CH. xii.] ADJUSTMENT, DIRECT AND INDIRECT. 61 



the genetic relationships between the two sub-classes. Did 

 the order of apes descend from the ape-like marsupials, the 

 monodelphian carnivora from the didelphian carnivora, the 

 higher rodents from the marsupial rodents, and so on ? If 

 so, it is difficult to see how the pouch should have been lost, 

 and the placenta developed in so many different orders 

 independently: such a number of exact coincidences seem 

 hardly probable. On the other hand, did all flie monodelphia 

 descend from one didelphian form ? If so, it is strange that 

 the differentiation into orders should have gone on so similarly 

 in the two sub-classes, resulting, for example, in the production 

 of marsupial mice which in general appearance are hardly 

 distinguishable from placental mice. 



Birds and reptiles present an equally puzzling cross- 

 relation. Upon no theory are these the direct ancestors of 

 mammals, although the lowest mammals are both bird-like 

 and reptilian in appearance. The duck-bill, belonging to the 

 mammalian sub-class of ornithodelphia, somewhat resembles 

 a lizard with a bird s beak. Embryology shows that the 

 three classes are divergent offshoots from an amphibious 

 or batrachioid ancestor ; but the birds and reptiles resemble 

 each other much more closely than either resembles the 

 mammalia, so that Prof. Huxley joins them together in the 

 super-class or province of sauroids. So far all is plain ; but 

 when we inquire by what forms the birds and reptiles are 

 linked most closely together, we are met by a difficulty. 

 Birds are divided into two sub-classes: the ostrich, cassowary, 

 emeu, dinornis, etc., are grouped together as strut/iious birds, 

 while all other existing forms belong to the sub-class of 

 carinate birds. Now until quite lately it was supposed that 

 ill birds were descended from an extinct reptilian form like 

 ~kat ancient reptile, the flying pterodactyl. For the resem- 

 tlances in structure between the pterodactyls and the carinate 

 birds are striking enough to have suggested an immediate com 

 munity of origin. Nevertheless, within the past seven years, 



E 2 



