the earliest vertebrate. Perhaps only cartilaginous neural or 

 hemal arches were present and the notochord had a thick 

 supporting sheath. Whether the agnath and gnathostome 

 lines share any common vertebral features is open to argu- 

 ment, the two neural arches and spines per segment of the 

 agnath are suggestive of the two arcualial elements per 

 segment of the gnathostome. These, however, may be paral- 

 lelisms. 



Vertebral evolution in the gnathostome appears to have 

 involved a basal radiation into chondrichthian, osteichthian, 

 and perhaps other lines. It is possible that in the osteich- 

 thian group several kinds of vertebrae were produced and 

 served to give greater support to the notochord. There is a 

 question as to whether the vertebral "spools" of some Devo- 

 nian dipnoans can be compared with the vertebral elements 

 of other choanates or for that matter whether the vertebral 

 body of the actinopterygian is comparable to that of any 

 other group. 



The evolution of the vertebral column within the tetrapods 

 is a better story but one cannot be certain which came first, 

 the embolomerous or the rhachitomous type. Occurrence of 

 the latter in the osteolepiform is not proof that it is the ances- 

 tral type. Again the fact that the stereospondylous (inter- 

 centrum largely) and reptilian (centrum largely) types of 

 vertebrae can best be derived from an embolomerous type 

 (or the proto-rhachitomous type, of Romer), with intercen- 

 trum and centrum about equally developed, does not mean 

 that this was the ancestral form. 



The primitive vertebra of the tetrapod was temnospondy- 

 lous and tended to become holospondylous, following a 

 number of pathways. The tendency among vertebrates thus 

 has been to develop a bony vertebral column and this end 



has been achieved in a variety of ways. Ways which are not 

 as yet fully understood. 



In fin structure the agnath is separated from the gnatho- 

 stome and in each of these groups are seen the beginnings 

 of lateral fins. In each group there were apparently several, 

 and sometimes parallel, pathways leading to the origin of 

 lateral fins. Among gnathostomes separate lines are more 

 evident but the pattern of evolution cannot be strictly de- 

 fined from the fossil remains. It is possible that lateral fins 

 arose either from membranes behind spines or from fin folds. 



Among the Osteichthyes there are two patterns of fin 

 structure, both of which are paralleled among the Chon- 

 drichthyes; these characterize the actinopterygians and choa- 

 nates. The origin of the tetrapod limb from the fin of the 

 latter appears to be a relatively simple step. Although these 

 two types of fins probably arose independently, the similar- 

 ities of the pectoral girdles suggest common ancestry. The 

 pectoral girdle thus becomes a strong point in the demon- 

 stration of relationship between these two kinds of fishes. A 

 rather distant relationship with the Osteichthyes is suggested 

 by the trunk, armor of Arthrodires; this reinforces the rela- 

 tionship suggested by the cranial roof patterns. 



In summary, the gap between agnath and gnathostome is 

 as well defined here as in the head. The Chondrichthyes are 

 quite distinct from the osteichthian line both in terms of 

 their vertebral column and in the fin skeleton. The actino- 

 pterygian, choanate fish, and tetrapod lines are well marked 

 but seem to share a basic ancestry. As in the case of the 

 head, the vertebral column and fin skeletons were forming 

 at the time of origin of these several lines. The gap between 

 choanate fish and amphibian is more apparent here perhaps 

 than in the head skeleton. 



192 • THE VERTEBRATE BODY SKELETON 



