ON THE ANATOMT OF FISHES. 
287 
an original and primitive connection between the air-bladder and the internal ear, 
which preceded the evolution of the Weberian mechanism, but is apparently still 
retained in some cases. The suggestion is certainly an ingenious one, but as regards 
the Characinidse, we are not sure that much can be said in its favour. It is quite 
possible that the ligaments in question are merely illustrations of the many and 
various ways in which the air-bladder in the Ostariophyseae becomes more or less 
intimately attached to the skeleton for physiological purposes, and, after all, may 
have no such significance as that suggested. On the other hand, the antero-lateral 
caeca of such Siluroids as Macrones aor, Osteogeniosus, and Platystoma, are hollow 
forward prolongations of the air-bladder, involving extensions of both the tunica 
externa and interna, and in their case Sagemehl’s suggestion may be reasonably 
accepted. 
The persistent and usually bilobed “ head-kidney ” usually occiq3ies a recess of 
corresponding shape in relation with the anterior wall of the air-bladder, to which 
indeed its posterior surface is closely applied. The peritoneum invests the ventral 
surface of the air-bladder, and, anteriorly, extends over the corresponding surface of 
the “ head-kidney,” but at the anterior margin of the latter is reflected backwards 
along the dorsal surface of the oesophagus. Outwardly directed peritoneal 
cul-de-sacs in relation with the antero-lateral regions of the air-bladder are almost 
invariably developed for the reception of portions of the lateral lobes of the liver. 
Their formation appears to be due to the unusual lateral extension of the anterior 
chamber of the air-bladder, and its apposition on each side to the external skin 
(lateral cutaneous areas) — a modification which has led to the displacement of the 
lateral lobes of the liver from their normal position between the bladder and the side 
walls of the body, and to the outward and forward hernia-like protrusion of a portion 
of each lobe into special peritoneal pouches. Between the peritoneum and the 
proper ventral wall of the air-bladder the latter is invested by a superficial coat of 
connective tissue which varies in consistency and often assumes the condition of a 
definite fibrous membrane, more particularly where it invests the anterior wall of the 
bladder, in which position it forms a fairly strong, inextensible, transversely disposed, 
fibrous membrane, varying in thickness in different Siluroids, and specially attached 
to the skeleton along its dorsal edge and lateral margins. To a varying extent this 
superficial coat may be traced on to the lateral and dorsal walls, as well as the 
anterior wall of the anterior chamber, and also, but less obviously, on to the 
corresponding surfaces of the lateral compartments. The skeletal attachments of the 
transverse membrane vary somewhat in different Sduroids. Dorsally, it is always 
attached near the median line to the complex centrum and especially to its oblique 
lateral ridges, and by a special backward prolongation to the anterior margin of each 
dorsal lamina, and also to the anterior margins of the transverse processes of the 
fourth vertebra ; laterally, it is firmly attached to the decurved distal extremities of 
the transverse processes and there blends with the ligamentous fibres by which the 
