THE COLORATION OF REEF FISHES 551 



which, even under the most favorable conditions, the least 

 intelligible results are obtained. Therefore, when all is con- 

 sidered, the surprising thing is not that the significance of color 

 change should have been imperfectly understood, but that its 

 meaning should have been grasped at all. 



Except as banded patterns sometimes appear upon fishes as 

 they come to rest, unconfined specimens give one little reason to 

 believe that psychic states, special activities, or uncontrolled 

 internal stimuli determine their color reactions to any appreci- 

 able degree. The following records are rather exceptional. A 

 hogfish (L. maximus) feeding in the gray phase over light colored 

 bottom is noted as flushing momentarily and then returning to 

 and remaining in its previous condition. Similarly a Nassau 

 grouper (E. striatus) lying in a dark phase in the shadow of a 

 large coral head came into the open over clear sand (fig. 3), 

 turned pale as it crossed it (fig. 4), and darkened somew^hat (fig. 

 5) as it commenced to feed upon the half of a crawfish (Panu- 

 lirus argus) which had been provided for it. Such irregularities 

 in behavior seem inconsistent with the impression conveyed 

 in the preceding pages, but they are apparently mere swdrls 

 upon the great current of evidence, and move on with the tide 

 they seem to oppose. 



Flounders adapted to black or white bottom will reverse 

 their coloration completely, while the whole body rests upon 

 the color to which it conforms, if the head is upon the other 

 (Mast, ' 14) . Sparisoma flavescens reacted for me in the same way 

 in a tank with a slate bottom, half of which was covered with 

 white sand. It was observed repeatedly swimming slowly 

 toward the white, where for a moment its passage was impeded* 

 at the border. Then, with only a part of its head across the 

 line, it assumed in full the color and pattern it regularly showed 

 over the light material before it. The whole system of chromato- 

 phores in the most changeable species seems to be in a state of 

 highly unstable equilibrium. Hence, dominance for even an 

 instant by a dark object in which interest might be centered 

 for the moment might very well induce the exact aberrations 

 which have been noted in the preceding paragraph. 



