272 VERTEBRATE ZOOLOGY 



tendencies in the direction of avian traits. The pterosaurs might seem 

 at first thought to be the ideal group from which to derive the birds, 

 but unfortunately these highly specialized flying reptiles are anatom- 

 ically too different from birds to offer any hope of using them as a 

 connection between the birds and the reptiles. The pterosaurs have 

 arrived at their flying mechanism in an entirely different fashion. 



The bipedal dinosaurs have been chosen by some authorities as the 

 group offering the strongest suggestion of avian affinities. It is argued 

 that the birds took their origin from some rather generalized offshoot 

 of the Triassic bipedal dinosaurs, which developed flight, and, after 

 a long period of transition, gave rise to Archceopteryx and other 

 primitive birds. This is possibly the best clue as to the pro-avian 

 ancestry of birds, but this is at best far from a satisfying phytogeny. 

 In lieu of a definite ancestral group of reptiles from which to derive 

 the birds the problem has become somewhat less concrete and concerns 

 itself with an attempt to explain the origin of the flying habit. Three 

 distinct theories of the origin of flight are held at the present time: 

 that of the cursorial, that of the arboreal, and that of the diving 

 origin of flight. 



The theory of the cursorial origin of flight was advanced by 

 Nopcsa, a Hungarian palaeontologist. This author considers that there 



FIG. 147. Restoration of a hypothetical pro-avis, supposed cursorial ancestor 

 of birds. (From Lull, after Nopcsa.) 



is a very fundamental distinction between flight based on membranous 

 planes, like those in bats and pterodactyls, and planes made up of 

 feathers; for the former involves marked adaptations of the hind 

 limbs, whereas the latter involves only the fore li^bs and leaves 

 the hind limbs unchanged. The hind limbs of birds are essentially 



