THE DINOSAURIA. 
transverse processes disappear in its hinder part, and the earlier 
caudal vertebrae are usually massive hut short (Fig. 7). There 
is nothing about the tail suggestive of resemblance to a bird, 
and the characters of the caudal vertebrae of Archaeopteryx 
are quite different, nor would it be worth while examining 
how far the tail may be paralleled among the Mammalia. It is, 
however, right to mention that in certain Dinosaurs, which are 
otherwise imperfectly known, the articular ends conform to 
the Crocodilian pattern in being procoelous. 
When the elements which form the vertebral column are 
freed from the matrix so as to be adapted together by their 
natural surfaces, they usually fall into three marked curves. 
First the cervical vertebrae form a curve which is convex on 
the under side and concave above, showing that the neck 
was usually arched upward. The dorsal and sacral vertebrae 
similarly form a curve in the opposite direction, that is to say, 
convex in length on its upper side, between the fore and hind 
limbs, indicating that the back was arched. While the tail 
vertebrae similarly form a curve in the reverse direction, or 
concave on the upper margin, and its concavity is usually so 
well marked as only to be intelligible on the hypothesis that 
the root of the tail was elevated to a considerable height 
above the ground. In the larger-tailed Dinosauria the depth 
and thickness of the tail was immense, and perhaps entirely 
Crocodilian in the relative proportion which it bore to the 
depth of the body. 
The ribs in the dorsal region appear to have been long, 
well curved, directed outwards, and sometimes upwards, so as to 
enclose a large visceral cavity. They are usually compressed 
from side to side, and sometimes flattened above, so that the 
cross-section of the upper part of a rib resembles a capital T. 
This character is especially seen in the Dinosaurs from the 
Greensand of England and Gosau ; and, judging from the way 
in which the transversely expanded neural spine of Belodon is 
correlated with the support of dermal armour, it seems probable 
that lateral stripes of dermal armour may be inferred from 
this condition of the dorsal ribs, especially as massive dermal 
plates are found in association with the remains in both those 
localities. The specimen of Ceteosaurus at Oxford includes a 
bone which Professor Phillips interpreted as the sternum. 
Such a bone may perhaps exist in some of the German triassic 
genera, but no other trace of it has been met with, and but 
few ribs could have been articulated to it. 
The scapular arch changes but little throughout the entire 
group (Fig. 9). It consists of a broad rounded or quadrate 
coracoid, which unites with the long and flattened scapula to 
form the glenoid cavity which receives the head of the humerus. 
