441 
oj Edinhurglb, Session 1883 - 84 . 
The experiments described in this paper were made in a small 
dust-box, blackened inside, glazed in front, and provided with a 
window at one side. For illumination two jets of gas enclosed in a 
dark lantern were used. The light entered the dust-box by the side 
window, and could be condensed on any part of the inside of the 
box, by means of two lenses fixed in a short tube, and loosely 
attached to the front of the lantern. Magnifying glasses of different 
powers were used for observation. The dusts experimented on were 
made, some of hydrochloric acid and ammonia ; some by burning 
sulphur and adding ammonia ; others by burning paper, magnesium, 
or sodium. Calcined magnesia and lime were also used, as well as 
ground charcoal. These three last substances were stirred up by 
means of a jet of air. 
For testing the effects of slight differences of temperature, tubes in 
some form or other were generally used. These tubes were closed 
at the front, projected through the back of the dust-box, and 
were brought close to the glass front, for observation under strong 
magnifying power. The tubes were heated or cooled by circu- 
lating water through them in a small tube passing through their 
interior. 
Suppose the experiments to be begun by introducing a round 
tube into its place in the dust-box, and filling the box with any 
dust, everything being then left for some time, so that all the 
apparatus may acquire the same temperature. If the light be now 
allowed to fall on the box, and be quickly brought to a focus on the 
tube, it will be found that the dust is in close contact with it, on 
the top and sides, but underneath a clear space will be observed i 
close examination will show that the particles are falling on the upper 
surface of the tube, and coming into contact with it, while under- 
neath a clear space is formed by the particles falling away from it. 
If the tube is now slightly cooled, a downward current is formed, and 
the currents of dustless air from below the tube meet under it, and 
form a dark plane in the centre of the descending current. It is 
shown that gravitation can, under favourable conditions, produce 
this separation of the dust quickly enough to keep up a constant 
supply of dustless air. Ho increase of effect is produced by a lower 
temperature. A temperature of - 10° C. makes the dark plane 
