590 W. H. LONGLEY 



impressed by the uniform absence of effort to demonstrate that 

 that conspicuousness exists, whose occurrence it is undertaken 

 to explain. 



Wallace's ('91, p. 228) statement, that to color for recognition 

 we owe most of the variety and much of the beauty of the colors 

 of animals; that it has caused at once bilateral symmetry and 

 general permanence of type, and that its range of action has 

 been perhaps equally extensive with that of coloration for con- 

 cealment, ultimately rests upon what "a, little consideration ^^'ill 

 show." Reighard has experimental evidence of the insufficiency 

 of the hypothesis of warning coloration as applied to certain 

 tropical fishes, but his belief in immunity colors is based upon 

 the alleged fact that "even casual observation shows that they are 

 highly conspicuous.^' Roosevelt ('11, p. 176) also declares that 

 "the most elementary study of prongbucks amid their natural 

 surroundings shows that there is not even the smallest founda- 

 tion for Mr. Thayer's theory." 



That the conspicuousness so lightly assumed is a subjective 

 phenomenon is capable of demonstration. Bristol ('03) has 

 expressed his opinion of seven or more Bermudian species, all 

 of which occur at the Tortugas, although only three are included 

 in table 5. Chaetodon capistratus he considers protectively 

 colored. Midway between such extremes as the squirrel fish 

 and the hind he places the angel fish and the tangs (Teuthis spp.), 

 yet to Reighard all three seem highly conspicuous. In addition 

 Bristol's distribution of his material points the same moral as 

 what has gone before. He recognizes, first, types of warning 

 coloration with bright hues, simple patterns and little ability 

 to change their appearance; second, protectively colored species, 

 whose scale of coloration is not so high, whose pattern is complex, 

 and whose power to change color is great. But to an inter- 

 mediate position he assigns others whose "color is medium, 

 pattern not complex, and range of color change less than in the 

 second group." Again, of the seven or more species which he 

 mentions the tangs, for example, have very great power of color 

 change, but their pattern is as simple as that of his warningly 

 colored forms. Neither, I think, is the blue parrot placed in the 



