206 BULLETIN OF THE BUREAU OF FISHERIES 



animal is independent of the intensity of the light. How is it possible to explain this 

 phenomenon ? 



LIGHT REFLECTED FROM THE SKIN AS A FACTOR IN ADAPTATION. 



Sumner says (191 1, p. 463) that to account for adaptation in shade "we are limited 

 to two alternative explanations: Either (i) the fish takes into account the degree of illu- 

 mination, just as we do, and makes due allowance for this in judging of the paleness or 

 darkness of the background; or (2) i: makes a direct visual comparison between its own 

 surface and that of the background and endeavors to bring the former into harmony 

 with the latter." 



To test the second hypothesis he made numerous observations on specimens, some 

 of which had an opaque cloth fastened over them so as to conceal all but the eyes, and 

 others had the pigmented surface stained. The results of all of these experiments he 

 considers inconclusive owing to the possible effect of the treatment. He holds, however, 

 that adaptation occurs normally in specimens bedded in sand, and concludes that this 

 comes "very near to refuting the visual hypothesis altogether." 



This would no doubt be true if all of the skin within the range of vision were con- 

 cealed when the animals are bedded, but this does not appear to be the case. While I 

 have not seen Rhomboidichthys, the form used by Sumner in his experiments, I have 

 made numerous observations on Paralichthys, a similar form, and I have never found 

 one, except for very short periods, in which the skin along the margin of the mouth was 

 not exposed. Moreover, the skin bordering the lower margin of the eye is also usually 

 exposed. Since both of these regions are within the range of vision, the fact that adap- 

 tation occurs in flounders after they are bedded does not seem to warrant Sumner's 

 conclusion. 



By removing the eye situated near the mouth and by taking special precautions to 

 keep the lower margin of the other eye well covered, I was able to correct the defects 

 in Sumner's method, and I found that after thus totally eliminating all of the skin from 

 the view of the fish, it still simulated the background, and, moreover, simulation was 

 quite as rapid and as extensive as it had been before the skin was covered. This was 

 observed in several different experiments, one of which follows: 



On September 3 the ventral eye of a specimen of P. albiguttus 27 cm. long was 

 entirely removed. On September 12 the wound was thoroughly healed and the speci- 

 men fed and acted normally in every way. It was put into a black aquarium, and 

 after it had become maximum black it was, at 9.40 a. m., suddenly entirely buried 

 with gray sand. The fish soon shook enough sand off to admit of respiration, but the 

 eye remained covered. At this time the sand was washed from the posterior portion 

 of the animal, so as to admit of direct observation of the skin. This was found to be 

 still maximum black, showing that no change occurs when the eye is entirely covered. 

 Soon after this the fish moved slightly and the eye was raised sufficiently to be seen, 

 but it was still covered with a thin layer of sand, and the upper surface was still somewhat 

 below the surface of the sand. The skin, however, began to turn lighter, and half an 

 hour later, 10.40 a. m., it was clearly somewhat Ughter. At 11.30 the shade of the skin 

 was considerably lighter and the pattern much broken so as to simulate the sand fairly 

 accurately. During this entire time the eye projected only very little above the surface 

 of the sand. The lower margin of the eye was continuously well covered with sand, as 

 was also all of the rest of the skin within the range of vision. It is consequently evident 



