210 
PROFESSORS T. W. BRIDGE AND A. C. HADDON 
surface of the body of the sixth vertebra, and also of two lateral cornua (fig. 88, a.h.') 
which extend forwards into the recesses between the modified transverse processes 
and the ventral outgrowths. Anteriorly, each cornu terminates cmcally in the 
shallow cul-de-sac, in which each recess ends in front, the entire bladder thus 
describing a horse -shoe-shaped curve with the concavity directed forwards.* The 
walls of the bladder are extremely thick, except towards the anterior extremities of 
the lateral cornua, where they become comparatively thin. The transversely dis- 
posed portion, and the immediately adjacent sections of the lateral cornua, are more 
or less completely filled by a delicate network of fibrous tissue, which, however, 
completely disappears towards the thin- walled cmcal extremities. We could detect 
no trace of any free or open communication between the anterior cmca through the 
intermediate portion, or of a ductus pneumaticus. 
The skeletal attachments of the air-bladder can be more or less closely identified 
with those of the normal organ when allowance is made for its peculiar shape and 
curvature. The partially solid, intermediate section of the bladder is firmly attached 
to the posterior margins of the ventral outgrowths, while the inner wall of each 
lateral cornu is firmly connected with a thin and somewhat oblique longitudinal 
ridge of bone which traverses each of the lateral surfaces of the complex and fifth 
vertebral centra, and terminates anteriorly in a free projecting spicule, situated on 
the dorsal side of the thin- walled extremity of each cornu. The free spicule in all 
probability represents the dorsal lamina, and is almost normal in its position and 
relations ; the longitudinal ridge itself may perhaps be regarded as an abnormal 
development of the oblique lateral ridge of the complex centrum, in which case its 
extension backwards along the lateral surfaces of the complex and fifth vertebral 
centra must be due to the curious infolding of the proper anterior wall of the 
bladder in the formation of its characteristic horse-shoe-like curvature, which has 
caused the ridge to lie behind, and to become confluent with, the posterior, instead of 
the anterior, extremity of the dorsal lamina. The hollow thin-walled extremity of 
each lateral cornu has much the same structure as the lateral half of an ordinary 
anterior chamber, and into its dorsal wall the crescentic process of the tripus is 
inserted. It is not possible, however, to identify any special tracts of fibres con- 
verging towards the crescentic process from the anterior and lateral walls of the 
cfEcum, but that portion of the dorsal wall which extends from the inner edge of the 
process to an insertion into the free spicular extremity of the longitudinal ridge, may 
perhaps be considered to represent the radial fibres of other Siluridse. 
The relations of the ventral outgrowths to the air-bladder suggest that they should 
be regarded as due to an excessive development of such structures as are represented 
in a feebler degree by the characteristic ventral processes of BagaHus, Glyptosternum, 
and Clarias, or more closely perhaps by the bony laminse which, in such extremely 
* The width of the air-bladder was 2'5 mm., and the length measured along its curvature, 10 mm., 
the specimen itself being 4| inches long. 
