ON THE ANATOMY OF FISHES. 
171 
of the foramen through which the tripus is connected with the scaphium, there is a 
small radial nodule attached to ‘the lateral surface of the complex centrum. No 
dorsal laminae could be made out, nor any obvious grooves for the posterior cardinal 
veins. 
The stem and ascending process of each post-temporal (fig. 68. pt.s. and pt.a.) have 
a somewhat curved posterior margin which strengthens, and to some extent deepens, 
the mouth of the funnel at its anterior and dorsal edges. The inferior limb (fig. 70, 
pt.i.) has the usual articulation with the lateral surface of the hasioccipital, and for a 
part of its extent is firmly applied, if not actually anchylosed, to the anterior wall of 
the funnel. The socket for the clavicle (cl.s.) is formed by a groove between the 
stem and inferior limb of the post-temporal, which is completely closed posteriorly by 
the distal margin of the anterior wall of the funnel-like transverse process. 
As in Glyptosternum and Auchenipterus the outer surface of each auditory capsule 
is marked by a deep cup-shaped depression. 
The air-bladder (figs. 68 and 70, a.s.) is reduced to the condition of two thin- 
walled, laterally situated air-sacs, which occupy the cavities of their respective bony 
funnels. Each sac is nearly spherical in shape but slightly contracted at the inner 
extremity where it is in relation with the tripus. Laterally, the sac bulges outwards 
into the wider portion of its funnel and has its outer wall nearly, if not actually, in 
contact with the inner surface of the adjacent lateral cutaneous area.* In fresh 
specimens, or when the air-sacs are not shrunken by the action of preservative fluids, 
it is probable that the two sacs are closely and extensively applied to these areas. 
Each sac is sufficiently large to be in close contact on all sides, except the outer wall, 
with the inuer surface of its investing bony capsule, but with the exception of the 
inner wall no fibrous connection between the two could with certainty be made 
out. In one of our two specimens of Callomystax we noticed that the two air-sacs 
were not symmetrically developed. The one on the right side had extremely thin 
and almost transparent walls, and appeared also to be somewhat the larger of the 
two, its outer wall being closely applied to the lateral cutaneous area of its side. 
The left sac had much thicker and more opaque walls, and, from its smaller size, the 
outer wall was widely separated from the external skin. 
We could discover no communicafion between the two air-sacs, neither were we 
able to detect any trace of a ductus pneumaticus. 
An examination of the structure of the walls of tlie two sacs at once shows that in 
the arrangement of their component fibres, and in the relations of the latter to the 
tripodes, each sac is substantially similar to the lateral half of an anterior chamber 
in a normal air-bladder. The crescentic process of the tripus is imbedded in the dorsal 
wall of the sac near to its anterior and inner extremity, and the fibres radiating from 
the convexitv of the process form, as usual, an inner and an outer stratum. The 
* The antcro-posterior extent of each sac was 8 mm., and the width 7'5 mm. The length of tlio 
Fish itself was 6 inches. 
z 2 
