habits.
Fig. 78. Ophiacodon. Proatlas, axis, and ribs.
The marine crocodile, with a fin-like tail, lost two, like the mosasaurs and aigialosaurs, having seven; Pleurosaurus probably had but five; and the ichthyosaurs, the most specialized of all aquatic reptiles, had practically no neck.
The first two or three of the cervical vertebrae are markedly differentiated in all reptiles, as in the higher animals. The first of these, the proatlas, is inconstant and vestigial, and has not been included in the numbers above given. The second, the first of our usual nomenclature, is the atlas. The third, more or less closely united with the atlas, is the axis, or epistropheus. The following cervical vertebrae, when present, are differentiated more or less from the dorsal series by their less erect or shorter spines, transverse processes, or the slenderness and mode of rib articulation. The cervicals of the later pterodactyls have additional articulations on their ventral sides, as has been described above (p. 91).
Fig. 79. Theromorph vertebrae: A, Dimetrodon, atlas and axis; B, the same atlas, from the front; C, the same proatlas, from the side; D,
Sphenacodon, neurocentrum of atlas, inner side. i, intercentrum; o, pleurocentrum (odontoid); n, neurocentrum (arch).
Proatlas. The proatlas (Figs. 79 c, 80 d, l) is a small, more or less vestigial neural arch between the arch of the atlas and the occiput, usually paired. It is believed to be the arch of a vertebra formerly intercalated between the atlas and the skull; by some, homologous with the so-called atlas of the Amphibia; by Baur, as the representative of a vertebra fused with the occiput in the reptiles; by others, as merely the separated spine of the atlas; by others, as the arch of a vertebra whose centrum is represented by the anterior end of the odontoid. Another theory, which has less to commend it, is that of Jaekel, namely, that the
