266 MR. W. CROOKES OK REPULSION RESULTING FROM RADIATION. 
needle in the centre of the bulb, and on this rests a glass cap, to which is attached 
four radial arms of aluminium. To these arms disks of mica or pith can be fastened, 
so as to form the moveable fly of a radiometer. The disks can be changed by un- 
cementing the glass top, and lifting the fly out with tweezers. The lower part of the 
Fig. 4. 
Centimetres. 
tube is drawn out for connexion with the mercury pump. The powders used for 
experiment were carefully painted on the opposite sides of pith or mica disks, only 
water or alcohol being used (222). 
245. The first, pair of substances tested in this apparatus were chromic oxide and 
precipitated selenium.'" In a tabular form, these stand as follows (243) : — 
Chromic oxide 
Selenium, precipitated 
Difference 
No screen. 
Water screen. 
71-5 
17 
35-8 
5-8 
35 *7 
-47 
A radiometer therefore made of these powders, on alternate sides of pith disks, 
should rotate in one direction when no screen is interposed between it and the light, 
and in the opposite direction when behind a water screen — the positive rotation taking 
place with a force of 357, and the negative rotation with a force of 47. 
When the apparatus was exhausted up to the point of maximum sensitiveness, a 
glass cell full of water was placed in front of it and a lighted candle brought near. 
Negative rotation immediately commenced, the selenium being repelled. On drawing 
away the screen the rotation stopped, and positive rotation commenced, which could 
* The selenium was precipitated as a brilliant scarlet powder, by boiling commercial selenium in fine 
powder, in solution, of cyanide of potassium, and then supersaturating with hydrochloric acid. Seleno- 
eyanide of potassium is formed, and I showed in 1851 that this salt is decomposed by almost any acid, 
Avith precipitation of selenium. (Chem. Soc. Journ., iv., 1852, p. 13.) 
