OE THE SKULL IN THE UKODELOUS AMPHIBIA. 
575 
The faintness of this attempt to form cranial vertebrae, is displayed in a peculiar 
manner ; for the conical segment is enclosed in an ectosteal bony sheath, exactly like 
that on the first following vertebrae of the spine ; this is what may be called a cembalo- 
style. 
But the neural laminae of the cephalostyle are formed by the trabeculae. 
The second occipital segment, so well defined, and so perfect in shape, has merely 
calcareous grains dispersed over its surface, and exactly resembling the semicrystalline 
points that gather and cluster round ossifying cartilage-cells. 
Although a bony cephalostyle occurs very constantly in the larvae of Caducibranchs, 
this is yet the most perfect I have seen ; in Salamandra maculosa it seems to be a 
tubular process of the parasphenoid, and in all it is transient, only continuing for some 
stages as a groove with lateral ridges on the upper surface of the parasphenoid. 
It is as well to say at once that here, in these types, the first segment belongs to the 
postpituitary region, therefore to the hinder half of the basisphenoid, and the second 
only to the basioccipital *. 
In this species the first vertebra is a large joint, as large as its successors ; but in 
several Caducibranchs there is a structure in this part, so curious and instructive, that 
I must mention it here ; although the copious illustrations of these types, prepared by 
me, must wait for publication at some future period. 
The Common Newt ( Triton cristatus), Spelerpes salmonea, S. rubra, and Notoph- 
thalmus viridescens have shown me this structure most clearly in their larval state. 
There is a deep notch between the occipital condyles, which, like those of the large 
vertebra to which they are articulated (its “ pro-zygapophyses ”), are pedunculated. 
Where the notochord lies between the occipital condyles, there the parachordal 
cartilage is deficient, for a time ; but a small oval posterior parachordal arises on each 
side. As the peduncles of the anterior articular facets of the first vertebra are very 
long, there is a considerable tract of the notochord left uncovered, in front by the para- 
chordals, and behind by that vertebra. 
That tract acquires its own bony sheath ; the small, intercalary parachordals become 
separately" ossified, like the exoccipitals, and then coalesce with the anterior end of the 
long notochordal style. 
We thus get what at first sight appears to be a mere “ odontoid process,” but which 
is a true vertebra, having all the essential elements thereof f. 
* I am now satisfied that in the Yertebrata, generally, the hind part of the trabeculae, which lies upon the 
front part of the parachordals, is the source from whence the posterior clinoid wall grows. 
t See Professor Huxley’s article “ Amphibia,” Encycl. Brit. vol. ix. p. 752. After saying that a similar 
process is seen in the Bays, he goes on to say, “ The first spinal nerve which has the distribution of the hypo- 
glossal of the higher Vertebrata passes out of the spinal canal either between the first and second vertebrae, or 
through the foramen in the arch of the first, in the Amphibia , which have no proper suhoccipital nerve. This 
is a very curious circumstance, and requires elucidation by the study of development.” 
I cannot help thinking that both the intercalary vertebra, just described, and also the one through or behind 
which the hypoglossal or suhoccipital nerve passes, are both undifferentiated from the occipital arch in the 
