138 HAROLD C. BINGHAM 



apparatus allows the experimenter to control the conditions of 

 setting by means of artificial illumination. His criticism would 

 be valid for similar experiments conducted in natural and un- 

 controlled light. With the visual discrimination uninfluenced 

 by setting, the perception could not have been of patterns. 



Another feature which Hunter has overlooked is one of method. 

 Referring to figure 1, he asks: 4 " If an animal is trained on dia- 

 grams 1 and 2, is it any wonder that he breaks down when con- 

 *fronted by diagrams 2 and 3 ? ' Assuming now that the animal 

 actually sees ' ' a triangle or a circle each in more or less of a square 

 setting," no explanation is offered for the breaking down of the 

 discrimination when merely the size of the form was changed, 

 i.e., when the triangle of 1 was a circumscription or an inscription 

 of the circle of 2. (Witness table 8, series 12, 13 and 14, March 

 21-22, and series 10, March 28; also table 9, series 5, April 21). 5 



Now in these tests the patterns remained the same except in 

 size, but the reactions changed from a high percentage of cor- 

 rectness to a relatively low percentage. 



Besides this mis-statement of conditions there is an obvious 

 lack of agreement in the matter of defining " form." The so- 

 called " abstract sense " in which I have used the term has called 

 forth objections. 



In my study of form perception I was not concerned with genetic 

 phases of the problem. My task demanded an answer to the 

 . question : Does the chick perceive forms ? 6 Consequently, it 

 makes no difference whether or not the conception of form, to 

 which I have given expression, is the result of development. One 

 might consider that phase of the subject, but in my problem I 

 was justified in determining whether or not the chick perceives 

 form in this " abstract sense." 



Now if our animals fail to perceive circularity and triangularity 

 as such, there are several principles that we should not lose sight 

 of. In the first place, we should not try to excuse our animals 

 nor become over-dogmatic in theorizing about extraneous, or 

 even allied problems. We should accept as a fact the conclusion 

 to which the evidence points. 



Moreover, we should seek to determine and define just what 



• Op. Cit., p. 331. 



' Join. Animal Behavior, 1913, vol. 3, No. 2, pp. 106 and 109. 



• The task would have taken on other complexities had positive results been 

 secured in the initial problem. 



