MR. W. CROOKES ON REPULSION RESULTING FROM RADIATION. 
313 
At 1 00 millims. pressure the ascending current of hot air is not strong enough to 
turn the disk round, and it is therefore carried along with the vanes by the viscosity 
of the air, which is practically the same as at 760 millims. 
At a pressure of 10 millims. the vanes become still, but the disk rotates very slowly 
negatively. 
At 1 millim. pressure, the disk still rotates slowly negatively, whilst the vanes will 
not move. 
365. Pressure 824 millionths of an atmosphere. — When the current is first turned 
on, the vanes give half a revolution in the negative direction, / , they then stop. 
The disk continues to rotate in the negative direction, CUPP . An observation of 
viscosity at this pressure gives the figure 0T06‘. The repulsion produced by the 
candle on the viscosity plate shows a force of 4 ’8. 
It will be observed that the first movements of the vanes at this pressure is in the 
negative direction, but this soon stops. On referring to par. 348, where I describe 
an experiment on a similar apparatus, but where the platinum wire is uncovered, 
and the heat therefore more intense, it will be seen that the negative tendency is 
strong enough to cause continuous rotation of the vanes. This negative rotation 
of the vanes begins to be apparent at 14 millims. pressure ; it is very strong at 
760 millionths of an atmosphere (350), and disappears at about 140 millionths of 
an atmosphere (351). Between a pressure of 294 millionths of an atmosphere and 
129 millionths of an atmosphere, there is a great change in the movement of the 
vanes when under the influence of the hot, naked, platinum ring, the rotation 
changing from 12 5 revolutions a minute negatively at the former pressure, to 100 
revolutions a minute positively at the latter pressure. How narrow are these limits 
may best be seen by converting them into decimals of a millimetre : the negative 
rotation is good at 0'2234 millim., whilst the positive rotation is very strong at 
0‘ 09 7 8 m illim . The negative rotation between the above-named limits is not so 
apparent in the apparatus where the platinum wire is covered with mica (345) as 
when the naked wire is used. This behaviour offers an anomaly which I shall 
endeavour to clear up at a future time. As I do not believe in attraction, and con- 
sider that all these movements are caused by a force having the action of a push, 
many experiments now in progress will be required to explain the difficulties of this 
negative action. 
366. Pressure 530 millionths of an atmosphere. — A little negative rotation of the 
vanes takes place on first making contact, and they then become stationary. The 
disk rotates at a speed of one revolution a minute in the negative direction, CSS*. The 
viscosity of the air is 0T04, and the candle repulsion = 7'1. 
367. Pressure 470 millionths of an atmosphere. — The vanes remain stationary. On 
first making contact the disk makes 1-| revolution in the positive direction, and 
then rotates continuously in the negative direction, at the rate of one-third of a 
revolution a minute. Viscosity = ‘102. Candle repulsion = 8‘2. 
2 s 
MDCCCLXXVIII. 
