596 W. H. LONGLEY 



Similarly, there has been no conclusive reply to Roosevelt's 

 ('11) objection — which is not original, but is stated recklessly 

 by him in its most uncompromising form: 



The claim that a large number of birds each coloured entirely unlike 

 the others both in tint and pattern and all hving in substantially the 

 same surroundings are all concealingly coloured can never be tenable. 

 . . , . Here near my house, for example, there is a hedge or tangle 

 rmming on one side of the garden. ... In this are to be found 

 thrashers, towhees, catbirds, chats, indigo buntings, and Maryland 

 yellowthroats (together, of course, with other birds). The fact is, 

 not one of these birds really has a concealing coloration. 



Unprejudiced inquiry has exposed the fallacy involved in 

 these remarks, but Thayer's hypothesis will require the labor 

 of many professional naturalists, before it assumes the position 

 to which it seems justly entitled. As a contribution to its sound- 

 er estabhshment it is a pleasure to state that the coloration of 

 tropical reef-fishes is correlated with their habits, repeats in 

 general the color notes of their environment, and is obliterative 

 under natural conditions. 



SUMMARY 



Fishes are countershaded ; color changes, which are common 

 even among the most gaudy, tend to assimilate them with their 

 environment; and, in general, their colors repeat those of their 

 surroundings. Specially defended types are not unlike others 

 in pigmentation, nor inferior to them in their ability to effect 

 adaptive color adjustments. Finally, there is no evidence that 

 bright colored species enjoy greater immunity from attack than 

 their fellows, for they constitute a large proportion of the food 

 and may be readily identified in the stomach contents of pre- 

 daceous forms. 



These statements which rest upon a great body of verifiable 

 observations, are consistent with the Darwinian hypothesis, but 

 inconsistent with the assumption that animals of high color 

 possess more than minimal conspicuousness under natural con- 

 ditions. They impel one to reject the hypotheses of warning 

 and immunity coloration, signal and recognition marks, and 



