THE COLORATION OF REEF FISHES 595 



Thayer's second thesis is: 



Colors, patterns, and appendages are the most perfect imaginable 

 effacers under the very circumstances wherein such effacement would 

 most serve the wearer. . . . Patterns on animals' coats are the 

 utmost that nature can do in opposition to the potent vicissitudes of 

 silhouetting. , . . Their bold coloring minimizes, not increases 

 their conspicuousness. 



The optical principles involved in this series of propositions 

 are explained with precision, and their necessary effects demon- 

 strated by a multitude of devices of surpassing ingenuity. Yet, 

 when all is done, the most friendly critic, who has not undertaken 

 supplementary researches, hesitates to believe, for it is per- 

 fectly clear that, if Thayer's reasoning is correct, the colors of 

 animals should repeat those of their environment, and this is 

 not known to be so. Great numbers of positive illustrations 

 are offered, to be sure, but there are greater hosts of animals 

 to draw upon ; it is not difficult to find among others explanations 

 that seem fanciful, and one animated by nothing more than a 

 spirit of scientific doubt may suspend judgment, rather than 

 become the victim of a plausible hypothesis supported by selected 

 and unrepresentative material. 



It is at this point that Thayer's method reveals its intrinsic 

 weakness. If he had been able to supplement his demonstration 

 of the effect of countershading with proof rather than suggestion 

 that color is intimately correlated with habit, his theories would 

 probably have displaced those that preceded them, as soon as 

 they could be disseminated among zoologists. As it is, when an 

 explanation of sexual dimorphism among birds is required, he 

 falls back upon difference in the nesting habits of the sexes, 

 which is certainly permissible when these are known to be differ- 

 ent; but Dewar and Finn ('09) assert that there is dimorphic 

 coloration is species in which the male shares the burden of 

 incubation. Now and hereafter the existence of other differences 

 in habit may be inferred tentatively, but this has not been legiti- 

 mate heretofore, since it would have involved the support of one 

 hypothesis by another for which no shadow' of justification ap- 

 peared. 



THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 23, NO. 3 



