should contain more than the slightest trace of water. If they do, the
gelatine subsequently made from them will most certainly exude, and become
dangerous and comparatively valueless. It will also be much more difficult
to make the nitro-cotton dissolve in the nitro-glycerine if either
contains water.

In order to find out how long any sample of cotton requires to be dried, a
sample should be taken from the centre of several boxes, well mixed, and
about 1,000 grms. spread out on a paper tray, weighed, and the whole then
placed in the water oven at 100° C., and dried for an hour or so, and
again weighed, and the percentage of moisture calculated from the loss in
weight. This will be a guide to the time that the cotton will probably
require to be in the drying house. Samples generally contain from 20 to 30
per cent. of water. After drying for a period of forty-eight hours, a
sample should be again dried in the oven at 100° C., and the moisture
determined, and so on at intervals until the bulk of the cotton is found
to be dry, i.e., to contain from 0.25 to 0.5 per cent. of moisture. It is
then ready to be sifted. During the process of removing to the sifting
house and the sifting itself, the cotton should be exposed to the air as
little as possible, as dry nitro-cotton absorbs as much as 2 per cent. of
moisture from the air at ordinary temperatures and average dryness.

The drying house usually consists of a wooden building, the inside of
which is fitted with shelves, or rather framework to contain drawers, made
of wood, with brass or copper wire netting bottoms. A current of hot air
is made to pass through the shelves and over the surface of the cotton,
which is spread out upon them to the depth of about 2 inches. This current
of air can be obtained in any way that may be found convenient, such as by
means of a fan or Root's blower, the air being passed over hot bricks, or
hot-water pipes before entering the building. The cotton should also be