° 1912 " J Allen, The Concealing Coloration Question. 499 



of forest cover is such as to make the detection of an animal of 

 any color difficult so long as it remains motionless. (It will be 

 observed that Roosevelt in this gives up the claim that animals 

 under such conditions are actually conspicuous. He admits that 

 they are not, but holds that their coloration has no significance 

 under such circumstances and that therefore it could not have 

 been brought about through natural selection.) One obvious an- 

 swer is a statement of the well-known fact that the birds of the 

 treetops run quite largely to the l^rilliant colors, while sea-birds, for 

 instance, show an entirely different style of coloration, largely 

 white and gray and black. As before stated, it is difficult to 

 account satisfactorily for such differences without involving the 

 theory of natural selection. Certainly sexual selection will not 

 explain them, and without one or the other of these methods of 

 selection we should have to fall back upon the vague and unsatis- 

 factory theories of 'harmony in nature' which, if they have any 

 definiteness at all, are really more metaphysical than scientific. 

 But wholly aside from probabilities and theory, what are the facts 

 in the case? Is it true that a bird of any color would be incon- 

 spicuous in the forest and that no coloration scheme can make 

 any difference in its conspicuousness? It seems to me that the 

 obvious way to settle this question is by experiment; observation 

 under purely natural conditions being unavailable in this case. 

 Distribute a number of bird -skins, forest birds and sea-birds, im- 

 partially in the treetops in some thick wood and see whether there 

 actually is any difference in their conspicuousness or not. It 

 ought not to be a difficult experiment. I am not aware that Mr. 

 Thayer has ever tried it in any of his demonstrations. I hope he 

 will, and I hope that others will. 



This suggestion of Roosevelt's, — or rather this pronouncement, 

 for of course he does not offer it as a mere suggestion, though we 

 can accept it as such, — this suggestion that all colors, bright or 

 dull, may be inconspicuous in a forest landscape is to my mind 

 the most important point he makes. Except for this one point, 

 indeed, it seems to me that the value of his paper depends almost 

 entirely on whether we can accept his interpretation of his own 



' Op. cit., p. 214 (6). 



