442 
Proceedings of the Royal Society 
thinner, because it increases the rate of the descending current, and 
carries away the purified air more quickly. 
A form of apparatus was arranged to get rid of this separating 
effect of gravitation. It consisted of an extremely thin and flat 
piece of metal. This test-surface was placed vertically in the 
dust-box. The air in passing over this piece of metal did not take 
up a horizontal movement at any part of its passage. The result 
was that even with a temperature - 10° C. the dust kept close to 
its surface, and no dark plane was formed in the descending current. 
The dark plane in the cold descending current seems, therefore, not 
to be an effect of temperature, but is the result of the action of gravi- 
tation on the particles under the body. A dark plane was, however, 
observed when working with this flat surface, when cooled \ but it 
was not formed in dusty, but in foggy air, and was found to be due 
to the evaporation of the fog particles when they approached the 
cold surface. 
If a very little heat, instead of cold, as in the previous experiment, 
is applied to the round tube, then the dark space under the tube rises 
and encircles the tube, and the two currents of clear air unite over 
the tube, and form the dark plane in the upward current. But in 
addition to this, heat has been found to evert a repelling effect on 
the dust. This was proved by putting the thin vertical test-surface 
in the dust-box, and heating it ; when it was found that the dust 
was repelled from its surface, and a dark plane formed in the 
ascending current ; neither of which effects was obtained with cold. 
The dust begins to be repelled with the slightest rise of temperature, 
and the dark space in front of the test-surface becomes thicker as 
the temperature rises. An experiment is then described in which 
the dust particles in the air flowing up between two parallel glass 
plates is caused to pass from side to side of the channel by the 
repelling action of heat at different points. 
For testing the effects of higher temperatures a platinum wire 
lieated by means of a battery was used. The platinum wire was 
bent into a U-shape, the two legs being brought close together. 
This wire was fixed in the dust-box with the bend to the front, and 
the legs in the same horizontal plane, the two copper wires to 
which it was attaclied being carried backwards and out of the box. 
