90 
MR, W. CROOKES OK REPULSION RESULTING FROM RADIATION. 
of the clear disks and fall on the black, the black disk immediately retreats, the clear 
disk remaining stationary for a moment, and then approaching the light. As soon as 
the disks have swung apart their full distance they keep there, and revolve like the 
fly of an ordinary four-vaned radiometer. The reason why the clear disk does not 
immediately move towards the light in proportion to the movement of the black disk 
away from the light, appears to be that the slight friction between the upper needle 
point and cup opposes, at first, such movement. 
Another explanation has frequently occurred to me, which, however, I give with 
diffidence. This is, that when radiation first falls on a black surface the repulsion is 
produced, not by the chain of molecular impacts between the black surface and the 
solid body in face of it, but by the reaction between the black surface and the layer 
of normally-moving gaseous molecules near it, these forming a resisting cushion for 
the extra active molecules to react against and press back the black surface. This 
cushion of inactive molecules is, as it were, rapidly eaten into by the active molecules, 
until in a very short space of time the pressure spans across from one surface to 
the other. 
If the candle is allowed to shine on the plain side of the black disk no movement 
takes place at first. Very soon the disks both move in the same direction away from 
the candle, the speed of the clear disk gradually increasing over that of the blacked 
disk. 
392. In the instrument just described, the two pairs of disks were freely pivoted on 
separate points. Another radiometer was now made, in which the plain and blacked 
disks could be fixed beforehand in any relative positions. Fig. 4 shows the instru- 
Fig. 4. 
merit; the arms carrying the 1 lack disks, a a, are supported on a cup revolving on a 
needle point; the arms carrying the clear disks, b b, have a circular hole in the 
centre, which fits over the glass cap. The relative positions of the arms can easily be 
adjusted by inverting the radiometer and tapping it; when replaced in position they 
are held together lay friction, and the fly moves as a whole. 
Experiments with this radiometer fully confirm the results obtained with fixed 
screens. In position A, with the two plates touching, the black surface being inside, 
