XV THE RELATION OF FACTS TO THEORIES 323 



man will," the " man " in question being apparently 

 the all-compelling force of Natural Selection. 



The facts so stated are certainly sufficiently 

 remarkable and seem at first sight at least to 

 warrant Weismann's conclusion that they are only 

 explicable by the action of Natural Selection, but 

 detailed reflection shows many difficulties. The 

 prime assertion of the immunity of the Danaidae, etc. 

 is denied by many, the persecution of the pro- 

 tectively coloured butterflies is also, as we have seen, 

 doubted by field entomologists. The relation be- 

 tween the colour of the wings and the position taken 

 up by them in repose seems a very striking fact, 

 but I have noticed a somewhat similar occurrence in 

 the feathers of birds. In the humming-bird Cyno- 

 lesbia gorgo the tail is forked and the tail-feathers 

 overlap one another ; the tips of the feathers are of a 

 gorgeous metallic colour, but this is confined to a 

 simple band at the exposed part, the part of the 

 feather which is overlapped being a deep black. 

 Owing to the forking of the tail, the overlapping is 

 such that in each quill more of the vane is covered 

 on one side than on the other, the distribution of 

 black and metallic colour corresponds exactly to 

 this overlapping, so that the metallic colour extends 

 farther back on one side of the rachis than on the 

 other. Here is a case very like that of the butter- 

 flies' wings, and yet it is almost impossible to believe 

 that it can have been produced and maintained by 

 its utility (see also p. 167). 



We have confined our study of protective colora- 

 tion and mimicry to the Lepidoptera, because they 

 are admitted on all hands to exhibit the phenomena 



