OF EVOLUTION. 27 



scarcely be gainsaid ; and if it can be shown with 

 equal force that many of the earlier reptiles pos- 

 sessed characters belonging to birds, have we not 

 the right to assume that the two classes of 

 animals are very closely related, and that they be- 

 long to one and the same stock ? And since the 

 modern birds have practically dropped all their 

 reptilian characters, have we not the right to 

 assume further that birds are descended from rep- 

 tiles, of which they represent only a diverging 

 group ? Is it not merely a repetition of the tale 

 that is furnished by the development of the am- 

 phibian from the fish so beautifully shown, apart 

 from geological history, by the tadpole before 

 our eyes and the reptile from the amphibian ? 

 It is true that we know of no modern bird 

 which passes through an absolute reptilian stage, 

 but does not embryology tell us that one of 

 the primary structures separating birds from rep- 

 tiles, the feather, is merely a modified scale, and 

 that it originates as a true scale ? 



If the combination of the modern and ancient 



