COLOR VISION IN AMPHIBIANS 493 



could he very well apply Buddenbrock and Friedrich's principle of pair- 

 ing off colors that matched the same gray. He did find that yellow-green 

 and its spectral neighbors could be equated to grays, and decided that in 

 the case of the frog (in contrast to the lizard) this meant that they were 

 gray to the animal, yellow-green thus being a neutral point in his spec- 

 trum. At intensities below .04 lux, any color could be equated to a gray 

 — this being the realm of pure rod activity, A marked Purkinje phenom- 

 enon was found, but above 30 lux there were no further changes in the 

 relative brightnesses of different colors. When the drum was striped with 

 alternate red and blue, and the stripes made progressively narrower, the 

 reaction was inhibited when the visual angles subtended by the stripes 

 were twice the threshold values for black and white stripes. Birukow con- 

 cluded that the 'visual acuity for colors' was only half of that for black 

 and white. Ignoring the fact that the differences in albedo of the adjacent 

 stripes were far from the same in the two cases, he correlated these find- 

 ings with the fact that the cone-to-rod ratio in the frog's area centralis is 

 1 :2. Again, ignoring the fact that while rods may play no part in color 

 vision, cones do play a part in black-and-white vision, he related his find- 

 ings to the fact (earlier demonstrated by himself) that it is the rods 

 rather than the cones which, in the frog, set the retinal limits of resolving 

 power. From all this he drew confirmation that the rods play no part in 

 color vision. 



The use of the optomotor reaction as a means of studying animal 

 visual acuity has been severely criticized. Apart from this however, does 

 the frog's compensatory reaction to red and blue versus gray stripes of 

 any and all albedos prove color vision? There is grave doubt of it. The 

 average investigator, finding that he could obtain a yellow-gray match 

 that abolished the reaction, would certainly not give up on red-gray com- 

 binations until he had tried many close grades of gray. And as a matter of 

 fact, a close perusual of Birukow's report reveals that his animals did have 

 matches of gray and blue, despite his conclusion drawn to the contrary. 



Again, the reaction to blue versus red of any and all albedos could 

 have a purely physical basis. Considering such factors as chromatic aber- 

 ration, it is hard to imagine how the parade of contours between the red 

 and blue stripes could be made to disappear even for a totally color-blind 

 animal for whom the red and blue were exactly matched in brightness. 

 The optomotor reaction is no more reliable as a means of studying color 

 vision than for tests of visual acuity. 



