Il8 COMPARATIVE MORPHOLOGY OF VERTEBRATES 



reduced, persisting only as a small projection, the coracoid process, 

 anchylosed to the ventral end of the scapula, where it often forms a 

 part of the glenoid fossa. The scapula is always well developed, and 

 in the placental mammals bears a strong crest (spina scapulae) on its 

 external surface, terminating ventrally in an acromion process. The 

 clavicle varies with the freedom of motion of the limb. Thus in 

 rodents, insectivores, bats, some marsupials and the higher primates 

 it forms a strong brace between shoulder and sternum. In ungulates, 

 whales, and a few carnivores it has entirely disappeared, while in 

 other mammals it persists as a rudiment without functional value. 

 In development two small elements frequently intervene between 

 the clavicles and the sternum (fig. 62). They are preformed in 

 cartilage but eventually fuse with the sternum. Their homology 

 is very uncertain. They have been called episternaha, supra- 

 sternalia, etc. 



The Pelvic Girdle {Pelvis) 



In its broader features the pelvis {cf. fig. 119) is much like the 

 shoulder girdle, and in its full development, may be compared, part 

 by part, with the anterior arch. Thus the acetabulum or socket 

 where the appendage is attached, is comparable to the glenoid fossa. 

 Dorsal to this is the ilium in the position of the scapula, while ventral 

 and medial to the acetabulum are, on either side, an os pubis in front, 

 an ischiimi behind, with a gap (ischio-pubic fenestra) between them, 

 just as between coracoid and procoracoid. An important landmark 

 is the point of passage of the obturator nerve through the pelvis. 

 This may have its own (obturator) foramen through the pubic portion 

 or the foramen may unite with the fenestra, the condition in the mam- 

 mals, where the common opening is called the obturator foramen. 



The phylogenetic history of the pelvis is more clearly indicated 

 than is that of the pectoral girdle, for in many fossil fishes as well as in 

 the sturgeon, there is little advance over Cladoselache (p. 113). The 

 basalia of a side have fused to a single basal, often perforated for the 

 obturator nerve, and bearing the radialia on its distal surface. The 

 basalia of the two sides have not met, but there is frequently between 

 them a pair of small cartilage plates, possibly the homologues of the 

 epipubis of the tetrapoda {infra). There is no acetabular joint. In 

 the other ganoid sand in teleosts there is little advance, aside from 

 ossification of parts, while no epipubic elements occur. A noticeable 

 feature in many acanthopterygians is the forward migration of the 



