COtOR CHANGES AND ADAPTATION IN FISHES. 1 93 



When the specimens were put into the boxes, the shade, color, aud pattern of the 

 skin were nearly alike in all. Nearly all had been in a white aquarium for some time. 

 No color, save gray of various shades, could be detected in any of them except in 

 Ancylopscita, and in these there were only some slightly irridescent patches, and at 

 times a faint trace of brownish color in and around the four conspicuous ocelli, which 

 are responsible for the common name "four spot," by which the members of this genus 

 are known. 



Immediately after they were put into the boxes they still appeared alike. There 

 was no evidence whatever that colored light reflected from the sides of the boxes or 

 transmitted through the fish from the bottom affected in any way the appearance of the 

 fish, with the exception of the fins and tail. Within half an hour, however, there was 

 a marked difference. There was much individual variation, but in general those in the 

 yellow boxes were distinctly yellowish; those in the brown, brownish; those in the red, 

 grayish or brownish; and those in the blue and green, grayish."* The pattern in the 

 skin of all the specimens remained essentially the same throughout the experiment, but 

 the shade and color continued to change greatly in all until, in the course of from a few 

 days to several weeks, the skin in most of them came to resemble the background in 

 color, as well as in shade to a most remarkable degree. This was especially true of 

 those on various shades of yellow and brown, and of those on light and dark blue and 

 chrome green.* (PL xxxiii-xxxv.) It was in general more marked in Paralichthys 

 than in Ancylopsetta. 



On red of different shades and dark green, tested in another connection, the fish 

 did not assume a color very much Uke that of the background, but in each case, including 

 the different shades of brown, red, and yellow, the color of the skin was unquestionably 

 different from that in any other case, and it showed no resemblance to anything obtained 

 on any shade or intensity of white, gray, or black. 



Thus the effect of each of the nine different colors tested, ranging from dark red 

 to dark blue, is specific, and since no combination of black and white, regardless of the 

 intensity of the light, produces anything similar in effect, it is evident that this specific 

 effect produced by the different colors can not be accounted for on the basis of differences 

 in the intensity or quantity of light reflected by the various colors, but must be due to 

 differences in the length of the waves. The fact that the light reflected from the colored 

 paint was not monochromatic can affect in no way the validity of this conclusion. Pure 

 colors, so necessary in much of the work on color-vision, are not even essential in work 

 of this sort. 



Experiments and results in detail. — These general statements and conclusions are 

 supported by the following details regarding changes in color under different con- 

 ditions, selected from my notes and by the autochromes " and photographs reproduced 

 on plates xxxiii-xxxv. 



* The tendency to become gray is particularly marked on those colors which are apparently not readily simulated. On such 

 backgrounds the animals become adapted in shade first and later in color, if at all. The response resulting in a gray shade is 

 imdoubtedly, in some way, associated with the quantity of light reflected from the backgroimd regardless of its quality, i. c., the 

 length of the waves. The response, on the other hand, which results in the production of color is associated with the quality of 

 the light. There are therefore two mechanisms involved in the process of adjustment on colored backgrounds. 



ft After this part was completed the same was found to be true for pink. 



•^ It is a pleasure to acknowledge my indebtedness to W. P. Hay for his generosity in putting at n]y command an excellent 

 rapid lens, without which the autochromes would have been practically impossible. Even with this lens the plates, all of which 

 were Lumiere. required, under most favorable conditions, an exposure of nearly a minute, and it was often necessary to exercise 

 no small amount of patience to keep the fish quiet that long. After this part was written some exposures were made in direct 

 sunlight and in this way the time necessary was greatly reduced. 



