88 
ON THE BEEATHING APPAEATUS 
the next moult. A similarly progressive change may he noticed 
in the larva of Corethra plumicornis, the well-known glass larva. 
In the larval condition there is no special breathing apparatus, 
unless the beautiful plumose tail, or the bundles of setae, are in 
some way used for that purpose. 
It seems, however, far more probable that aeration takes 
place through the skin itself, or the joints, as in the foregoing 
instance of Cloeon. There are four kidney-shaped air bladders, 
arranged two at the tail and two near the head of the larva, 
which are of a tracheal nature, being lined with the ordinary 
spirally arranged hair ; but there is no tracheal tube connecting 
either of these four with each other, each appearing to be 
independent of the other, and being probably for the purpose of 
preserving a balance, and enabling the animal, which has no 
legs, to maintain a horizontal position when suspended in the water. 
When, however, the larva has changed its skin for the last 
time, before changing to the pupa state, a definite tracheal 
system may be seen running the whole length of the body, and 
in front and above the two thoracic air sacs, are two spindle- 
shaped bodies ; but as yet none of these newly developed organs 
contain air. 
After the next moult has taken place the tracheae are found 
full of air, and the kidney-shaped air sacs have disappeared, 
whilst the two spindle-shaped bodies, above mentioned, having 
been liberated from their confined position, are filled with air, 
and stand out straight above the pupa, which now assumes 
an upright position. These floats are not filled by rising to 
the surface and drawing in the external free air, as is the case 
with gnat pupae, but from the trachea, the air absorbed from the 
water first passing through the body. 
I can speak with certainty on this point, as I have seen the 
larva change to a pupa, and the floats gradually filled, without 
its once rising to the surface. 
