THE COLORATION OF REEF FISHES 575 



fishes as a whole. There are 52 species, for example, listed in 

 table 3. Of these eight are included in the group under con- 

 sideration. Among the forty-four remaining there is only one 

 whose vertical range is coextensive with those above. It is 

 questionable, indeed, whether this much should be conceded, 

 for the gray snapper has been observed only twice actually break- 

 ing the surface as the others do. Each occasion was in the very 

 early morning when the water was teeming with Annelids — 

 the palolo worms^ — in their annual swarm. But even if the 

 record be accepted without qualification, the proportion in the 

 two cases stands as 40 to 2.27 per cent, or, in other words, the 

 reaction in point occurs more than seventeen times as frequently 

 among fishes of the blue series as it does among others selected 

 at random. 



There are additional observations which tend to support the 

 hypothesis that a correlation exists between the habit of swim- 

 ming well above the bottom and the development of blue pigment. 



Gnathypops aurifrons (Jordan and Thompson, '04), a species 

 known from only one specimen, taken at the Tortugas, is not 

 very common, but one sometimes chances upon a colony of 

 these interesting creatures in water 8 to 10 feet deep, or even 

 shallower. They seem to spend a large part of their day float- 

 ing inactively well up over their burrows. When undisturbed, 

 they maintain themselves at an angle of 70 to 80 degrees with 

 the horizontal for indefinite periods. Upon being alarmed they 

 retreat into their shelters tail foremost, and remain in these tiny 

 vertical pits until the disturbance has subsided. Even if un- 

 molested they seem to withdraw toward dusk, and probably 

 remain under cover all night. Hence, in spite of their fossorial 

 and tubicolous instincts, it would appear that they are pre- 

 eminently fishes of the middle depths, living in moderately 

 shallow, open water, above the bare bottom in which all but one 

 of some dozens of burrows observed were located. 



Thalassoma bifasciatus and Iridio bivittatus occur so commonly 

 together, that of them, if ever, one might say, ''Here are two 

 animals living in the same environment, but differing in color 

 and pattern, therefore, it is perfectly apparent, that if one type 



