366 
MB, W. CEOOKES ON EEPULSION EESULTING EEOM EADIATION. 
black by holding it over the flame of burning turpentine. The movable disk may be 
of any substance. 
The apparatus was fitted up in a recess built of brick, and closed in front with a 
glass window. In the brick wall at the side j holes were pierced opposite the disks 
and the central mirror. These were lined with card tubes ( k ) blacked inside. The 
interstices were packed with cotton-wool, and the apparatus was closely surrounded 
with Winchester quart-bottles filled with water. In front of the card tubes wooden 
shutters (l) were fastened, so that either one could be opened independently of the 
other. The apparatus was sealed on to the pump by means of the arm shown at the 
upper part of the torsion-apparatus. 
With nothing in the pan, and only the standard black disk on, the beam is in equi- 
librium ; consequently each new disk requires a special counterpoise. 
199. The following weights and measurements of the different parts of the apparatus 
were taken : — 
Weight of glass beam, central stirrup, and mirror ..... 3-0485 grains. 
Standard black disk and aluminium bar 0-4635 grain. 
Pan at other end of glass beam, acting as counterpoise for above. 0-5263 „ 
Weight of white pith disk and support 0-2670 ,, 
Counterpoise for ditto 0-3085 ,, 
Length of arm from centre of support to centre of disk ... 99 millims. 
Length of arm from centre of support to centre of counterpoising 
pan 84 „ 
Diameter of disks 17‘25 „ 
Torsion of suspending fibre, in air, with glass rod (186) hanging 
to it ^ oscillation in 15-75 seconds. 
200. The experiments were tried as follows : — The exhaustion having been carried 
to the utmost point, so that no increase in sensitiveness was produced by further 
working the pump, a standard candle was adjusted opposite the black disk and at a 
definite number of millimetres off. The deflection of the index ray of light was then 
taken on opening the shutter. After completing this observation, the beam was allowed 
to come to rest ; the candle was then lowered till it was opposite the lower disk, and 
the deflection caused by it was again taken. These were repeated several times. The 
mean results of the first series are given in the following Table (the lower disk being 
plain white pith) : — 
