146 
PROFESSORS T. W. BRIDGE AND A. C. HADDON 
surprising that these structures should be so well developed when the diminutive and 
rudimentary condition of the air-bladder is taken into consideration. It must be 
admitted, however, that our suggestion receives some support from the fact that in 
the allied genus Glyptosternum, where a definitely recognisable dorsal lamina is 
present, there is no trace of the dorsal ridge. 
The post-temporal'"* has its ascending process and stem somewhat expanded, but 
the inferior limb (pt.i.) is an extremely slender process, which, after leaving the stem, 
is firmly applied if not actually anchylosed to the distal extremity of the transverse 
process of the fourth vertebra, but nevertheless becomes detached therefrom as it 
passes inwards, and eventually fuses with the anterior end of the dorsal ridge on the 
lateral surface of the basioccipital. The socket for the clavicle is in part formed by a 
deep groove in the post-temporal at the junction of the stem with the inferior limb, 
and in part by the distal extremity of the modified transverse process which converts 
the groove into a tubular socket. 
The air-bladder is represented by two thin- walled oval sacs, which occupy the 
bony recesses formed by the spout-like transverse processes of the fourth vertebra 
(fig. 53, a.s.). In proportion to the bulk of the Fish the size of the air-bladder is 
extremely insignificant, and it is evident that beyond a certain period, which is 
probably reached at a very early stage, the bladder does not increase in size with the 
growth of its possessor. In our specimen a very immature Fish, of not more than 
6 inches in length, and probably weighing less than 3 ounces even when fresh, each 
air-sac was about 4 mm. long and 3 ‘5 mm. wide, that is to say, about the size of an 
ordinary garden pea, which, according to Taylor (38), was also the size of an air-sac 
in a specimen weighing 10 pounds. We could discover no trace of any communication 
between the two sacs, or of lateral compartments, or of the existence of a ductus 
pneumaticus. The ventral surface of each sac is invested by a thin but tough fibrous 
membrane, which stretches between and is firmly attached to the decurved anterior 
and posterior margins of the modified transverse process, and would seem also to 
extend from one sac to the other across the ventral surface of the complex centrum. 
The distal openings of the bony recesses are closed by their respective lateral 
cutaneous areas, but the outer wall of each air-sac is separated from the latter by a 
certain quantity of fatty tissue, as well as by a portion of the lateral lobe of the liver. 
As in other Siluridse, the w'alls of the air-bladder are firmly attached to the contiguous 
skeletal elements at certain points. The posterior wall of each air-sac blends with the 
investing fibrous membrane, which we must identify as the ventral portion of the 
superficial coat of the air-bladder, and the two are dorsally attached to the anterior 
margin of the transverse process of the fifth vertebra, and also to the posterior 
portion of the dorsal ridge as the latter passes on to the side of the complex centrum, 
much in the same way that the lateral portions of the posterior wall (i.e., the primary 
transverse septum) of a normal anterior chamber are inserted into the dorsal lamina 
* In fig. 53 only the inferior limb is represented, the remainder of the bone having been removed. 
