278 
MR. W. CROOKES ON REPULSION RESULTING FROM RADIATION- 
Artificial and natural 
tourmalines. 
Parallel. 
CrpssecJ.. 
Group 1 
40- 
46- 
2 
as- 
36. 
„ 3 
se- 
30- 
Mean . , 
ss- 
37 r 3 
Here the difference is in favour of the parallel tourmalines, but it is so slight that, 
like the former result, I do not consider it as the expression of a natural law. 
262. The idea when these experiments were first tried was that a shoe of tourma- 
line, being black to a ray of light polarised in one plane, and white to a ray polarised 
in the other plane, would be repelled when the incident light was quenched by it, and 
not affected when the incident light passed through it. The idea was a perfectly 
reasonable one some years ago ; but recent researches have proved that the repulsion 
resulting from radiation is almost entirely a surface action, whilst the action of a 
tourmaline on a ray of polarised light is one in which thickness is necessary. 
The above experiments prove that the special action originally thought possible 
does not exist in a degree appreciable with the present experimental means. 
263. That a polarised ray of light can produce repulsion in a vacuum as well as 
a ray of common light is proved by the fact that radiometers move readily when 
exposed to candle light which has passed through Mr. Spottiswoode’s large prism. 
A torsion balance with black pith at one end was also exposed to candle light, two 
tourmalines being interposed between the candle and the pith. When the tourma- 
lines were crossed there was no repulsion, but when parallel the pith was driven back 
against the side of the glass tube, 
EFFECT OF SHAPE IN INFLUENCING THE AMOUNT AND DIRECTION 
OF REPULSION. 
264. In a preliminary note on the theory of the radiometer, which I had the honour 
of experimentally illustrating at the meeting of the Royal Society, November 16th, 
1876,* I described an experiment in which the vanes of a radiometer were made of 
gold leaf, lampblacked on one side. The apparatus in which this .and most of the 
succeeding experiments in this division were tried, is the one shown at fig. 4, par. 244. 
Through the open top, access can readily be obtained, and disks, plates, &c., can be 
quickly tested by being fixed to the extremities of a pair of aluminium arms with 
a glass cap in the centre, rotating on the needle point. On submitting the pair of 
blacked gold leaf disks to experiment in this apparatus, after exhaustion, I found an 
anomaly in the action. A candle repelled the black surface of one of the disks, but it 
seemed to repel with almost equal energy the metallic surface of the other disk ; one 
* Proc. Roy. Soc., No. 175, 1876, vol. xxv. p. 304. 
