196 CHARLES A. CO-BURN 



disturb them. The results on these days of calm steadiness, 

 as a rule, showed an increase in the number of correct choices. 

 On other days, their behavior would be practically the opposite. 

 While still in the entrance-box they would walk impatiently to 

 and fro before the woven wire side of the box. When the en- 

 trance door was opened, they would often start several times 

 to enter only to turn back into the entrance-box. When they 

 finally did enter, they would rush to one of the exit doors, and, 

 in a crouching attitude, wait until it was opened. On these 

 days, great care had to be taken in opening and closing the 

 doors for an unusual noise or sudden movement would greatly 

 increase their excitement. During this behavior they were very 

 likely to develop a position habit. Series, in which this excited 

 behavior resulted in a considerable number of incorrect choices, 

 have been noted in the tables. 



The ability of the crow to pass directly from one set of cir- 

 cles to another with no great difference in the number of cor- 

 rect choices (see Table 2), was further tested by a series of 

 experiments, the results of which appear in Table 3. 



In these experiments, the attempt was made to determine 

 whether the crows were reacting to a certain specific stimulus, 

 or whether they were reacting to it because of its relation to 

 another stimulus. For instance, if the 6 centimeter and the 4 

 centimeter circles were presented, and the crow trained to react 

 positively to the 6 centimeter circle, would it continue to do 

 so when the 6 centimeter circle was presented with a 9 centi- 

 meter circle, or would it, instead, choose the larger area in each 

 instance? 



As in the preceding series the crows were trained to choose 

 the larger of two circles. When they had gained the ability to 

 choose correctly, they were given ten trials with a different pair 

 of circles. During these ten trials, they were rewarded after 

 each test, regardless of the correctness or incorrectness of the 

 reaction. A reaction was considered correct if the crow chose 

 the larger circle. These series are designated, in Table 3, " rela- 

 tive reactions." The training series which preceded the relative 

 series of August 26th are given in Table 2. 



The results of these experiments indicate fairly clearly the 

 relativity of the crows' reactions. Especially is this true of 

 crow no. 1. For example, on August 24th and 25th, when the 



