192 BULLETIN OF THE BUREAU OP FISHERIES. 



COLORED BACKGROUND, ARTIFICIAL. 



General statements. — Among those who hold that fishes simulate the background in 

 color, Frisch (1912, 1913) has probably presented the most conclusive evidence. He 

 made an extensive study of the reactions to color in Phoxinus and Crenilabrus and 

 maintains that there is in these forms clearly adaptation to color. (Farbenanpassung.) 

 Hess, however (191 3), while admitting that fishes change color, maintains that there is 

 no evidence showing that these changes bear any specific relation to the color of the 

 environment. He asserts that the strongest point Frisch has in favor of his contention 

 is found in his statement that Phoxinus on a yellow bottom becomes yellow, but, he 

 concludes, after working on this form one and one-half years, that Frisch is in error. 

 He kept specimens over colored bottoms for several weeks and found no specific effect 

 of the color of the bottom on that assumed by the fish. He says (p. 407) : " I did indeed 

 find that after having been over red, yellow, or orange bottoms for a number of hours 

 some of the specimens appeared slightly more yellow than those kept over gray bottoms, 

 but often the opposite was true." And those on blue, he maintains, frequently appeared 

 more yellow than those on red. Hess asserts that his opinion regarding the color of 

 these specimens was confirmed by his colleagues, and he says that one of them even 

 maintained that those on the gray were more yellow than those on the yellow bottom. 



After the appearance of the criticism of Hess, Frisch repeated some of his experi- 

 ments. He tested 22 specimens of Phoxinus on yellow, and asserts that 20 became 

 yellowish and 2 did not. One of these, he says, was later found to be sick and the other 

 abnormal, in that it contained only a very small amount of yellow pigment. Frisch 

 states that his judgment concerning the color of these fish was confirmed at the time 

 by his colleagues, Hertwig, Goldschmidt, and Buchner. 



While the results recorded in table i include neither Phoxinus nor Crenilabrus, they 

 indicate clearly, as previously stated, that adaptive color changes do occur in some 

 fishes. A thorough study of Paralichthys and Ancylopsetta, on artificial backgrounds 

 variously colored, proves that this conclusion is correct and shows that color-adapta- 

 tion in these forms, especially in Paralichthys, is highly developed and extends over a 

 wide range. 



In this study nine wooden boxes, 30 cm. long, 25 cm. wide, and 7.5 cm. deep, were 

 used. Each box was colored on the inside with oil paint ranging from dark red to very 

 dark blue. There were thus nine boxes in all, each difi'ering in color from the rest. 

 The nine different colors represented in these boxes were compared with Bradley's 

 standard colors. The common name of each and that in Bradley's classification with 

 which it most nearly corresponds follows: Maroon (red, shade no. 2), vermilion (orange 

 red), dark brown (orange, shade no. 2), light brown (orange, shade no. i), chrome yellow 

 (orange yellow), light yellow (yellow, tint no. 2), chrome green (yellow green, shade no. 2), 

 light blue (green blue, tint no. i), dark blue (blue, shade no. 2). 



These boxes were all placed in strong diffused sunlight and so arranged that the 

 water in them was constantly about 4 cm. deep. All of them received a constant supply 

 of sea water from the storage tank. One or more moderately small specimens of Para- 

 lichthys were put into each box, and one of Ancylopsetta into the maroon, the chrome 

 yellow, the chrome green, and the dark blue. Under these conditions the fish fed 

 normally and thrived, some being in the boxes more than two months. None were lost 

 except by accident. 



