15 
rendering the turnip-leaves very watery, which soon 
kills them. 
We clo not clearly know in what way sudden or violent 
rainfall acts on the larva); after looking at all the 
observations accessible on the subject it appears to me 
to act in many obvious ways at once. The heavy rains 
fairly beat many of the caterpillars from the plants, and 
their moving and breathing apparatus are so clogged 
with wet and dirt, that some are suffocated and some 
cannot stir until all the supplies within are exhausted, 
and they are starved to death. We know also that 
unsuitable food acts immediately on larval life ; therefore 
it is likely that, if the caterpillars are suddenly pre¬ 
sented with a diet changed by rain to watery sap, and 
also accompanied by water lying on the leaves, that 
they should suffer ; but it does not appear that this can 
be wholly the cause of the sudden sickness, for we do 
not hear of it (as far as I am aware) similarly affecting 
caterpillars feeding below ground, or in the hearts of the 
cabbage, unless they are fairly drowned out. 
The peculiar arrangement of the breathing apparatus 
of insects renders some of the kinds very liable to being 
destroyed by any means, such as dust, mud, water, or 
other substance that may choke the pores. The opera¬ 
tion of breathing is carried on by insects not by means 
of a mouth and lungs, but by means of small openings 
called spiracles (from the word spiro, to breathe), and for 
the most part arranged along the sides. Each spiracle 
forms the opening to a single air-tube or trachea, and 
from these branched tubes wander through every part of 
the insect, conveying air to the limbs, and also to act on 
the visceral fluid (commonly called blood) which fills 
the cavity of the body. In the perfect stage of the 
insect most of the air enters the insect through the 
air-pores in or near the thorax, and therefore the pinch 
beneath the wings, which we know to be a simple way 
of killing our specimens, is considered to do so by suffo¬ 
cation. 
In the larval or chrysalid stage breathing is stated to 
