21 8 BULLETIN OF THE BUREAU OP FISHERIES. 



Table IV. — Effect of Adaptation to a Given Color on the Selection of Colors. 



This was true, however, only for those specimens adapted to pink, and the fact 

 that these specimens were adapted to pink (Ridgway's alizarin pink) and that the test 

 boxes were still the original red (Bradley's red, shade no. 2) may have had something 

 to do with it. The preponderance of turning toward the yellow in individuals adapted 

 to yellow was also insignifiant. But in those adapted to green and blue the selection 

 of these colors, respectively, was quite marked, especially the latter. In all the selec- 

 tion of white was practically perfect. 



The order of brightness of these four colors, beginning with the darkest, was as 

 follows: Red, green, blue, yellow. Between the red and green and especially between 

 the green and the pink, which was considerably lighter than the red, the difference 

 was not great, but it was marked in reference to the other three colors. 



The fact that the individuals adapted to green selected green in preference to red, 

 yellow, or blue, seems to indicate that the selection was not made on the basis of bright- 

 ness, for the green was lighter than the red and darker than the yellow and the blue. 

 The same argument applies to the reaction of those adapted to blue. 



The only way that these reactions could be accounted for, solely on the basis of 

 brightness, would be to assume that the animals were adapted to a color of a given 

 degree of brightness, and they reacted negatively to colors either darker or brighter than 

 the one to which they were adapted. This assumption, however, does not accord well 

 with the results obtained in the black and white box ; for it demands that the brighter the 

 color to which the animal is adapted the greater the tendency to turn toward the white. 

 Thus one would expect the greatest relative number of positive reactions to white in 

 individuals adapted to yellow and the least in those adapted to red. This expectation, 

 however, was not realized. These results are not in themselves absolutely conclusive 

 regarding color-vision, but they do seem to lend support to the conclusion reached with 

 reference to the problem in the study of the simulation of colors in the background. 



Hess (1913b) holds that the spectrum for certain fishes is shortened at the red end 

 and that the region of maximum brightness is shifted toward the blue. That is, he main- 

 tains that the distribution of brightness for these animals is similar to that for color- 

 blind persons. But even if this is true for flounders, the reactions recorded in table iv 



