AND CLASSIFICATION OF THE FOSSIL REPTILIA. 
249 
The Verteljral Column. 
The only part of the vertebral column in which the number of vertebra) is imper¬ 
fectly known is the dorsal region. A specimen in the British Museum, of which a 
portion has been figured by Sir B. Owen (‘Cat. South African Reptilia,’ Plate 52), 
indicates not fewer than fifteen vertebrfB contained between the head of the humerus 
and the head of the femur, and in Deuterosaurus the number of dorsal vertebrae 
preserved is eleven, so that the vertebrae in the several regions of the body may be 
stated with some probability at seven or eight cervical, twelve or thirteen dorsal, one 
to six sacral, and about twenty caudal. 
The forms of the vertebrae appear to differ a little in the several types, but they all 
show a remarkable approximation to Sauropterygian genera. There are the same 
biconcave articular surfaces to the centrum, only the mode of ossification of the inter¬ 
vertebral or proto-vertebral substance appears to be different, for the unossified part has 
a tendency in Anomodonts to contract to a tubular-conical form, while in Plesiosaurs 
and Nothosaurs the tendency is for the base of the cone to disappear, so that the 
conical excavation becomes shallower, till the articular surface is perfectly flat, and 
the cavity becomes obliterated. 
The positions of the articulating surfiices for the ribs are more like that in Plesiosaurs 
than in other animals, since there is a double attachment in the cervical region, a 
single attachment in the dorsal region, which is entirely on the neural arch (except in 
the Pareiasauria), and a single caudal attachment, which descends again to the 
centrum. But, since the cervical diapophysis in Anomodonts is formed by the neural 
arch, the affinity is obviously closer with the Crocodilia, especially the Teleosauria, 
and in that group chevron bones are equally well developed ; so that, notwithstanding 
the general resemblance of these vertebrm to those of Plesiosaurs, the divergence in 
the cervical region is as absolute as in the sacral region. Yet the sacral vertebrae are 
quite unlike those of Crocodiles ; and the affinities indicated by these resemblances 
may be no more important than the affinities with Ilatteria. 
' The Vertebrce of Dicynodonts. (Plate 12, figs. 2, 4.) 
No specimen has hitherto been described which shows the actual association of the 
vertebrae attributed to Dicynodon with a skull ; for, although there is a strong 
probability that the vertebrae and limb-bones from the Gonzia River, attributed to 
Dicynodon tigricepsf are parts of the same animal as the skull from the same locality, 
the specimens are separated. It is, therefore, interesting that the skull fragment 
already described in this memoir (Brit. Mus., R. 866) has the earlier cervical 
* ‘ Catalogue South African Reptilia,’ No. 66, p. 40, Plate 35. 
2 K 
MUCCCLXXXIX.—B. 
