AND CLASSIFICATION OF THE FOSSIL KEPT ILIA. 
209 
extended to have articulated with more. The tail is equally devoid of characters which 
suggest affinities. In depth the centrum has more in common with the Crocodilian 
and Dinosaurian types than with Nothosaurs, which have the centrum shorter, or with 
Lacertilians or Ornithosaurs, which have it more depressed. The divided condition 
seen in the neural spine of the later caudal vertebrae probably indicates a complete 
development of the neural arch upon each of the proto vertebral elements which go 
to make up the centrum ; though the vertebra must still be regarded as highly 
differentiated, since the caudal ribs are given off from its anterior moiety as processes 
directed transversely, while the chevron bones, which represent them on the posterior 
moiety, have already descended to the inferior visceral margin. The theory of the 
double-headed articulation of dorsal and cervical ribs is not unconnected in some 
animals with the hypothesis that the transverse process and chevron bone blend in 
the anterior part of the body to form one rib with two articulations, and sometimes 
with a pair of sternal ribs to each distal extremity. Tire attachment of the ribs being- 
high up, as well as the length of the ribs, would indicate that the respiratory and vital 
organs in Protorosaurus were well developed ; and the mode of attachment of the ribs 
in most Reptiles and higher Vertebrates appears to depend partly on the way in which 
they are elevated by the lungs, and partly on the muscles which come into play in 
connecting the ribs with the vertebrae. So that the double-headed attachment of the 
dorsal ribs in modern Crocodiles is fundamentally different from the attachment in Birds, 
only because the transverse processes have become so much elongated as to remove the 
rib from the side of the centrum. But the Mammalian and Avian ribs are typically 
single-headed, and the second head or tubercle is obviously only a consequence of the 
rib being brought into contact with the neural arch ; so that, if no transverse platform 
is developed, the rib cannot have a second articulation. And it is on this condition 
that I account for the single-headed ribs of Protorosaurus, since nothing is needed to 
make the rib double-headed but a transverse development of the neural arch, such as 
I shall subsequently describe as partially developed in the genus Mesosaurus. 
The sternal ribs are imperfectly known. Von Meyer represents them as rods, of 
which two, or possibly three in some cases, are attached to the enlarged sternal end 
of each costal rib. The sternal ribs, however, were probably composite ; and I am 
disposed to believe that each consisted of two lateral pieces on each side, united by 
squamous overlap with a median piece in the middle line of the abdomen. Sternal 
ribs are seen in Lariosaurus, in Mesosaurus, in Stereo sternum, in Rliynchosaurus, and 
other Triassic and Permian types, as well as in Plesiosaurus. Their structure is in 
every case substantially the same when it can be observed ; though the number of 
sternal ribs to each costal rib varies. The nearest approximation to this condition 
among existing Reptiles is seen in Platteria. I have no doubt Rhynchosaurus is a 
Rhynchocephalian ; but Lariosaurus, Mesosaurus, &c., are Notliosaurians. 
The, pelvis and hind-limb .—The pelvis is not complete in any specimen. But the 
ilium appears from the sum of the evidence to have an antero-posterior elongation 
2 E 
MDCCCLXXXVII.-B. 
