78 AFRICAN GAME ANIMALS 



only is his elaborate theory (based purely on artificial ex- 

 periment) unsupported by one shred of evidence, but it is 

 shown to be the direct reverse of the truth by the simple 

 expedient of studying the facts in the field. Yet it is on 

 just such theories as this, backed by just such experiments, 

 that the ultraselectionists rely for support concerning much 

 of their contention about concealing coloration. If Mr. 

 Thayer in this instance were sound in his premises he would 

 have proved the exact reverse of what he announced he has 

 proved; for he would have proved that natural selection 

 had developed a revealing coloration in the animal which 

 lives in the reeds. As a matter of fact, what is really proved 

 is that the coloration of these animals has no survival value 

 one way or the other. 



Experiments of such an artificial character are usually of 

 doubtful value; carried on as Mr. Thayer has carried them 

 on they serve merely to darken understanding. No closet 

 or laboratory experiment begins to approach in value first- 

 hand field observations by a competent man. Mr. Charles 

 Sheldon's preliminary study of the white and black moun- 

 tain-sheep, in his "Wilderness of the Upper Yukon/' affords 

 a model of what such a study should be, and, when read, 

 there is no need further to explain why no amount of in- 

 genious closet guessing can take the place of trained first- 

 hand observation of animals in their native haunts under- 

 taken, not to twist facts into the support of a theory, but 

 with the ability and purpose to find out the truth. His 

 conclusion concerning the subject with which we are dealing 

 is that the colors and coloration patterns of these northern 

 sheep are not caused by natural selection of their protective 

 values — that they are not survival factors. On the whole. 



