558 W. H. LONGLEY 



trieved; but, obviously, each condition limits the space in which 

 the fish secured must have lain. 



Still more vivid impressions of the actual abundance of the red 

 fishes, in spite of their infrequent appearance by day, may be 

 obtained by watching them come out of hiding at dusk. If one 

 goes at that time to a suitable place, and flashes a strong light 

 down into the water, it is possible to demonstrate that they are 

 more in e\ddence than all other species combined. One may 

 easily have several in sight at once through a glass-bottomed 

 pail. Just before dawn the fishes may be seen again with the 

 same frequency in the same places, as they are about to retire 

 into seclusion. In neither case is there any possibility of undue 

 concentration of individuals being induced by the light, for the 

 first flash may reveal the situation described. 



Farther, and decisive evidence to the same effect may be 

 obtained by studying their feeding habits, which have not yet 

 received the attention they merit. As matters stand, an ex- 

 amination of the stomachs of nine specimens gives no reason for 

 supposing that they feed at all by day. The facts are stated in 

 detail in table 2. It is worthy of note that the nearest relatives 

 of these red fishes, also red, are large eyed and live at consider- 

 able' depths. Red and yellow appearing during the breeding 

 season of fishes which spawn at depths which the sun's rays of 

 those colors do not attain (Hess, 1913) may have the same 

 significance as the other reds under discussion. 



These discoveries, if it is proper to use such a term in con- 

 nection with facts which may be common knowledge among 

 fishermen, but do not seem to have been used in any attempt to 

 arrive at an understanding of the function of animal coloration, 

 lead one to oppose the application to the red fishes of hypotheses 

 of warning or immunity coloration, or signal or recognition marks. 

 But it is not really probable that any one will maintain that, 

 natural selection having failed to curb the tendency of color to 

 run riot in this group of animals of the same habit, the same color 

 has been produced independently and accidentally in each case; 

 and this is the essence of the immunity hypothesis. Nor is it 

 more likely that any one can be found who will seriously assert 



