21 6 BULLETIN OF THE BUREAU OP FISHERIES. 



He maintains (191 2) that Phoxinus fed for some time exclusively on yellow meat 

 was able to distinguish between yellow and any one of 24 different shades of gray, some 

 of which were very much darker and others very much lighter than the yellow. This 

 discrimination, he holds, could only have been made on the basis of difference in color. 

 He further contends that it is possible to understand the changes in color during the 

 breeding season, so conspicuous in the males, and adaptation in color to the bottom 

 only on the assumption that fishes have color- vision. 



All of this evidence, however, does not convince Hess. He maintains (1913) that 

 some of the fishes, which during the breeding season become most highly colored, red 

 and yellow predominating, live at such a depth that these colors are entirely eliminated 

 by absorption and consequently that they can have no bearing on vision whatever. 

 He repeated and extended Frisch's experiments on feeding and adaptation, using the 

 same species but, he asserts, improved methods, and maintains that he obtained no 

 evidence of discrimination among different colors and no evidence of adaptation in 

 color to the bottom. • He concludes (p. 439) that Frisch's statements regarding the 

 behavior of Phoxinus are wrong, "samtUch unrichtig." 



In order to answer these charges Frisch extended his feeding and other experiments 

 on Phoxinus. He used 50 shades of gray ranging from black to white and also blue, 

 green, yellow, and red. He maintains that the fish distinguished between any of these 

 colors and all the shades of gray, also between any combination of blue, green, and red, 

 but not between yellow and red. He also maintains, as previously stated, that the fish 

 assumed a color similar to that of the bottom; and that, while Hess may be right in his 

 assertion that some species which become highly colored during the breeding season 

 live in water so deep that yellow and red rays do not penetrate, nearly all fishes in which 

 this occurs live in shallow water. Moreover, he holds that the fact that yellow and red 

 at a given depth appear gray to the human eye does not prove that they affect the eye 

 of the fish in the same way. 



Unless the results of Frisch's observations, especially those on feeding, can be shown 

 to be erroneous they seem to support strongly the idea of color-vision in fishes. The 

 experiments here made on Paralichthys and Ancylopsetta support this idea quite as 

 strongly and prove conclusively that Paralichthys and Ancylopsetta assume a color 

 similar to that of the background, ranging from dark blue to dark red, and that this 

 correlation can not be accounted for on the basis of difference in brightness or energy, 

 and that it is dependent upon the length of the light-waves, and, moreover, that the 

 color assumed is dependent upon stimuli received through the eyes. 



Now the essential objective characteristic of color-vision in man consists in the fact 

 that the stimulation resulting in sensation of color is dependent upon the length of the 

 waves and not upon luminous intensity, at least not in certain respects. On the basis 

 of this phenomenon as a criterion it necessarily follows, from the facts stated above, 

 that flounders also have color-vision. It may, however, be contended that this is not 

 a satisfactory criterion of color-vision, and, in a certain sense, this is perfectly true. 

 But is there any criterion that is more satisfactory ? 



In the process of adaptation in color we have a response dependent upon the quality 

 of light, wave-length, and upon the visual apparatus in the eyes. It is known that the 

 stimuli involved pass through the optic nerves to the brain and that they are distributed 



