53 
newly discovered genus of serpentiform fishes, 
mous sac or air vessel, which extends in length from the 
extremity of the snout about 20 inches. The great delicacy 
of the parietes of this sac, and its apparent liability to rupture 
from the action of the spirit, prevented my inflating it to its 
full extent; but when partially filled with air, it measured 
about 9 inches in circumference below its union with the tail, 
and its greatest diameter, including the slender body to which 
it pertained, was 4 inches. At about one inch below the last 
point of its attachment with the body, the rectum was ob¬ 
served to perforate the sac, the tenuity of which rendered the 
course of that intestine, as indeed that of aL the digestive 
organs readily traced. They are apparently sustained beneath 
the ribs (which latter appear very imperfect), by a membra¬ 
nous expansion, as they are not affected in their position by 
any inflation or emptying of the sac. That the sac itself 
communicates with the pharynx, is sufficiently proved by the 
fact, that if the blowpipe were further introduced, the diges¬ 
tive organs were alone inflated, the sac undergoing no change 
in its dimensions. The nearest point of analogy therefore to 
this structure observable in the class Pisces, is seen among 
the diodons and tetraodons, where a kind of ingluvies, or crop, 
formed of a very thin and extensible membrane, adheres 
closely to the peritonaeum throughout the whole extent of the 
abdomen, by means of which their curious and rapid power of 
inflation has long excited surprise, though we cannot in the 
smooth Ophiognathus so satisfactorily account for its use, as 
among these fishes, which by its distention, mechanically 
elevate their spines, and thus float about in safety. This 
external sac might again be compared with the more internal 
air vesicle of most other kinds of fishes, and especially that 
