ME. W. CEOOKES ON REPULSION RESULTING EEOM RADIATION. 
353 
179. The battery was disconnected from the instrument, and one end of a wire was 
attached to one of the platinum loops, d, the other end of the wire being connected 
to the prime conductor of a frictional electrical machine. A few turns of the handle 
sent the arms flying about wildly ; sometimes they spun round violently in one direction, 
then they stopped and went round the other way, finally one pointed steadily to the 
platinum spiral and refused to move. A candle was brought near, and all means were 
tried to discharge the disk, but with no effect. When the candle was quite close it 
overcame the interference, and the disks revolved in an irregular jerky manner. 
The spiral was ignited by a battery, in the hope that this would discharge the elec- 
tricity, but with no avail, and there was nothing to be done but stop the experiments 
and put the apparatus in water. 
In three or four days the electrical disturbance was sufficiently diminished to enable 
me to proceed experimenting ; but I could detect the influence for weeks after. 
180. One pole of a small induction-coil capable of giving half-inch sparks in air was 
fastened to the platinum loops d, the other pole being held by an insulating handle. 
The loose pole was then brought near the bulb. The nearest disk rushed round to it 
and followed it a little, then it stuck as if the glass were electrified. By gently moving 
the loose pole round I could get the arms to rotate in either direction with a little 
practice, and they would keep on for five minutes or more when once started. It seemed 
a matter of indifference whether the black or the white surface went first. The results 
with the induction-coil were only a little more under control than those with the friction 
machine. The movements appear all to be explained by the known laws of static elec- 
tricity ; the rotations have no connexion with the instruments under the influence of 
radiation, but are of the “ electrical fly ” kind (34, 35, 36). 
181. Before leaving the subject of the radiometer, it may be of interest if I describe 
a few forms of this instrument which I have made for special purposes. 
It is easy to get rotation in a radiometer without having the surfaces of the disks 
differently coloured. A radiometer was made similar to the one described in par. 152, 
but somewhat larger, and having the pith disks lampblacked on both sides. Its weight 
was 1-25 grain. It was exhausted with a charcoal-tube attached. When it was 
exhausted and a candle was brought near it, the arms moved until two of the disks 
were equidistant from the flame, and no amount of initial impulse in one or the other 
direction would set it in rotation. A piece of ice caused it to move until one disk 
pointed to the ice, when it also stopped. By shading the candle with a screen, so that 
the light shone on only one half of the fly, rapid rotation commenced, which was 
instantly stopped, and changed into as rapid rotation in the opposite direction, by 
altering the position of the screen to the other side. 
182. It is difficult to exhibit the movement of a radiometer to any large number of 
people at once. To enable me to do this I have made an instrument, the disks of 
which are thin glass, silvered and polished on one side, and coated with lampblack on 
the other. This, owing to its great weight (65 grains), is somewhat slow; but in the 
