238 Fortieth Annual Meeting 



experimental purposes, and during the latter part of my 

 stay the supply gave out completely, so that certain im- 

 portant tests were left untouched. 



When just brought into the laboratory by the collectors 

 the fishes were usually of a rather dark brown color, with 

 inconspicuous darker and lighter markings. This appear- 

 ance makes it seem probable that they were commonly taken 

 upon the mixed sand, composed of finely divided lava and 

 tufa which is so common in the bay of Naples. 



In my experiments, I began with natural backgrounds. 

 I used the dark mixed sand just referred to ; a fine gravel, 

 composed of gray and brown pebbles of various shades; 

 and again a coarse gravel, made up of stones an inch or 

 more in diameter. I also used several other sorts of ma- 

 terial, including a jet-black sand of almost pure magnetite 

 crystals, and a white artificial product, obtained by grind- 

 ing up marble. Bottoms of slate and of white opaque glass 

 were employed in some of my experiments; and finally 

 plates of glass which had been painted with various colors 

 and with black and white, distributed in various patterns. 

 It is with these last that I obtained some of the most pic- 

 turesque results. 



What these results were can best be told by reference to 

 the lantern slide pictures, which I am going to show you, 

 but it may be well at this point to sum up the leading facts 

 which were revealed during my experiments. 



( 1 ) Fishes became nearly white upon a white back- 

 ground ; dark brown and in some cases nearly black upon a 

 black one. 



(2) The animals were limited, however, to black, 

 white, brown and gray tones.* Bright red or yellow back- 



* More radical changes of color have, however, been described for 

 certain other fishes, as, for example, one cyprinoid species, which has 

 been found to vary from light green to dark brown, depending upon the 

 condition of the bottom on which it lives ; and our own cunner, when 

 young, may be bright green, pink or of less showy colors. I do not 

 think, however, that a transition from one color to the other has ever 

 been observed in the latter case as a direct result of a change in the 

 nature of the background, and the supposition that the same fish can 

 undergo this change is only an inference. 



