1889 - 90 .] 
Mr J . Aitken on Dust Particles. 
223 
the cell was surrounded by an uncooled part of the plate, and any 
change in the appearance of the cooled dust could be easily detected 
by contrast with the surrounding area. 
The first thing to be done was to test the working of the 
apparatus, and see how it acted with dusts of known composition 
and known condensing power. For this purpose smoke of burning 
magnesium, gunpowder, and sodium were used — the first on 
account of its small affinity for water vapour, the second for its 
condensing power. It is well known that the smoke from gun- 
powder is far more dense in damp than in dry weather, the difference 
being due no doubt to the condensation which takes place on the 
smoke when the air is damp. While sodium smoke was selected 
on account of its great affinity for water. The smoke of these sub- 
stances was produced by burning a little of them in an enclosed 
space, the test plates were placed in the enclosure, and the smoke 
allowed to fall on them. The mirrors with the deposited dusts 
were then tested, and their condensing powers measured, with the 
following result : — Magnesia was found to condense at almost 
exactly the same temperature as the glass, but gunpowder smoke 
began to show signs of condensing at a temperature 5° above the 
dew-point ; while the soda condensed vapour from air at a tempera- 
ture 17° above its dew-point. In making these tests it is necessary 
that the air in the room, where the testing is done, be very dry, 
otherwise the beginning of the condensation on the dust cannot be 
detected, because the dust surrounding the cooled surface already 
has some moisture condensed on it. To overcome this difficulty it 
was customary to heat and dry the plates before testing them. 
The different kinds of dust, when tested in this way, having thus 
shown a distinct difference in their affinities for vapour, tests were 
now made of the dust collected from the atmosphere. It is un- 
necessary to go into the detail of these experiments, as although 
they distinctly point to atmospheric dust having an affinity for 
water vapour, and this affinity seemed to vary with the dust, yet the 
method of testing is not very accurate. As the condensation begins 
by imperceptible degrees, it is very much a question of quickness 
of perception, and carefulness of working, that determines when the 
first appearance of condensation will be detected ; and even the first 
appearance is not the real beginning, but only the state at which it 
