364 Transactions of the South African Philosophical Society. 



the mechanism was evolved at a still earlier stage and retained by 

 Archceopteryx, even though probably not altogether perfect for its 

 purpose ; for though the bird foot is a good foot for perching it is 

 not the best arrangement for climbing. 



The immediate ancestor of Archceopteryx and other birds is likely 

 to have been a small animal, since the power of flight is not 

 likely to have originated except in a form small enough to take 

 flying leaps, yet it seems impossible that a tibio-tarsus can have 

 arisen except in an animal of very considerable weight. Small 

 animals can hop about very satisfactorily with the ordinary 

 mammalian type of tarsus. Even the fairly large kangaroo finds 

 the mammalian arrangement with a little modification satisfactory 

 enough without any anchylosis of the bones. It therefore seems 

 probable that the avian mechanism arose in an animal which walked 

 on the ground on its hind feet, and was sufficiently heavy to require 

 a great degree of firmness in the tarsal joint. From the structure of 

 the avian fore limb we may infer that this hypothetical bird ancestor 

 had at least four digits in its manus. It probably also had a long 

 tail, since the ostrich embryo shows evidence of at least twenty-six 

 caudal vertebrae. 



To what group would this bird ancestor belong ? Clearly it must 

 have closely resembled the carnivorous dinosaurs. Its hind limb 

 was essentially similar ; like some of them it no doubt had abdo- 

 minal ribs ; and most likely it had a fixed quadrate. In accentuating 

 the differences of birds and dinosaurs a good deal has been made of 

 the fact that the bird's quadrate is movable, but this is probably a 

 secondary arrangement of very little moment. It seems very 

 doubtful if Archceopteryx had a movable quadrate ; and it is likely 

 that there is some connection between the lengthening of the bird's 

 beak and the movableness of the quadrate. As the beak lengthened 

 the cranial arches became reduced, allowing of movement in the 

 prefrontal region, and with it some movement of the quadrate. The 

 pelvis of the ostrich embryo shows that though that of the ancestor 

 was somewhat similar to that of the carnivorous dinosaur it was 

 distinctly more primitive. It had the pubis and ischium joined 

 ventrally enclosing an obturator foramen. 



The resemblances between the avian ancestor and pterosaurs 

 must have been much less than between it and carnivorous 

 dinosaurs, and yet there is pretty clear evidence of some affinity. 

 In the pterosaur there is no evidence of any bipedal stage either in 

 the digits or tarsus, and it seems probable that it is descended from 

 a quadripedal form of arboreal habits. The pelvis of the pterosaur 

 has an expanded ilium, a united pubis and ischium with a small 



