—I. The Dinosauria, with the cervical vertebræ relatively short, and the femur as long as, or longer than the tibia. II. The Compsognatha, with the cervical vertebræ relatively long, and the femur shorter than the tibia.
3. The affinities of the Ornithoscelida with other Reptiles.
If we consider the relations of the Ornithosauria to other reptiles, it is at once obvious that they belong to that great division of the class in which the thoracic vertebrae have distinct capitular and tubercular processes, the latter being longer than the former, and springing from the arch of the vertebræ, as in the crocodiles. These reptiles may be termed Suchospondylia, to distinguish them from another great group, in which the thoracic vertebræ have the capitular and tubercular processes fused together into one process or facet, and which may be termed the Erpetospondylia,—from a third, in which the capitular and tubercular processes are both mere tubercles springing from the centrum of the thoracic vertebræ, Perospondylia,—and from a fourth, Pleurospondylia, in which the thoracic vertebræ have neither capitular nor tubercular transverse processes, but the ribs are sessile upon, and fixed to, the vertebræ.
The last-named group consists of the Chelonia; the Perospondylia contain only the Ichthyosauria; the Erpetospondylia comprise the Ophidia, Lacertilia, and Plesiosauria; while the Suchospondylia embrace the Crocodilia, the Dicynodontia, the Pterosauria, and the Ornithoscelida.
The closest relations of the Ornithoscelida within this group are with the Dicynodontia on the one hand, and the Crocodilia on the other. The sacrum and the iliac bones of the Dicynodonts more closely resemble the corresponding parts of the Ornithoscelida than they do those of any other Reptilia, except the Pterosauria[1]; and there are a good many points of resemblance in the skull and dentition. Our knowledge of Rhopalodon and of Galesaurus is hardly sufficient to afford grounds for a safe opinion: but it seems probable that they will turn out to be annectent forms between the Dicynodontia and the Ornithoscelida.
The connexion of the Crocodilia with the Ornithoscelida is probably to be sought in some common form, more Lacertilian in its character than any of the known members of either of these groups. The oldest known Crocodilians, Belodon and its congeners, exhibit modifications which approximate them rather to the Lacertilia than to the Ornithoscelida.
If we seek for reptilian allies of the Ornithoscelida in formations of older date than the Trias, the Permian forms alone present themselves. Our knowledge of these is almost entirely due to the researches of Von Meyer, the results of whose investigations have hardly received the attention they deserve. They prove the existence of two very distinct reptilian genera, Proterosaurus[2] and Para-
