Larva of Chaoborus Grystallinus ( de Geer). 
367 
which swell on taking up water, the swelling probably being 
assisted by a hypothetical substance discharged into the blood. 
Assuming the gas is secreted by the cells as suggested, its 
collection and distribution would follow on the lines already 
indicated. 
With the descent of the larva, the sacs being non-compressible, 
diminution of pressure inside the sacs follows, due to the difference 
of hydrostatic pressure between the inside of the sac and the 
surrounding fluid. With these conditions an immediate equaliza- 
tion would take place by the available gas passing from the tubes 
through the one-way valve into the sac. 
On ascending we have a reversal of conditions — an increased 
pressure inside the sac, and diminished pressure outside. The 
natural outlet for the gas would be through the tracheae attached 
to the sac. The two actions, respiration and stabilization, synchro- 
nize. 
We have satisfied ourselves there is a proper connexion between 
the sac and capillary tracheae by discharging the entire air contents 
of the sac through a short length of tracheae still intact. We are 
also of the opinion there exists some check action to prevent air 
being rushed out of the sac by sudden excessive pressure. 
Kepeated experiments on air sacs in a live box, to which 
pressure can be applied, has disclosed the fact that the contained 
air, when the sac is gently pressed, will issue forth from the 
tracheal tube end in a series of intermittent bubbles, an appreciable 
period elapsing between their appearances. 
On the other hand, if there was no check we would expect the 
air to issue forth in a continuous stream. 
It has been observed that no air escapes from the broken end 
of the tracheal tube attached to the sac, when the latter is under 
observation in a live box, without pressure. 
Here we have a condition where pressure inside and outside 
the sac has been equalized, and the contained air would, therefore, 
remain in the sac and trachea. 
We have referred to the valve as a one-way valve. It is con- 
ceivable the small circular bodies in the tube passing through its 
centre, as already described, act as the round glass stoppers used 
in mineral water bottles, allowing the fluid and gas to pass in, but 
resisting any escape. The entire body of the knob-shaped 
appendage is full of gas (fig. 27), being in direct communication 
with the sac, and arising from its neck is an outlet capillary 
trachea, part of which spreads over the surface of the sac itself. 
Correctly termed, this knob-shaped appendage, or valve, acts as an 
intake as well as an outlet valve. 
About the period of completion of the transition stage from 
larva to pupa, the main tracheal tubes of the pupa can be seen 
forming at the valve end of the sac ; the now shrivelled tubes are 
