ON THE ANATOMY OF FISHES. 
295 
characteristic of all the Siluridae normales, and as equally conspicuous by its absence 
in all other Ostariophysese, may possibly have some physiological significance. 
Sagemehl {loc. cit.) has already referred to this feature in certain Siluroids, and 
suggests that, physiologically, it enables certain advantages to be realized, which, in 
other Ostariophysese, are gained by the bipartite structure of the air-bladder. In 
referring to the two-chambered air-bladders of the Cyprinidse, Characinidae, and the 
Gymnotidae, he directs attention to the fact that this subdivision of the air-bladder 
occurs in no other Teleostei, and is always associated with the presence of a Weberian 
mechanism. The variable ratios of pressure of the surrounding medium condition 
the increase or decrease of the volume of gas contained in the whole air-bladder, and, 
owing to the peculiar structure of that organ, these variations of volume will find 
their expression almost exclusively in corresponding oscillations in volume of the 
small elastic anterior chamber, to which alone the Weberian ossicles are related, 
the larger posterior sac being scarcely, if at all, affected. A similar bipartition, the 
same writer also remarks, occurs only in a few Siluridae, among which are mentioned, 
apparently on the authority of Valenciennes, Auchenoglanis^^ and Auchenipterus. 
But in the case of the latter family Sagemehl says that the same physiological result 
is arrived at in an entirely different way. In many Ariinae, and also in such Siluridae 
as Callichrous, Cryptopierus, and Schilbe, the air-bladder is not confined to the 
abdominal cavity proper, but its lateral portions extend outwards between the ventral 
and dorsal portions of the lateral trunk muscles, quite close to the external skin. In 
this manner, Sagemehl remarks, the air-bladder acquires a tolerably direct connection 
with the external medium, and will, therefore, much more promptly react to pressure 
oscillations than if entirely confined to the abdominal cavity. 
Sagemehl is scarcely correct in attributing any important physiological distinction 
between the obviously bipartite air-bladders of the Cyprinidae and the Characinidae 
and the air-bladders of the normal Siluridae. In both cases the air-bladder is 
physiologically two-chambered in the sense that it consists of a smaller distensible 
anterior portion and a larger posterior and indistensible part, but in the Cyprinidae 
the bipartite character is obvious externally, whereas in the Siluridae it is only by 
internal examination that the essentially bipartite nature of the air-bladder becomes 
obvious. In fact, in all the normal Siluridae the air-bladder is, physiologically, as 
truly bipartite as in other Ostariophyseae, and the former derive precisely the same 
physiological advantage in the registration of hydrostatic pressure variations from 
this fact as do the latter. Neither is Sagemehl quite accurate in supposing that 
Auchenoglanis^' and Auchenipterus possess an air-bladder which is bipartite externally 
as in the Cyprinidae, inasmuch as we have already shown, in the morphological section 
of this paper, that in both genera the bladder is substantially similar to that of all other 
Siluridae normales so far as this point is concerned. Sagemehl’s suggestion that the 
* Auchenaspis, Bleek. 
