ON THE ANATOMY OP FISHES. 
303 
minor differences in the degree of development and the mutual relations of the 
different Weberian ossicles in the two families. Perhaps the most important differ- 
ence is the removal of the two claustra from any obvious physiological relations with 
the atrial cavities in the Siluridse, and it is by no means clear whether this should be 
regarded as a loss or a gain from a physiological point of view. 
Of the remaining families of the Ostariophysem the Characinidae closely resemble 
the Cyprinidae in so far as the points now under discussion are concerned, and of the 
Gymnotidae it may be said that they exhibit a somewhat less specialized condition of 
the air-bladder and Weberian ossicles than any of the preceding families, although 
obviously related to the Cyprinoid type. 
The conclusion suggested by these facts seems to us to be this — that, physiologi- 
cally considered, the most important distinctive features of the Weberian mechanism 
in the Siluridae as compared with other Ostariophyseae are mainly related to the air- 
bladder, which in the former attains its maximum degree of specialization as an organ 
adapted for the registration of varying hydrostatic pressures. 
B. Silui'idcE abnormales. 
In considering, from a physiological point of view, the significance of the structural 
modifications presented by the air-bladder and its adjuncts in this section of the 
family, we may, in the first place, briefly capitulate certain anatomical facts already 
mentioned in a previous section of this paper. 
Apart from a very considerable reduction, both in absolute and relative size, the 
absence of lateral chambers and the persistence of the anterior chamber only, the most 
noteworthy features in the air-bladder of the various genera of Siluridse abnormales 
are — its constriction, partial or complete, into two laterally-situated air-sacs, which may 
not only lose all connection with each other, but with the oesophagus also ; the 
tendency of the air-sacs to become more or less completely enclosed within variously- 
formed bony capsules ; the occasional atrophy of the ductus pneumaticus, or its persis- 
tence as a solid fibrous cord ; and frequently also, in addition, the more or less complete 
atrophy of the fibres which normally form the dorsal walls of lateral halves of the 
bladder, and connect the lateral walls with the crescentic processes of the tripodes. 
To these may be added the partial solidification of the organ in certain species, either 
through the thickening and matting together of its walls, or the development of 
internal fibrous trabeculae, or through the partial obliteration of its cavity through 
the growth of a thick central column of fibres. With the exception of the inter- 
calaria, which are often absent, without, however, in any way interfering with the 
complete anatomical connection of the tripodes with the scaphia and the claustra, 
which apparently have no functional significance in the Siluridae, the remaining 
Weberian ossicles are fairly well developed, and retain their normal relative size. 
Traces of degeneration are, however, to be met with. The tripus generally loses its 
