24 BULLETIN OF THE BUREAU OF FISHERIES. 



ficial layers of the skin. The guanophores also undergo some changes in their relation- 

 ships with the chromatophores, by which many of them become more intimately asso- 

 ciated with melanophores, leaving relatively few intimately associated with xantho- 

 phores. Relatively few guanophores show a greenish-yellow tinge due to their prox- 

 imity with xanthine pigment. On the other hand, many guanophores show a greenish- 

 blue metallic tone, probably due to their intimate association and particular spacial 

 relationships with well-expanded melanophores. Obviously the dark shade assumed 

 by specimens adapted to a dark-blue background is due to the degree of expansion of 

 the melanophores. The greenish-blue tone is probably due largely to the optical effect 

 produced by the guanophores which are most intimately associated with melanophores. 



DISCUSSION. 



A comparative study of living material and preparations of the skin of specimens 

 of Paralichikys albiguttus adapted to backgrounds of the various shades and colors used 

 in this investigation indicates that shade depends primarily upon the degree of distribu- 

 tion of the melanin pigment in the melanophores and the spacial relationships of the 

 guanophores with these bodies in the superficial layers of the skin. The xanthophores 

 probably play no important part in the determination of shade. The melanophores in 

 the deeper layers of the skin always respond less promptly and to a lesser degree to 

 changes in the shade of the background than those in the superficial layer. 



The most obvious response to a change in the color of the background is a change 

 in the distribution of the xanthine granules in the xanthophores. Shades of yellow 

 and orange probably depend primarily upon the relative degree of expansion of the 

 xanthophores containing yellow and orange pigments. In general the particular 

 quality of the color assumed depends on a complex group of factors which do not 

 lend themselves readily to a detailed analysis. Some of the colors assumed may be 

 duplicated by mixing pigments of the colors represented in the pigments contained in 

 the chromatophores. Doubtless, these colors depend largely on the relative degree of 

 expansion of the melanophores and xanthophores. Colors which can not be duplicated 

 in this manner, doubtless, depend on the relative degree of expansion of the melano- 

 phores and xanthophores plus the optical effects due to the diffraction of light by the 

 guanin crystals in the guanophores and possibly the absorption and diffraction of light 

 by other structural elements in the skin. The optical effects produced by the guanin 

 crystals are probably modified to some extent by the particular spacial relationships of 

 the guanophores with the chromatic organs. 



Obviously, certain colors are simulated much more perfectly than others. Among 

 the colors used in this investigation yellow and green were simulated much more per- 

 fectly than dark red and dark blue. Autochrome plates published by Mast (1914) show 

 that some of his specimens simulated the color of blue backgrounds much more per- 

 fectly than the specimens used in the present investigation. However, his specimens 

 which were adapted to a red background did not simulate the color of the background 

 any more closely than those adapted to a red background in this investigation. As 

 indicated in an earlier section of this paper, none of the specimens placed on a dark-red 

 background showed any color which approximated red more closely than the orange- 

 colored pigment in the xanthophores. In view of these facts, the conclusion that all 

 colors can be reproduced in the skin of the flounder is unwarranted. 



