It is barely twenty years ago since the discovery was made, first in
France by Coulier and Mascart, but more thoroughly worked out by Mr.
John Aitken in 1880. He found that if a jet of steam is admitted into
two large glass receivers,--one filled with ordinary air, the other with
air which has been filtered through cotton wool so as to keep back all
particles of solid matter,--the first will be instantly filled with
condensed vapour in the usual cloudy form, while the other vessel will
remain quite transparent. Another experiment was made, more nearly
reproducing what occurs in nature. Some water was placed in the two
vessels prepared as before. When the water had evaporated sufficiently
to saturate the air the vessels were slightly cooled; a dense cloud was
at once formed in the one while the other remained quite clear. These
experiments, and many others, show that the mere cooling of vapour in
air will not condense it into mist clouds or rain, unless _particles of
solid matter_ are present to form _nuclei_ upon which condensation can
begin. The density of the cloud is proportionate to the number of the
particles; hence the fact that the steam issuing from the safety-valve
or the chimney of a locomotive forms a dense white cloud, shows that the
air is really full of dust particles, most of which are microscopic but
none the less serving as centres of condensation for the vapour. Hence,
if there were no dust in the air, escaping steam would remain invisible;
there would be no cloud in the sky; and the vapour in the atmosphere,
constantly accumulating through evaporation from seas and oceans and
from the earth's surface, would have to find some other means of
returning to its source.

One of these modes would be the deposition of dew, which is itself an
illustration of the principle that vapour requires solid or liquid
surfaces to condense upon; dew forms most readily and abundantly on