COLOR CHANGES AND ADAPTATION IN FISHES. 215 



VISION. 

 COLOR-VISION. 



Several investigators have maintained that they have experimentally demonstrated 

 that fishes have color-vision, but in every case the validity of the evidence has been 

 seriously questioned. 



Washburn and Bentley (1908, p. 140), working with the creek chub (SemoHlus 

 atromaculatus) , found that it could be trained to distinguish food associated with green 

 from food associated with red of different shades; and Reighard (1908) found, in experi- 

 ments cjn feeding, that the gray snapper {Lutianus griseus) can distinguish blue as well as 

 green from red, even if the red appears much darker or much brighter than the blue. 



The fact that these animals distinguish the blue from red that is brighter, as well 

 as from red that is darker than the blue, shows, the authors maintain, that the selection 

 could not have been solely on the basis of difference of intensity or brightness such as a 

 color-blind person can perceive in the different colors, and that the animals consequently 

 have color-vision. This conclusion is valid, however, only if the brightness at the red 

 end of the spectrum is practically the same for the fishes as it is for man. If this end 

 has a lower stimulating efficiency for fishes, as is found to be true in color-blind persons, 

 it is evident that the red which, to the normal human e3'e, appeared brighter than the 

 blue may have actually appeared darker to the fishes; and if this is true the discrimina- 

 tion observed may have been made on the basis of brightness. 



This idea is in full harmony with the conclusion reached by Hess, who has probably 

 done more work on vision in animals in general than anyone else. In his experiments 

 on fishes (19-13) he studied their response in the spectrum as well as their reactions to 

 colored objects. In these experiments he tested Atherina, Phoxinus, and Mugil. Young 

 Atherinas are positive in white light of all intensities above the threshold. When they 

 are exposed in a spectrum, Hess maintains, they aggregate in the yellow near the green, 

 but that they aggregate in any other region except the red if it is made more intense 

 than the rest. This, he asserts, is true for light-adapted as well as for dark-adapted 

 specimens, provided the spectrum is sufficiently strong. He holds that they respond 

 just as color-blind organisms would be expected to respond, and, he contends, the same 

 is true in regard to the reactions of adult Atherina, Phoxinus, and Mitgil to colored 

 objects, in feeding experiments. In the feeding experiments he used food and other 

 objects colored red, yellow, green, blue, and gray of different shades, in various combi- 

 nations. He maintains that he found no evidence whatever of discrimination on the 

 basis of color. He asserts that he has demonstrated that the methods of earlier investi- 

 gators, supporting the idea of color-vision, were inadequate, and concludes that fishes, 

 contrary to other vertebrates, are color-blind. In this conclusion, however, Hess seems 

 to stand practically alone among investigators of this subject. 



Bauer (191 3), on the basis of results obtained by means of methods similar to those 

 used by Hess, concludes that fishes have color-vision when their eyes are adapted to 

 light but not when they are adapted to darkness. And Frisch is even stronger in his 

 support of color-vision. He bases his conclusion on three lines of evidence- — discrimi- 

 nation of food of different colors, change in color during the breeding season, and adapta- 

 tion in color to the bottom. 



