STRUCTURE OF THE VERTEBRATES 161 



In the first known amphibia the bones of the arm and hand 

 were specialized for land life and had assumed their typical 

 tetrapod (four footed) form. The humerus was short, the radius 

 was on the preaxial (thumb) border of the arm, and the ulna 

 was on the postaxial border. The carpals were numerous. Little 

 is known regarding the original number of digits. The oldest 

 evidence is a foot-print made by an early amphibian (Thino- 

 pus), which might indicate a foot with three toes. It is more 

 probable that the print was made by some five-toed animal 

 walking on the margin of its foot. The crossopterygian larva 

 rests on the outer (postaxial) border of its fins (page 44), and 

 the young frog also holds the foot in this position. Therefore, 

 as the earliest skeletons had the typical five toes, it is believed 

 that Thinopus was simply an amphibian which had not become 

 flat footed. 



The first reptiles had lost the dermal bones, and had the 

 "complete" girdle: (1) dorsal scapulae not connected with the 

 head, and (2) coracoids and clavicles meeting (3) the median 

 ventral sternum. The heavy coracoids held the humeri at right 

 angles to the body, with the forearms bent forward. The recent 

 reptiles are similar in structure. When quiet the body of the 

 animal rests upon the ground, but when running the legs are 

 pulled under the body as it is lifted from the ground. 



This complete girdle was carried over to the monotremes with 

 almost no changes, and is a large part of the evidence that these 

 animals arose as a separate line of evolution. The mammal- 

 like reptiles had a marked reduction of the coracoid bones, with 

 the result that the legs were drawn under the body into the run- 

 ning position, and in this respect were more like the higher mam- 

 mals than are the monotremes. The evolution of the separate 

 bones is traced below. 



1. Coracoid. The mammal-like reptiles had greatly reduced 

 coracoids which still met the sternum. The embryo of the living 

 marsupials shows a similar condition, with cartilaginous cora- 

 coids meeting in the mid-line. But as development continues 

 these cartilages recede from the sternal position and become at- 

 tached to the scapula as coracoid processes. In the placentals 

 the coracoid cartilages never meet the mid-line, and develop as 

 processes of the scapula. 



