354 
Transactions of the Society . 
hoops have buckled into a number of apparently short rods tilted 
at different angles (PI. XIX., fig. 25). The band of tissue at the 
fracture appears as a collar ; if the fracture is bad it suggests an 
elbow joint. This buckling is due to the inability of the hoops to 
expand. An increase in depth is especially noticeable in the 
central hoops, which gradually lessens as those occupying the 
extremities are approached. (PI. XVIII., fig. 21.) 
This formation of the hoops produces a curious effect when 
viewing the sac as a whole ; the impression given is that of a tube 
or cord running round the periphery of the sac. (PI. XVII., 
fig. 20.) 
The hoops are entirely covered with a transparent elastic coat, 
the outer surface of which is slightly corrugated. This transparent 
tissue also passes between each hoop, holding them together. 
(PI. XIX., fig. 26). The longitudinal sections show quite an 
appreciable thickening of this elastic coat on the middle portion 
only of the ventral side. The air sac has only two coats, the outer 
one being of a gelatinous nature which carries the pigment spots, 
and an inner transparent one strengthened by the embedded hoops. 
Viewing the larva with the head to the right, we find on the 
left-hand end of the thoracic, and on the right-hand end of the 
posterior sacs, there is attached a knobbed shaped appendage con- 
stricted towards the top and tapering away to a very fine point, 
from which issues an exceedingly delicate capillary tube. This 
latter, with the whole of the appendage, is enclosed in a large 
transparent tube. (PL XVIII., fig. 17.) At its greatest circum- 
ference, we find out that the appendage has a circle of roundish 
patches of a different structure to that of itself (fig. 33). From 
its neck there is given off a capillary trachea, a branch from 
which is carried over the air sac. The other end of the sac is an 
exit, and to this is attached capillary trachea. If both ends of 
the sacs are exits only , why this important difference in structure ? 
The appendage, which we will refer to as a valve, is filled with 
air, and when under examination, surrounded by water, the optical 
conditions are not all favourable for observing delicate structure. 
We have, however, made out the continuation of the capillary 
tube referred to as passing through the centre of the valve, and 
then no doubt into the lumen of the air sac. At intervals along 
the tube inside the valve appear highly refractive circular bodies 
(fig. 27). We will refer to this structure again. 
A great deal has been written concerning the question of the 
ability of the Corethra to stabilize itself under various pressures, 
and it will be of interest at this stage to refer to W. N. F. Wood- 
land’s paper {11) on “The Structure and Function of the Gas 
Glands associated with the Gas Bladder of some of the Teleostean 
Fishes,” which will give us an opportunity of comparing the 
methods of these fishes in maintaining hydrostatic equilibrium 
