376 
adjoin the ventral nerve-cord. The large 
pair (which we may call the pneumatic 
tracheal trunks) suddenly gives out a 
crowd of branches in the abdomen, which 
subdivide again and again, breaking up 
into a spray of fine filaments which supply 
the branchial organs. 
The branchiae consist of a dozen longi- 
tudinal columns around the rectum, each 
bearing about thirty-five pairs of oval 
branchial leaflets, pinnately arranged and 
Each leaflet 
receives about two hundred very delicate 
imbricating over each other. 
filaments from the tracheal branchlets ; the 
filaments are attenuated towards their ex- 
tremity and end caecally in the sac-like 
leaflet. Dufour found (in Aeschna) thate 
the leaflets are enclosed in pockets or invo- 
lutions of the intestinal wall; in Libellula 
they seem to lie loosely in the rectal cavity. 
The larva respires by drawing in through 
the anus a gentle flow of water, which it 
then expels with force, driving the exhaus- 
ted water and contained impurities to a safe 
distance. The inflow and outflow may 
occur about fifty times per minute; and 
occasionally the process will stop for a 
while, especially when the larva is at rest. 
There are sphincter valves at the anus and 
muscles for regulating the opening and 
closing of the anal lobes and armature, 
and a ganglionic enlargement to supply 
this. 
the branchiae and no large ganglia in the 
abdomen to suggest special muscular action. 
The muscles which line the abdominal wall 
regulate the respiration. By relaxing they 
permit the abdomen to expand, causing a 
partial vacuum: then the water flows gen- 
tly in, and the air flows from the pneumatic 
tracheal trunks into the many thousands 
But there are no large muscles about 
PSYCHE. 
of filaments which crowd the branchial 
leaflets. This action may be aided by 
pressure in the front part of the body which 
drives the air to the reat; and the pneu- 
matic pressure in the tracheae swells the 
delicate filaments and the enclosing bran-. 
chial leaflets simultaneously with the in- 
coming tide of water. The fine membrane 
permits the passage of gases, but not of the 
water. The contraction of the abdominal 
muscles by reducing the vacuum expels 
both air and water from the branchial 
region, driving the air forward to the tra- 
cheal system. In a specimen imprisoned 
with water in a glass tube we found that 
four air-bubbles were expelled with the 
outgoing tide of water in as many minutes 
Thus the me- 
chanical principle is nearly the same as in 
lung-respiration. The simultaneous inflow 
of air and of a fluid follows the expansion 
of the body-wall, and the contraction of the 
wall induces a reciprocal outflow. In the 
insect the fluid comes from without, the air 
from within; in the air-breathing verte- 
brate the air comes trom without, and the 
fluid (blood) from the system, being aided 
in its progress by the heart.® Cutting 
through the abdominal wall of the larva 
has the same effect as piercing the human 
(a result of excitement). 
pleura, causing the respiratory organs to 
collapse and stopping their function. 
prevent the expansion of the abdomen 
of the larva, breathing is temporarily 
arrested. 
M. Jousset de Bellesme’ discovered that 
the larval dragon-fly swells out into its 
adult form by a process which is a kind 
_ ® Foster’s Physiology, 1880, 4th edit., book 2, 
chap. 2, § 7. 
7 Harper’s Annual rec. of sci. . . . 1878, p. 447. 
If we 
