1870. | AXIAL SKELETON OF THE URODELA. 261 
All the species of the order agree in possessing a spine made up, 
in the adult condition, of more or less similar vertebrae varying in 
number from 45 to 112, according to the species or individual. 
These vertebrze increase slightly and very gradually in size from the 
second vertebra till about the middle of the body. Thence they 
decrease again, at first gradually and slightly, but afterwards rapidly, 
and to such a degree that the last ones are only minute rudimentary 
ossicles. With the exception of the genus Siren, these vertebree 
may be arranged in four categories. 
1. Cervical.—This includes only one vertebra, namely that which 
articulates with the skull. 
2. Dorsal.—This includes almost all the trunk-vertebree, 7. e. all 
the vertebrze behind the cervical vertebra, and anterior to the sacral 
vertebra or vertebrze. 
3. Sacral.—This includes the vertebra or vertebrze to which the 
pelvis is attached. 
4. Caudal._—This includes all the vertebrze posterior to the sacral 
vertebra or vertebree. 
In the exceptional genus just mentioned (Siren) there is no sacral 
vertebra, and a cervical, and more or less arbitrarily divided dorsal 
and caudal regions are all that can be distinguished. 
Very rarely two contiguous vertebree will more or less completely 
anchylose together. I have observed this in the large species Crypto- 
branchus japonicus, where sometimes the last two presacral are fused 
together, and sometimes the sacral and the first caudal. -In a skeleton 
in the British Museum both these unions occur, so that the four 
originally distinct vertebree form actually but a pair, though each 
shows evident signs of its complex nature. 
Every vertebra, except the abortive ones towards the end of the 
tail, consist of a body (centrum), and of a neural arch ossified, I 
believe, continuously with it. The neural arches constitute epaxial 
parts. 
Every vertebra, except the cervical one and the very last caudal 
vertebrze, is furnished with lateral prolongations, never uniting below 
and forming complete arches, except by the intervention of the limb 

those arches (the vagus &c.) seem to be serially homologous with that portion 
of the spinal nervous system which is called sympathetic. 
The chevron bones of Mammals, Reptiles, and Amphibia I stated to be, in 
my opinion, hypaxial parts, and serially homologous with those hypapophysial 
processes which are so largely developed in the Pelican and the Great Auk, and 
which, in their azygos condition, are evidently situated in the line of suspension 
of the inner laminz of the ventral plates of theembryo. According to this con- 
ception, in vertebrates generally we have, at the anterior end of the axial skeleton, 
hypertrophied epaxial and hypaxial parts, z. ¢. the brain-case and visceral arches. 
Further back we have hypertrophied paraxial parts with much diminished 
hypaxial ones. Finally, towards the hinder end of the body (except in tailless 
forms) we have, in vertebrates above fishes, a reappearance of hypaxial elements 
generally accompanied by coexisting but distinct paraxial parts. In fishes, in 
the same region, we have generally paraxial parts in union with more or less of 
the hypaxial element, or we have paraxial parts only, or, much more rarely, only 
hypaxial parts. 
