556 TEE POPULAR SCIENCE MONTHLY 



being, as in the case of the butterflies A and B, midway between the 

 extremes, favored by the momentary illumination's being neither too 

 great nor too small. Even in the climaxes of conspicuousness his 

 patterns still perpetually lessen his visibility in direct ratio to their 

 strength. And when he is chased by an enemy every instant of con- 

 fusion as to where he leaves off and the background begins must often 

 save him, so that the brighter his light marks and the deeper his dark 

 ones, the greater the range of background he can meet without sil- 

 houetting as an entirety, and being for the instant conspicuous. One 

 advantage which patterns do certainly sacrifice in purchasing the above 

 advantages is that, although their wearer is never seen entire until the 

 background is too dark for his black, or too light for his white, yet it is 

 true that, on the other hand, some note of a pied costume is always to 

 be detected moving when the wearer moves. In this respect, monotone, 

 whenever at rare moments it exactly matches its background, has the 

 advantage. This fact additionally condemns the aerial animal to detec- 

 tion when he moves, yet it is often rather his motion than his form that 

 becomes noticeable, because each of his patterns still has the chance of 

 passing for something beyond him. 



Fig. 7 shows one of the cardinal effects of patterns. C is a bird 

 patterned in white, black and gray. Seen against the sky he loses his 

 white part — against the dark he loses his dark part— and against the 

 gray, his gray part. 



Now when we find that pattern works always for concealment in 

 direct ratio to its own conspicuousness and elaboration, there remains 

 no vestige of evidence that the specific recognizability of the of course 

 constant pattern of each species has had, even to the slightest degree, a 

 hand in the evolution of such pattern. And those who would still 

 claim for conspicuous patterns any other reason for being than conceal- 

 ment of their wearer must first show what patterns could in the slightest 

 degree better serve procryptic ends, under the circumstances, than the 

 very ones now in use; and also what ones would less aid identification. 3 



There are two groups of supposed warningly-colored animals that 

 seem particularly to lend themselves to the exposure of the weakness of 



3 Naturalists confound identification with mere detection. Our identifica- 

 tion of familiar objects depends, fundamentally, upon unvaryingness of their 

 appearance. We know the mink just as well by his slim form and sleek dark 

 monochrome as the skunk by his fatter form and bushy black and white. The 

 skunk, by the way, varies in appearance far more than the mink, ranging from 

 nearly all black to half white, and this is another evidence that his pattern is 

 not for identification. Naturalists' assertion that patterns serve equally both 

 purposes is like two men claiming the same dog. The wise judge puts it to the 

 test as to which man the dog will obey against the commands of the other. 

 If we call animals' costumes the dog, we find that he always obeys Master Con- 

 cealment, but obeys Master Warning-color only when Master Concealment has. 

 commanded the same thing. 



