276. MR. ST. GEORGE MIVART ON THE [Apr. 28, 
into each arm of the notch (fig. 16), ending in two strongly project- 
ing processes, like the similarly conspicuous productions of the caudal 
neural arches before described. These processes are also especially 
marked in Proteus, Amphiuma (fig. 17), and Spelerpes rubra, and 
more or less so in Triton palmatus, Pleurodeles waltlii, &e. In 
Spelerpes rubra these prominences descend almost vertically, corre- 
sponding with the position of their vertical homologues above. The 
median inferior crest is often more vertically extended than is the 
corresponding neural one, as is the case in the species last men- 
tioned (except Amphiuma, where there is hardly any median inferior 
crest, and then only in the very posterior caudals) and in Aneides. 
These backwardly projecting lateral processes extend over the anterior 
part of the hypaxial arch next behind, and thus somewhat simulate 
posterior zygapophyses ; but these arches have no true articular 
processes, as is well seen in such forms as Cryptobranchus, Meno- 
poma, and Amphiuma, where the contrast is marked. Very generally, 
as in Triton, Amphiuma (fig. 17), Pleurodeles, and, above all, in 
Fig. 17. 

ay 
Last trunk- and anterior caudal vertebree of Amphiuma (No. 579 a in College of 
Surgeons’ Museum). 
H. Hyperapophysis. Hy. Hypapophysis. 7. Interzygapophysial ridge. 
Chioglossa, an oblique ridge extends backwards and downwards from 
the anterior point of attachment of each half of the hypaxial arch to 
the backwardly projecting inferior process of the same side. These 
ridges appear to be the only parts of the hypapophyses that are left 
in those forms in which the hypapophysial arches are very narrow. 
They somewhat resemble the interzygapophysial ridges of the caudal 
neural arches, but are more oblique in direction. 
Tue CervicaL VERTEBRA. 
The vertebra which comes first and articulates in front with the 
skull differs importantly from all the vertebree which succeed it. 
Like the others, it constantly presents a centrum which is concave 
behind, and from which a continuous, ascending neural arch is de- 
veloped. But not only are there no hypapophyses, but transverse 
processes are completely absent, except in Szren (fig. 18), Proteus, 
and Menobranchus ; and in these genera they are quite rudimentary, 
while the vertebra is invariably destitute of any costal appendage or 
rib. But it is the anterior face of the vertebra which is most re- 
markable. This presents on each side (at the junction of one half 
