THE REDISCOVERED COUNTRY 351 



ma tic are ^'conceaKngly" coloured, in that sense of the term 

 — nature may use a double instead of a single barrel — 

 widely diverse optical laws may have the same affect — and 

 their differences in marking may have quite a different pur- 

 pose. Concealing coloration — again, I must repeat, in the 

 case of larger game animals — ^may prove to be a sort of by- 

 product of other purposes, a supplemental use thrown in by 

 an economical nature for good measure, an added principle 

 on the side of safety that may work at times and may fail 

 to work at others. Just as nature creates a thousand indi- 

 viduals in order that one may survive, so she may invent a 

 dozen expedients of safety in order that one may work. 

 The occasions wherein coloration fails to work — which may 

 be the majority of cases — does not necessarily stultify the 

 scheme of decoration — as they would were concealment the 

 sole or principal reason for it. And when it does work, why 

 there is so much gained. 



5. The theory of imitative coloration. This means 

 simply that the animal is so coloured as to imitate its 

 background. My remarks as to the preceding hypothesis 

 apply almost verbatim to this. Excellent examples are a 

 nighthawk on the ground, treefrogs against bark, wood- 

 cock on the nest, etc. So pretty is this theory that it, too, 

 lends itself to over-refinement. If a man seriously starts 

 to drawing analogies in the mental world, he speedily and 

 logically arrives at the absurdly fanciful; if a man sets out to 

 trace resemblances in the physical world, he arrives with 

 equal speed at the fantastic. Thus because a flock of 

 flamingoes a- wing at a distance of several miles looks to the 

 poetic eye like a pink cloud of morning — as it does — 



