ON THE ANATOMY OF FJSHES. 
199 
Group ; — Chacina. 
Chaca lophioides. 
Cuvier and Valenciennes (8) describe the air-bladder of this species as very 
large, broader than long, and composed of two lobes, situated one on each side of the 
vertebral column. Its large size may, perhaps, justify the inference that the air- 
bladder is of the normal type. 
Sub-family : — SILURIDiE HETEROPTERiE. 
Group : — Silurina. 
Saccobranchus fossilis. 
There is a short, but in some respects very inaccurate, description of the air-bladder 
of this species by Cuvier and Valenciennes (8, vol. 15, p. 301). Taylor (38) has 
also briefly referred to the organ as being very similar to that of Clarias magur. 
Day’s account (9) is equally brief: “air-vessel tubular, placed transversely across the 
body of the anterior vertebrae, where it is entirely enclosed within a bony capsule” 
(P- 712). 
With regard to the osseous capsules in which its diminutive air-bladder is enclosed, 
Saccobranchus presents an interesting gradation between Clarias and Callomystax, 
but the bladder itself is almost precisely similar to that of the former genus. The 
somewhat elongated funnel-like capsules are very similar on the whole, both in shape 
and mode of formation, to those of Clarias nieuhojii. In one or two features, however, 
Saccobranchus closely resembles Callomystax. The anterior margin of the transverse 
process of the fourth vertebra is strongly deflected throughout its entire length, but 
to a greater extent in its distal two-thirds, where it forms not only the anterior but 
also the ventral wall of each funnel, and meets behind the deflected and anteriorly- 
curved posterior margin of the transverse process of the flfth vertebra. As in 
Callomystax, the decurved anterior margin of the first-mentioned process unites 
internally with the superficial ossification investing the lateral surface of the complex 
centrum, and consequently converts what in Clarias is a mere slit for the passage of 
the tripus from the air-bladder to the scaphium into a complete foramen. The 
characteristic ventral processes which in Clarias help to complete the ventral walls of 
the funnels are absent in Saccobranchus, unless the slightly-retroverted edges of the 
lateral ridges bounding the aortic groove may be regarded as rudiments of them. 
The two funnels will therefore have complete bony walls except over a small area in 
the proximal portion of the ventral surface, and distally where they are closed only by 
the lateral cutaneous areas. The passage for the posterior cardinal vein (cardina. 
groove) is well marked on the right side, and, anteriorly both the vein and the slender 
