COLOR CHANGES AND ADAPTATION IN FISHES. 211 



was repeated August 14, with essentially the same results, although according to my 

 notes there was a tendency in the fish to become darker with the light from the mirror 

 intercepted. There was little or no difference noted in the shade of the light regions, 

 but the ocelli and spots appeared to become somewhat less dense and brownish when 

 the background was illuminated from below. 



4. Background consisting of direct reflection from the mirror. — August 13, 5.30 a. m., 

 the cylinder was put in place and covered. The sheet of white paper with black. spots 

 was still under the fish. At 8.30 a. m. the three ocelli appeared black and the rest of 

 the surface very light gray. The sheet of paper was then removed, exposing the fish 

 directly to the light from the sky reflected by the mirror. At 9.15 a. m. the ocelli were 

 light grayish and the rest of the surface light reddish gray. The reddish cast was 

 probably due to the reddish sediment on the mirror. The cylinder was now removed, 

 a sheet of white paper put under the dish, and the light from the mirror intercepted. 

 At 2.45 the reddish cast had disappeared and the fish was maximum white. This shows 

 that the effect of illumination from below is in all probability due solely to the increase 

 in the amount of light from the background. 



5. Colored background. — I have in several places pointed out that when the light 

 from the mirror was colore'd the skin of the fish also became colored. On August 18, 

 4.30 p. m., a piece of brown wrapping paper was put under the dish, the cylinder cov- 

 ered and put in place, and the light from the mirror turned on. After the paper became 

 wet it was considerably lighter in shade, but it still appeared distinctly brown as seen 

 through the cylinder. August 19, 3 p. m., the fish was very light gray, probably maxi- 

 mum white, with no indication of a brownish cast. 



The brown paper was now replaced by a maroon-colored glass. As seen from the 

 opening at the upper end of the cylinder, the background appeared bright red. August 



20, 9 a. m., the fish was maximum white, except the three ocelli and a row of ten spots 

 about I mm. in diameter near the base of the fins. All of these were brown, not black 

 or gray, as they had been on the gray and the white and black backgrounds. August 



21, II a. m., no change could be detected. August 22, 8.30 a. m., it was still the same. 

 At this time the cylinder was removed and the light from the mirror intercepted. The 

 fish very quickly became much darker and assumed a distinct brownish color. So 

 rapid was the change that it could be readily detected within two minutes after the 

 illumination had been altered. No further change occurred during the following three 

 days, except an increase in the density of the shade. The fish became dark brown in 

 color, but it was not so reddish as those which had been on a red background for several 

 weeks. These tests were repeated several times during the following five days with 

 essentially the same results. The brownish ocelli and spots mentioned never failed to 

 appear. These were not seen in any of the tests with a gray background illuminated 

 only from below, no matter how dense it was. 



It is therefore evident that the appearance of the brownish ocelli and spots on the 

 red background could not have been due to the quantity of light transmitted by the 

 colored glass, and that it must have been due to the quality of the light — i. e., the 

 length of the waves. But, under normal conditions on a maroon background, with 

 certainly no more colored light striking the eyes from below and none received from 

 above, just as in these tests, the effect of the color is very much greater. The entire 



