is the matter of pattern perspective. The elucidation of this will mark another 

 step in the differentiation of obliterative coloration from all forms of mimicry. 

 For it will show that not an exact reproduction of the actual background- 

 pattern, but a picture of that pattern as it looks when more or less altered and 

 refined by distance, is essential to the concealing of an object. Or, in other 

 words, that the object's obliteratively-shaded surface must bear a picture of 

 such background as would be seen through it if it were transparent. The 

 diagram, Fig. 18, represents a flat, bird-shaped model, vertically placed, 

 seen against a horizontal background. The background-pattern is supposed 

 to be actually uniform throughout, but diminished to the eye as it recedes on 

 the horizontal plane. The model, vertically interposed between the eye and 

 this receding ground-plane, must, for concealment, bear a pattern graduated 

 from larger on its lower borders to smaller on its upper. For the highest 

 parts of the model are seen against the most distant and therefore most dimin- 

 ished portion of the uniform background-pattern — and vice versa. Further- 

 more, the markings of the background, being on a receding plane, are fore- 

 shortened throughout, and this effect also must be imitated on the model. 



These diagrams and photographs will serve to illustrate, in a crudely sim- 

 plified form, some of the main principles of obliterative pattern which prevail 

 in Nature. Instead of one unvaried pattern on a single plane, however, Nature 

 furnishes backgrounds of rich diversity. Mud, grasses, pebbles, bushes, tree- 

 trunks, branches, leaves, living and dead, and vistas amid vegetation to the 

 bright sky beyond — these, all of them subject to endless variations of com- 

 minglement, of distance, and of lighting, are a few of the numberless details 

 of the backgrounds against which ground-haunting animals are seen. To 

 achieve the highest degree of inconspicuousness, these animals must wear, 

 superadded to their obliterative shading, yet in the main conforming with it, a 

 sort of compound picture of their normal backgrounds — a picture seemingly 

 made up by the averaging of innumerable landscapes. Further, this land- 

 scape-picturing must be suited variously to different portions of the animal's 

 surface. The top-planes, being seen in full only against the nearer ground, 



3i 



