DR. A. GUNTHER’S DESCRIPTION OF CERATODUS. 
527 
of the skull (Plate XXXV. fig. 2) ; yet in a vertical section of the skull the course and 
termination of its tapering extremity can be distinctly traced, as the cartilage composing 
it is of a somewhat lighter colour than that of the skull. The boundary line between 
the cartilage of the skull and the notochord is still more distinct under the microscope 
(Plate XXXVI. fig. 1). The cranial cartilage (c) is distinguished by a considerable 
number of cells, many of which are spindle-shaped, having each end produced into a 
narrow process. These cells are entirely absent in the central parts of the notochord 
(m), the substance of which appears in the form of bundles of undulated fibres running 
in the direction from the central cylinder towards the periphery. 
The central gelatinous body (Plate XXXV. fig. 2, cd') is continued nearly to the 
extremity of the notochord, which lies opposite to the entrance into the acoustic cavity. 
The caudal portion of the notochord is tapering behind ; and its extremity terminates 
in a thread which is gradually lost between the upper and lower series of neural and 
haemal elements, which, in one specimen, coalesce into two tapering bands, and are per- 
sistent further backwards than the notochord. 
Wherever an organ is reduced to a rudimentary condition, individual variation occurs. 
Thus scarcely two specimens of Ceratodus will be found in which the caudal termina- 
tion of the vertebral column is exactly alike. One of the most remarkable variations is 
figured on Plate XXX. fig. 3. The notochord, with its whitish fibrous sheath, terminates 
here abruptly at a considerable distance from the end of the tail, its termination ( n ) being 
rather obtuse. The neural and haemal arches are continued beyond the end of the noto- 
chord, but, being no more separated by it, are now confluent, form one tapering band 
which extends to within half an inch of the tail, and which shows a distinct vertical seg- 
mentation. This is all the more worthy of notice, as a similar segmentation of the pos- 
terior extremity of the vertebral column has been observed in some other fishes with 
notochordal skeleton. In Ceratodus such a segmentation is evidently within the limits 
of individual variation ; it is confined to a continuation of the neural and haemal ele 
ments, and does not extend to the notochord. 
The notochord forms the base for about sixty-eight sets or rings of neural and haemal 
elements, the hindmost being quite rudimentary and so indistinct that it is impossible 
to give the exact number. Twenty-seven of these rings bear ribs; and the first (caudal) 
ring in which the haemal apophyses coalesce into a spine is the twenty-eighth. The 
bases of the apophyses are so deeply imbedded in the fibrous sheath of the notochord 
as to be in immediate contact with the cartilage of the notochord. The boundary line 
between them may again be discerned by the colour, the cartilage of the notochord 
being yellowish and slightly iridescent, whilst the cartilage of the apophyses is bluish. 
Under the microscope the latter appears as cartilage without fibrous basal substance, 
but with a great number of regularly ovate cells; whilst the cartilage of the notochord 
shows a distinctly fibrous structure with similar cells sparingly distributed near its peri- 
phery, the cells disappearing entirely towards the central parts. We have seen, above 
that no cells whatever could be found in the attenuated foremost portion of the noto 
MDCCCLXXI. 4 D 
