556 W. H. LONGLEY 



appearing in flecks and vermiculations upon closer scrutiny. 

 The method is crude; allowance for the personal equation of the 

 observer must be large, and the number of species examined was 

 neither great nor thoroughly representative. Moreover, as 

 will appear later, the implicit assumption that all hues that accord 

 with those of the environment must repeat those of the bottom 

 is without foundation. The results obtained are not, however, 

 without interest, and may be stated as follows: 



Upon 30 species considered, yellow occurred nineteen times; 

 brown and gray, sixteen times each; blue, eight times; red and 

 green, five times each; and black, three times. 



These facts might seem to warrant the conclusion that, roughly 

 speaking, colors occur upon the fishes in the same proportions 

 as those in which they appear upon the reefs ; for the gray of bare 

 sand and dead corals, and the brown of large algae, or the micro- 

 scopic ones living symbiotically in some corals, and the brown 

 gorgonians, are the commonest colors from the shore line to 

 depths where the bottom becomes invisible. Turtle grass is 

 abundant over some parts of the reef flats, and at places upon 

 the reefs its color is supplemented by the vivid green of Zoo- 

 anthus or Halimeda. Yellow heads of Porites astraeoides and 

 yellow gorgonians are common upon the rougher bottoms where 

 fishes most abound, and red and blue are not wholly absent, 

 though forming only an infinitesimal part of the color mass as 

 a whole. But the conservative inference that, at least, no 

 positive evidence appears that different laws prevail in the 

 distribution of color upon animals in the Sargassum and on 

 the reef, is all that is permissible; for it may be shown that 

 some greens, most blues and apparently all true reds are not 

 related at all to the colors upon the bottom, but have another 

 significance. 



It was noted next that the remarkably large eyes of the 

 squirrel fish, Holocentrus ascensionis, and of Priacanthus cruen- 

 tatus are correlated in each case with red body color. This 

 observation, and the knowledge that the upper hmit of the range 

 of certain red, deep-water animals is also the limit within which 

 most of the sun's rays are absorbed (Murray and Hjort, '12, p. 



