THIRTEENTH ANNUAL REPORT 95 



One of the lai\i;e llcior pools eontains twelve specimens of the 

 striped bass {Rocciis lijicatus) which have never been moved 

 since they were first placed in the pool fifteen years ago. This 

 jionl is lined with while tiles, and all the specimens are pale and 

 colorless while their characteristic stripes are scarcely noticeable. 

 The only normall}' colored fish in the lot is blind, and it is also 

 luidersized, since it has never been able to secnre its full share of 

 the daily supply of food. 



.\nother wdiite-tiled pool cnntains skates and rays, all of 

 which became pale and colorless within two days of their cap- 

 ture, and ceased to exhibit their original rich brown tones. A 

 newly arrived specimen is remarkably cons])icuous when placed 

 among- them. The changes of color depend upon the eye. 



In all these northern fishes, which reflect the tone of their sur- 

 roundings, the changes are slow, requiring days or even Aveeks, 

 but the color changes of tropical species are sadden, or actually 

 insfantaiicoiis. 



A year ago the following label was attached to all tanks con- 

 taining fishes subject to c[uick changes in appearance: 



"This species ;;(«_v change color at any moment. 



A few minutes' observation of the fish is usually enough to reveal a 

 cliange either in color or in pattern of marking. 



The color cells of the inner skin are under the instant control of the fish. 



Under natural conditions the changes of color are made chiefly for the 

 purpose of concealment from enemies. They are also used for the capture 

 of prey, for signaling, warning, mimicry, courtship and other purposes." 



Through the agency of the label many of the visitors to the 

 Aquarium learned to detect the changes, and found a new en- 

 joyment in watching the more showy species. 



The methods by which fishes produce changes in appearance 

 are well understood. The color cells in the skin are contractile, 

 and, as they may contain red, blue, yellow, or other pigment, the 

 diiiferent colors result from muscular action u])on one or more 

 kinds of cells. 



The fishes of tropical seas are usually of brilliant coloration, 

 perhaps more so than tropical birds or butterflies, but their re- 

 markable colors are not so well known, as they jicrsist only dur- 

 ing life. Outside of the tropics these fishes are to be seen no- 

 where, except in a few public aquariums. 



Six months ago I began a series of daily observations on their 

 colors, which have been continued to the present time. While 

 the present article is merely a preliminary account of the be- 

 havior of certain tropical fishes with respect to the display of 

 color, enough has been learned to show that many of those under 



