154 
PROFESSORS T. W. BRIDGE AND A. C. HADDON 
a normal anterior chamber, more particularly as they occur in such Siluridae as Arius 
and its allies. The attachment of the anterior and posterior walls of each sac to the 
corresponding margins of the modified transverse process of the fourth vertebra are 
respectively represented in Arius j^iclctda for example, by the almost precisely similar 
skeletal attachments of the dorsal edges of the anterior wall of the anterior chamber, 
and of the primary transverse septum or posterior wall. The attachment of the 
mesial part of the dorsal wall of a normal anterior chamber to the ventral surface 
and sides of the complex centrum is apparently in part represented in both Bagarius 
and Glyptosternum by the sheet of fibres which extends from one air-sac to 
the other, and is at the same tmie firmly adherent to the ventral surface of the 
centrum in question, and which we caimot but regard as the consolidated mesial 
portion of an originally normal anterior chamber. If, as we have suggested, the 
dorsal ridge of Bagarius is the equivalent of the oblique lateral ridge of the complex 
centrum and the dorsal lamina of one side, the attachment of the inner wall of 
each air-sac to it might reasonably follow fi-om the constriction of an originally 
normal anterior chamber into t'svo laterally-placed sacs, and the consequent approxi- 
mation of their anterior and posterior walls at the inner extremity of each sac. The 
same remark will apply also to Glyptosternum, although in this Siluroid the dorsal 
lamina is more easily recognised. The superficial coat of the air-bladder is obvious 
enough in both genera, but while it retains its normal attachments anteriorly to the 
decurved anterior margin of the transverse process of the fourth vertebra, the 
atrophy of the lateral compartments has enabled it to enter into a similar attachment 
posteriorly to the hinder margin of the same process. It is by no means improbable 
that the marked downward curvature of the anterior margin of this process is largely 
due to the actual ossification of that portion of the superficial coat which we have 
hitherto called the transverse membrane. The peculiar ventral process which, on 
each side, grows out from the ventro-lateral edge of the complex centrum may also, 
with some probability, be regarded as due to the extension of bony deposit from the 
superficial ossifications of the complex centrum into the superficial fibrous coat. 
But while retaining in a fashion the usual rigid skeletal attachments the relations 
of the walls of the air-sacs to the movable tripodes afford further indications of the 
effects of degeneration. We entertain no doubt of the practical atrophy of the 
dorsal wall of each sac, and the almost complete disappearance of the usual strata 
of fibres by which, in the normal anterior chamber, the lateral and antero-lateral 
walls are functionally connected with the tripodes. With the suppression of these 
fibres the absence of the peculiar crescentic curvature in the posterior division of the 
tripus must be associated. 
We shall subsequently have occasion to discuss these facts from a physiological 
point of view, and we shall now merely remark that in our opinion it is quite clear, 
from the partial enclosure of the two air-sacs within bony capsules, combined with the 
skeletal attachments of their anterior, posterior, and ventral w^alls, that no volumetric 
