PATTERN-DISCRIMINATION IN VERTEBRATES 323 



would vary directly as the area exposed. If the luminous 

 intensity of the two screens were made equal, the brightness of 

 each screen would be inversely as its area. 



Suppose the two screens are covered by plates each containing 

 a window, the one 3 cm., the other 6 cm. in diameter. If the 

 two screens are equal in brightness, and if a complete image of 

 the larger be formed on the retina, it is evident, ceteris paribus, 

 that the animal will be stimulated by four times as much light 

 from the larger as from the smaller form. Thus it is possible 

 for the animal to acquire discrimination on the basis of difference 

 in quantity of luminous flux entering the eye, independently of 

 the difference of extensity of the stimulated parts of the retina. 

 Lashley 3 reports that his white rats began to react to the difference 

 of illumination of the respective alleys of the Yerkes box. To 

 swamp this difference he introduced an additional lamp directly 

 above the Yerkes box. This served the purpose, apparently, by 

 adding enough illumination (an equal amount to each alley) to 

 reduce the proportional brightness-difference to a point below 

 the threshold. It is evident that such procedure does not neutral- 

 ize the difference of luminous intensity of the respective screens. 

 If an equal brightness be added to each screen by the additional 

 lamp, or if the screens are protected from it, their respective 

 luminous intensities would be directly as their respective areas. 

 It is not clear that this was Lashley 's intention. Bingham 4 , 

 however, in his work on the chick, assumed this precaution to be 

 sufficient. The fact is masked to some extent by his use of the 

 word " intensities " (op. cit. p. 76) synonymously with " bright- 

 nesses ' in the present discussion. His birds discriminated a 

 circle 6 cm. in diameter from one 4.4 -f cm. in diameter. Bingham 

 reports no control tests to show that discrimination was on the 

 basis of size, rather than of luminous intensity of the stimuli. If 

 he made no control tests his conclusion is of course unwarranted. 



If the luminous intensity of the respective screens be equalized 

 by reducing the illumination of the larger, or increasing that of 

 the smaller, a difference in brightness is ipso facto established. 

 This difference also may serve as a basis of discrimination inde- 

 pendently of difference in size. 



3 Lashley, K. S.: Visual Discrimination of Size and Form in the Albino Rat. 

 Journal of Animal Behavior, 1912, pp. 310 ff. 



4 Bingham, H. C: Size and Form Perception in Gallus Domesticus. Ibid, 1913, 

 pp. 65 ff. 



