60 
SYSTEM OF NATURE. 
on the contrary, there is a flexibility equally remarkable : in 
fine, there is scarcely any character of the skeleton, respi¬ 
ration or generation, in which the two groups closely cor¬ 
respond, so that it becomes doubtful whether the crocodiles 
more nearly approach the tortoises or the lizards. Latreille 
has uhited them with the Testudinata under the name of 
Cataphracta; and I observe that Mr. Gray, in a late Sy¬ 
nopsis of the contents of the British Museum, retains the 
group under Latreille’s name, but adds the Amphisbaenia, 
which, like the crocodiles, are covered with square bone¬ 
like plates. I am quite disposed to agree with these emi¬ 
nent zoologists in giving full weight to structural characters, 
rather than yield to superficial resemblances. As some¬ 
thing in the way of a connecting link between the tortoises 
and crocodiles,—mainly distinguished by the bony carapax 
of the former,—it maybe mentioned that it has been asserted 
that, in the Tilgate Forest strata, the ribs of turtles occur 
quite unconnected with any portions of a carapax. Unless 
there is a mistake in this, the animal must combine the 
structures of crocodile and tortoise. 
Now supposing it were established that the Cataphracta 
form one distinct class, and the Squamata of Merrem—in¬ 
cluding lizards and snakes—another, how shall we dispose 
of the fossil genera Ichthyosaurus, Plesiosaurus and Mosa- 
saurus, all of which were closely allied to lizards in the 
form of their heads. It will be observed, from the very 
interesting particulars collected by Dr. Buckland* con¬ 
cerning the osteology of these creatures, that they were 
huge marine lizards, every portion of their skeletons exhi¬ 
biting similarity to the lizards rather than the crocodiles. 
* Bucklancl’is Bridgw. Treat, i. 168—220. 
