COLOR VISION IN MAMMALS 509 



to yellow (A,595m^) and negative to darkness. When a blue (A,478m|i) 

 of low intensity was substituted as the negative stimulus, the rats con- 

 tinued to go to the yellow. But when the intensity of the blue light was 

 increased to a certain point, the rats broke down and made chance scores. 

 One rat was trained positive to red versus green. Removal of the green 

 stimulus confused the animal, which made chance choices; but removal 

 of the 'positive', red, stimulus had no effect. Obviously, the red stimulus 

 was no stimulus at all — the rat was blind to it. This shortening of the 

 red end of the spectrum is quite in keeping with the fact that the rat, like 

 all other known rodents, exhibits no Purkinje phenomenon either electro- 

 retinographically or pupilloscopically. This, despite the unquestionable 

 presence of some cones, in a proportion of perhaps one to every hundred 

 or more rods. 



Munn, in 1932, used colored papers with the rat and obtained only 

 negative results. Several years later, with Collins, he reinvestigated the 

 rat's perception of red light. The red stimulus was paired with a 'nega- 

 tive' white light and with darkness in alternate sets of trials, the object 

 being to make impossible any step-wise response always to the brighter 

 stimulus, and to avoid giving the rat any constant stimulus to which he 

 could become negative. The animal was thus forced to react positively 

 to redness alone if it could; but it proved unable to do so with any regu- 

 larity. The authors concluded that for the rat the brightness-relation of 

 the stimuli was of most importance, their absolute brightnesses secondary, 

 and that color discrimination — if any — was indeed weak. These results 

 verified those of Muenzinger and Reynolds, whose technique had been 

 similar except for the use of red, white, and black papers instead of red 

 and white lights and darkness. The rat had shown an ability to discrim- 

 inate red from gray, but with great difficulty when the gray was close to 

 black. This again would be expected if the rat is blind to long-wave light. 



The work of Coleman and Hamilton has been considered, by psy- 

 chologists, a model investigation. In 1933, they trained rats positive to 

 black versus red. When gray was substituted for the black, and when the 

 red was exchanged for a darker shade, the animals reversed their prefer- 

 ence. Reversal of the brightness relationship in other pairs of color 

 stimuli also inverted the responses of the rat. With some pairs, only 

 chance scores were ever made, showing that the two stimuli were not only 

 matched in brightness but had no difference for the animal as to hue. 

 WTien new rats were introduced to these 'confusion pairs' of colored 

 papers, they could never learn to go to one paper and avoid the other. 



