and the light
revealing the dust. Cork the flask, stuff its neck with cotton-wool,
or simply turn it mouth downwards and leave it undisturbed for a day
or two. Examined afterwards with the luminous beam, no track is
visible; the light passes through the flask as through a vacuum. The
floating matter has abolished itself, being now attached to the
interior surface of the flask.

Were it our object, as it will be subsequently, to effectually detain
the dirt, we might coat that surface with some sticky substance. Here,
then, without 'torturing' the air in any way, we have found a means of
ridding it, or rather of enabling it to rid itself, of floating
matter.

We have now to devise a means of testing the action of such
spontaneously purified air upon putrescible infusions. Wooden
chambers, or cases, are accordingly constructed, having glass fronts,
side-windows, and back-doors. Through the bottoms of the chambers
test-tubes pass air-tight; their open ends, for about one-fifth of the
length of the tubes, being within the chambers. Provision is made for
a free connection rough sinuous channels between the inner and the
outer air. Through such channels, though open, no dust will reach the
chamber. The top of each chamber is perforated by a circular hole two
inches in diameter, closed air-tight by a sheet of India-rubber. This
is pierced in the middle by a pin, and through the pin-hole is pushed
the shank of a long pipette, ending above in a small funnel. The
shank also passes through a stuffing-box of cotton-wool moistened with
glycerine; so that, tightly clasped by the rubber and wool, the
pipette is not likely in its motions up and down to carry any dust
into the chamber. The annexed woodcut shows a chamber, with six
test-tubes, its side-windows w w, its pipette p c, and its sinuous
channels a b which connect the air of the chamber with the outer air.

The chamber is carefully closed and permitted to remain quiet for two
or three days. Examined at the beginning by a beam sent through its
windows, the air is found laden with floating matter, which in three
days has wholly disappeared. To prevent its ever rising again, the
internal surface of the chamber was at the outset coated with
glycerine. The fresh but putrescible liquid is introduced into the
six tubes in succession by means of the pipette. Permitted to remain
without further precaution, every one of the tubes would putrefy and
fill itself with life.