260 
MR W. CROOKES ON REPULSION RESULTING FROM RADIATION. 
The complicated nature of these actions is well shown by an examination of the 
results given in the first three lines. Lampblack on a pith disk being 100, plain 
white pith without any lampblack is 17 ‘7. But when lampblack is applied to the 
back of the pith, and the plain white surface is again exposed to radiation, the action 
sinks to zero. The repulsion exerted on the white pith must be the same in each 
case, but the presence of lampblack behind it has caused an absorption of heat by the 
back surface, producing molecular pressure just sufficient to neutralise the pressure 
in front. It is not sufficient therefore to take into account the kind of powder 
exposed to radiation ; its conducting power for heat, as well as that of the disk 
which serves to support it, and the absorbing and radiating powers of the back surface 
for heat, together with its distance from the bounding surface of the glass vessel, all 
have to be taken into consideration. For these reasons too much weight must not 
be attached to the results given in the foregoing tables. They convey a fair general 
idea of the variations in the power according to the colour and physical condition 
of the substances exposed to radiation, and considering the enormous difficulty 
in getting results which have any approach to quantitative accuracy, they are very 
satisfactory ; but it must be remembered that they are only comparable one with 
another under the exact conditions in which they were tried, and it is not unlikely 
that a repetition of the series, with an alteration in some apparently unimportant 
item, such as an increased diameter of glass tube, a thicker disk of pith, or the 
substitution for pith of a substance which will better conduct heat, might give a 
different series of numbers. The proportion between them, however, would probably 
remain the same as at present. 
In an ordinary radiometer, such as I described in my last paper on this subject, 
(145), the disks are of pith, lampblacked on one side. The present results show that 
the whole of the force of radiation exerted on the black surface is utilised in turning 
the fly round. 
238. The experiments were tried with mica to test a suggestion by Professor Stokes, 
that roasted mica, lampblacked on one side, would combine in a pre-eminent degree the 
conditions needed for the fly of a radiometer, namely, — • 
1. Good absorption on one face. 
2. Good reflection on the other, for heat as well as light. 
3. Bad conduction. 
4. Absence of organic matter. 
5. Extreme lightness. 
I had already used mica both plain and roasted, but had never tested them 
systematically. The “ silver flake ” mica is a natural mineral occurring in the United 
States. It is of a brilliant silvery lustre, and splits into exceedingly thin fibres, 
which are as coherent and tough as ordinary mica, whilst they have the bright 
reflecting surface of mica artificially calcined. The mineral is probably produced 
