COLOR VISION IN MAMMALS 511 



mice with both normal and 'hereditary rodiess' (Keeler) retinae. He 

 found reason to think that some individual mice have a rudimentary 

 color sense. Most of his animals could not discriminate colored papers 

 from grays or colored lights from white ones, but one mouse out of seven 

 could distinguish a red light from a white one — though the same individ- 

 ual confused red papers with gray ones. But we should expect that if red 

 light is of no stimulating value it would naturally be discriminated (as 

 darkness) from even a dim white light, whereas a red paper would not 

 be invisible, even though its redness was not registered, but would appear 

 gray and would be confused with gray papers. Despite this sort of criti- 

 cism of his work, Hopkins remains to the mouse what Walton is to the 

 rat: the sole claimant of color vision; and the same remarks apply to 

 both — their techniques are not discernibly superior, if equal to, those of 

 the larger number of other investigators who have found no reason to 

 think that murid rodents see hues as such. 



The rabbit has come in for some attention. The Watsons found, as in 

 the case of the rat, that after red-versus-green training, darkness could be 

 substituted for the red light without disturbing the animal in the slight- 

 est: red light is darkness for the rabbit. The animal was now trained 

 positive to a blue light and then was required to discriminate it from a 

 dim yellow, which was gradually brightened. The animal died before the 

 investigators were able to try a range of intensities sufficiently great 

 to exclude a brightness discrimination. They were only able to say that 

 blue light probably looks brighter to the rabbit than a yellow of the same 

 energy. 



At about the same time (1912) Washburn and Abbot were experi- 

 menting with six rabbits, using colored papers. The animals learned to 

 distinguish a red from a light gray, but could not tell the red from a dark 

 gray or a black. The results with blue-gray discriminations were not so 

 striking, but did permit a conclusion that only brightness guided the 

 rabbit to a choice. 



Again, as with the rat and mouse, there is conflicting evidence. R. H. 

 Brown, in 1936, claimed to have established a Purkinje shift in the 

 rabbit, from A,560m[i, to A,530m[l; but his procedure was altogether too 

 crude to support his conclusions. He established a reflex response to 

 colored light by conditioning with light and shock stimuli. His animals 

 were quickly made responsive to only one member of a pair of lights 

 (A,640m[l and ^490m[x), but his variation of the intensity of the nega- 

 tive stimulus was made in only three large steps — scarcely adequate to 



