262 MR. ST. GEORGE MIVART ON THE (Apr. 28, 
girdles, and by means of membrane. These lateral structures con- 
stitute paraxial parts. 
These paraxial parts, unlike the neural arches, are not only always 
incomplete below, but they are not usually ossified continuously with 
the centra respectively supporting them, but generally consist of 
two parts—a transverse process and a rib—the rib being somewhat 
moveably articulated to the distal end of the transverse process. The 
ribs end freely, except those articulated with the pelvic bones, which 
are intercalated parts of the appendicular skeleton. Certain vertebree 
have transverse processes only, their free ends more or less widely 
diverging. This is the case with the caudal vertebree, except the 
first two or three of them (which sometimes support ribs), and 
also occurs in many trunk-vertebree of Amphiuma, Proteus, and 
Stren. 
The caudal vertebree, except the first one or two and the very last, 
have almost always inferiorly extending processes and arches, con- 
tinuously ossified like the neural arches above them, which moreover 
they resemble, more or less closely, in shape and proportions. They 
are probably serial homologues of subcentral processes of the trunk- 
vertebrae ; and the whole of these inferior structures belong to a 
system of hypazial parts*, and are the hypapophyses. That such 
subcentral processes really do answer serially to the caudal hypapo- 
physes behind them, is well shown in Siren, where the vertebra in 
front of that which bears large hypapophyses is furnished with a 
pair of smallebackwardly projecting processes exactly like those of 
Spelerpes rubra, but at the same time developed from the posterior 
end of the ridges, which unquestionably represent the hypapophyses 
of the vertebre next behind (fig. 14). 
Sometimes instead of, or besides, lateral processes, the inferior 
surface of a centrum will develope a median longitudinal bony ridge. 
Such a structure is to be seen in the third and fourth vertebree of 
Stren (fig. 13, Hy), the second and third vertebree of Menobranchus, 
and in many of the trunk-vertebree of Proteus and Amphiuma. This 
ridge is hypaxial, and may also be spoken of as hypapophysialt, as 
sometimes in Stren and Menobranchus it seems, by becoming medianly 
grooved, to divide into a pair of hypapophyses. The propriety of 
regarding the subcaudal arches and processes as distinct from paraxial 
parts is justified by the frequent coexistence of the latter together 
with subcaudal arches in the tail. Moreover these arches are un- 
doubtedly the representatives of the ‘chevron bones’ of the Croco- 
dile ; and these latter were found by Professor Goodsir¢ to be, at the 
root of the tail, enclosed within the backwardly continued peritoneal 
folds and abdominal cavity, thus removing them altogether from the 
* As before said, I believe the hyobranchial apparatus, jaws and trabecule, to 
also belong to the system of hypaxial parts. 
+ The way in which the osseous extension (of the under surface of the centrum) 
related to the great arterial channels is represented by a single process, by a 
pair of processes, or by a triple development, is well shown by Professor Owen 
(Memoir on the Megatherium, Phil. Trans. 1851, part 2. plate lii. figs. 48-51). 
{ Edinb. New Phil. Journal, January 1857, p. 128. 
