188 
PROFESSORS T. W. BRIDGE AND A. G. HADDON 
of two rounded lobes, almost entirely separated the one from the other, and sunk in 
great hollows situated in front of the “ apophysis of the large vertebra.” These lobes 
are said to communicate under the body of the vertebra, but from the fact that the 
bladder had already been displaced, the writer adds that he was not certain how it 
was attached, and that the ligaments deserve a new and special dissection. 
Day (9) also briefly refers to the air-bladder of Clarias, and includes the genus in 
his list of Siluridte in which the organ is said to be constricted into two lateral 
portions, and more or less completely enclosed within bony capsules. 
Our own investigations have been confined to Clarias nieuhojii, C. magur, C.fuscus, 
and C. anguillaris. 
In Clarias nieuhojii the skull and the first, the complex, and the fifth vertebrae are 
rigidly connected together (i.) by the superficial ossifications which continuously invest 
the lateral surfaces of the various centra, and also extend on to and thicken the sides 
and under surface of the basioccipital ; and (ii.) through the firm sutural union of the 
correlated vertebral and cranial elements (figs. 78, 79). The body of the first vertebra 
(fig. 79, is very small ; its dorsal surface forms part of the fioor of the neural 
canal, but the ventral and lateral surfaces are covered by the forward extension of the 
superficial ossifications on to the basioccipital [h.o.). In a vertical longitudinal section 
of this region of the vertebral column the centrum in question appears as if completely 
enclosed within the contiguous concavities of the complex centrum and basioccipital. 
Tlie centrum of the complex vertebra (figs. 78, 79, c.c.) is nearly three times the 
length of that of the fifth, and a single pair of fairly large nutrient foramina are 
visible on its ventral surface. Except for the asymmetrical development of its con- 
cavities, the anterior being much deeper than the posterior, the body of the fifth 
vertebra (r’.^) in no way differs from the normal centra. The sixth vertebra is normal 
and free (fig. 80, v.^), but, unlike most other Siluridce, its well-developed transverse 
processes do not carry ribs, the seventii [v.'’) being the first of the rib-bearing series. 
The confluent spinous processes of the third and fourth vertebrae (fig. 79, n.s.®), 
11 . s.^), form a thin vertical plate of bone which is much higher behind than in front. 
The low, but much-thickened anterior margin of the plate, and the neural arch of the 
complex vertebra incline forwards, so as to abut against the supraoccipital and ex- 
occipitals, a small amount of intercalated cartilage, however, still persisting between 
the apposed elements. A thin lamina of bone derived from the under surface of the 
supraoccipital spine {so.') firmly articulates by an interdigitating .suture witli the 
dorsal margin of the confluent spines. The uncleft spine and arch of the fifth ver- 
tebra {n.s.^) are firmly, but by suture only, united to those of the preceding vertebra. 
As in nearly all other Siluroid Fishes, the walls of the neural canal are fibrous over a 
somewhat triangular area {at. a.) on each side between the arch of the complex vertebra 
and the exoccipital (fig. 79). 
The transverse processes of the fourth and fifth vertebrae are greatly expanded and 
so modified as to form on each side a more or less complete bony funnel, which is 
