 



