MR. W. CROOKES OK REPULSION RESULTING FROM RADIATION. 
247 
disturbances caused by change of temperature. When the disks have to be changed, 
air having been let in through the pump, access is easily obtained to the glass cap p 
(fig. 1 ), and the cement being softened by heat, and the cap removed, the disks are 
lifted out together by seizing the aluminium stirrup with forceps. A fresh set of disks 
being introduced, the apparatus is again packed up, and re-exhausted. 
A lamp at f throws a narrow beam of light on the mirror of the apparatus, through 
the aperture d. The ray is reflected to the scale g, where its deflection from zero 
shows the angular movement of the torsion beam when one of the disks is repelled by 
radiation. The scale is 1-| metre from the reflecting mirror. 
A standard candle (109) is supported on a heavy stand, h, and can be raised or 
lowered by means of the sliding piece, i. Another sliding piece, j, carries a pointed 
wire projecting from it. The upright rod of the stand is graduated and numbered, so 
that when the sliding piece j is at mark 1, the point of the wire is on the prolongation 
of the axis of tube and disk No. 1, and so on. Then by sliding the candle up till the 
most luminous part of the flame is level with the point of the wire, it is known that 
the light will shine full on the disk under experiment. A half cylinder, k, covered 
with black velvet, protects the candle from draughts. The candle stand, h, slides along 
a straight edge, l m, screwed to the bench, so graduated that by bringing a mark on 
the sliding stand to one of the divisions, it indicates the number of millimetres 
separating the surface of the experimental disk from the centre of the candle flame. 
222. The experimental powders are laid on one surface of mica or pith disks, 
17*6 millims. in diameter, the pith being 1 millim. thick. If unaffected by alcohol, 
the powders are ground up with this liquid in an agate mortar, and then laid on 
somewhat thickly with a camel-hair brush, so as to be certain that the whole surface 
of the disk is well covered. The back of the disk is left plain. If affected by alcohol, 
water is used • but in any case no gum or other adhesive agent is added. When disks 
of mica or thin metal are used, they are punched out with the same tool employed for 
the pith, and care is taken to completely flatten the disk and to remove the burr. 
Where the material for the disks cannot be punched (e. g. charcoal, selenium, rock salt, 
&c.) they are cut or filed to shape. The upper disk, h, in the apparatus is never 
removed ; it consists of a pith disk, coated with lampblack, by painting with a thin 
cream of lampblack and alcohol, and after drying smoked over burning camphor (147). 
The other disks are attached to the flat glass fibre, f g (fig. 1), by a minute piece of 
cement at their backs, and the whole series is lowered into its place, as shown in fig. 2, 
and fixed in a vertical position. The cap p is then cemented on, and the ah is rarefied 
by means of the pump to within one millimetre of a vacuum. It is kept at this 
exhaustion for about twelve hours, when the disks and powders become perfectly dry 
under the influence of the phosphoric anhydride which is in the drying tube of the 
pump (26, 51, 82,' 355)* The lamp and scale are now adjusted, and the luminous 
index brought to zero, if necessary, by moving the control magnet. 
* Proc. Roy. Soc., Nov. 16, 1876, No. 175, p. 306. 
