ADJUSTMENT, DIRECT AND INDIRECT 



marine environment has aided in producing 

 fish-like limbs, while it has not interfered with 

 the general likeness of the animal to certain 

 families of land carnivora. So in the case of 

 the pterodactyl as compared with carinate birds, 

 we begin with skeletons constructed on the 

 same plan, and we may expect to find that di- 

 rect adaptation to the necessities "of flight will 

 tend to produce similar modifications of the 

 shoulder structure. But since, before the ap- 

 pearance of pterodactyls, the dermal covering 

 of reptiles was very likely as different from that 

 of birds as it is now, so that a reptilian wing 

 could not be formed by a modification of the 

 dermal covering, we find, naturally enough, the 

 wing of the pterodactyl formed, like that of 

 the bat, by a modification of the skeleton. And 

 this fact seems to justify us in the alternative 

 which we have accepted, that the likeness of 

 the pterodactyl to birds is no proof of imme- 

 diate kinship, but only of secondary adaptive 

 variation, as in the case of bats. A similar 

 argument applies to the numerous likenesses 

 between the higher mammals and the marsu- 

 pials. At an ancient epoch the marsupials were 

 a dominant race of animals, extending all over 

 the world. But since they have been almost 

 everywhere exterminated by their hardier mo- 

 nodelphian descendants, there is no difficulty in 

 the view that direct adaptation to similar differ- 



8s 



