210 THE ORIGIN AND DEVELOPMENT 



ferent from those we meet in discussing the orna- 

 mentation of vertebrated animals. In birds and 

 quadrupeds, the feathers, hair, and other dermal 

 appendages have developed to maturity and even 

 wasted and been replenished under all the vicissi- 

 tudes to which animal life is exposed during a 

 period of several years. In butterflies, on the 

 contrary, the ornamentation we are considering 

 is confined to the brief final epoch of life, there is 

 no replenishing of the scale-tissue, and the scales 

 are formed rapidly and once for all, at a definite 

 period, viz., immediately upon the change from 

 larva to pupa ; and being then concealed from 

 light and excesses of temperature within a thick 

 integument, and often also behind the walls of a 

 dark chamber of silk, vegetable fibres, or earth, 

 they are as far removed as possible from external 

 agencies. In the depths of this retreat the scales, in- 

 cluding all the pigment of the wings, are complete- 

 ly developed, the insect appearing full-fledged and 

 perfectly caparisoned, subject to no further change. 

 Nevertheless, the general phenomena of orna- 

 mentation in vertebrates are so exactly repeated 

 in butterflies that no one can plausibly claim that 

 these phenomena originate, in the two groups, in 

 distinct proximate causes. The same relation of 

 color to locality which has been so well brought 

 out by Mr. Allen* reappears in butterflies, as I 



* In various papers upon our native birds and mammals. 



