iVKERlTlIEIlIUM. 
113 
of tins column arc: — the atlas, the last cervical, niueteon thoracic (in this scries there is 
a gap, one vertebra at least being missing), four luml)ars, and four sacrals. 
On the whole, the atlas is like that of M. Jjjonsi described above, hut ditfers in several 
minor points. Thus the transverse s])inous ridge of the neural arcli is wider ; the 
anterior border of the arch is less concave ; the transverse processes differ slightly in 
form and are more upturned ; the surfaces for the axis are deeper from above 
downwards and less extended from side to side. Some of tliese differences may be 
merely apparent and the result of abrasion or distortion. 
Ill the last cervical the centrum bears a fairly well-marked hypapophysial ridge ; 
its articular ends are oval in outline, tlie long axis being transverse, and are slightly 
concave in the middle. The transverse processes are large, rising partly from the 
centrum and partly from the arch ; they are stout, and somewhat thickened at the 
obliquely truncated end. The arch is high and is prolonged upwards into a 
prominent neural spine. ' There is no trace of an articular surface for the head of the 
first rib. 
In the first thoracic vertebra the centrum is of much the same shape as the last 
cervical, but the broad blunt hypapophysial ridge is almost restricted to the anterior 
end. The transverse processes are stout and short, terminating in a deeply concave 
rib-facet, looking downwards and forwards. The anterior edge of the transverse process 
is produced into a shelf-like projection of bone, which is continuous internally with 
the anterior edge of the neural arch and bears on its upper surface the broad and fiat 
anterior zygapophyses. The neural arch is broader than in the last vertebra and the 
pointed neural spine slopes strongly backwards. There is a slight cup for a rib-head 
on the upper angle of the anterior face of the centrum, and a much deeper and more 
sharply defined cup on the upper angle of the posterior face. 
The second thoracic is closely similar, but the tubercular facet of the transverse 
process looks directly downwards instead of downwards and forwards. At the same 
time the anterior shelf-like projection becomes smaller, and there is the first trace of a 
tuberosity on the upper side of the end of the process, which becomes larger till about 
the eighth thoracic. In this vertebra, owing to shortening, the transverse process 
seems to pass into the metapophysis which is largely developed in the following 
thoracic vertebrae. 
The following five vertebrae (text-fig. 44) are generally similar, but the transverse 
processes become shortened and the haemal ridge is wanting. Furthermore, in 4-7 the 
oval posterior capitular facet {c.f.) becomes very strongly concave and bordered by a 
prominent rim. In these first seven dorsals the neural spines {n.sp.) are comparatively 
slender, terminate in a point, and slope strongly backwards. In the eighth dorsal the 
form of the spine is transitional to the short broad type found in the posterior thoracic 
region. In this vertebra also the anterior capitular facet is merely a shallow concavity 
lying beneath and in front of the very short transverse process, which bears a small 
Q 
