ON THE ANATOMY OF FISHES. 
139 
of each cup is said to be covered over by the common integument which, at this point, 
is extremely thin and adheres to the surface of the bladder, presenting, when the 
latter is distended with air, an “ external elastic tumour ” of an oval figure. The two 
air-bladders, which have no communication with each other, or with the alimentary 
canal, apparently derive tbeir supply of air from a vascular tissue placed between the 
two cups where they are attached to the spine. The external coat is of a thin texture 
and argentine colour, and has a layer of fine adipose tissue interposed between it and 
the internal surface of the cup. 
This description is so different from anything met with in other species of Arius 
that we are not at all surprised at Day’s statement (9, p. 708) that the air-bladder 
Taylor examined “ could not have been in Arius yagovaA From an 'examination of 
an example of this species taken at Mandalay, in Upper Burniah, Day (Zoc. c^^.) gives 
the following description of its air-bladder : — “ Air-vessel large and somewhat heart- 
shaped, with a moderately thick external fibrous coat. On removing its front wall 
a longitudinal partition becomes apparent, but is not extended to its anterior portion. 
It has three transverse subdivisions forming it into five cavities, owing to the longi- 
tudinal partition commencing at the first transverse subdivision. These lateral cavities 
freely communicate with one another on the same side and with the opposite ones by 
means of the anterior chamber which does not possess any subdivision.” From this 
account it is obvious that Taylor’s description cannot apply to the air-bladder of a 
genuine Arius gagora, and, further, that this species possesses a perfectly normal 
bladder, subdivided internally by a primary transverse and a longitudinal septum, 
and, in addition, by two or more secondary transverse septa. It is not improbable 
that the species examined by Taylor was really Callomystax gagata ; at any rate 
his description agrees fairly well with what we found in an example of the latter 
species.* 
According to Johannes Muller (28) the air-bladder of Bagrus [Sciades) emphy- 
setus (Mull, and Trosch.) (= Arius emphysetus^ Gunther) is very long, and consists 
of three bladders disposed in a longitudinal series, and communicating through their 
connecting constrictions. 
Ilemipimelodus borneensis. 
In almost every respect the skeleton and air-bladder of this species closely conform 
to the Arioid type as illustrated by Anus pidada. 
The walls of the air-bladder are very thick, but somewhat thinner in the antero- 
lateral and lateral regions of the anterior compartment than elsewhere. The primary 
transverse septum is also very stout, and its width such as to render the orifices 
whereby the lateral and anterior chambers intercommunicate comparatively small. 
Three stout secondary transverse septa are present in each of the two lateral 
chambers, 
* Soc p. 171. 
T 2 
