338 
MR. W. CROOKES ON EEPULSION RESULTING EEOM RADIATION. 
Again, the variations in the luminosity of a “ standard ” candle will cease to be of 
importance. Any candle may be taken, and if it be placed at such a distance from the 
bar as to give a uniform deflection (say 100 millims.), the standard can be reproduced at 
any subsequent time ; and the burning of the candle may be tested during the photo- 
metric experiments by taking the deflection it causes from time to time, and altering its 
distance, if needed, to keep the index at 100 millims. 
141. When a strong light is brought near this apparatus the bar receives an impulse 
which, unless the magnetic control is very strong, spins it round and round several times. 
If two strong lights are presented to it on opposite sides the bar oscillates rapidly from 
one to the other. As the lights are withdrawn to a greater distance the oscillations get 
smaller, until the bar settles down to a fixed position, dependent on the relative inten- 
sities of the lights shining on it. 
142. Another instrument was constructed like the one last described (135), but the 
pith bar was blacked on alternate halves, instead of having the same half blacked on 
each side. By this construction an impetus given to the bar by a beam of radiation 
would always act in the same direction of movement, the right half of the pith surface 
presented to the light being always black, and the left half of the pith always white, so 
that, if the impulse were strong enough to carry the bar beyond the dead centre, con- 
tinual rotation would be produced. Experiment fully confirmed this supposition. 
When even imperfectly exhausted, the suspended bar rotated when a candle was brought 
near it ; and after more complete exhaustion it spun round rapidly, under the influence of 
radiation, so that the suspending fibre was twisted up, and ultimately stopped the move- 
ment by the accumulated torsion. 
143. Were the black and white surfaces mounted on a pivot, like a compass-needle, 
instead of being suspended on a silk fibre, the movement would not be stopped by 
torsion. The friction, however, would possibly interfere. To test this an apparatus 
was fitted up, as shown in fig. 5. a is a glass vessel open at the 
top, and attached to a hollow glass stem ( b ), which is sealed up at 
the lower end. At the side of b a tube ( c ) is attached, which is 
connected by the glass spiral to the mercury-pump. To the 
hollow stem a piece of glass tube (d) is cemented by fusion so as to 
remain fixed in the position shown. The upper part of d is 
drawn somewhat narrow before the blowpipe, and in it is cemented 
a small cup-shaped ruby. The top of the vessel a is ground 
quite flat, and a ground-glass cover (f) can be cemented on (83). 
The movable part of the apparatus is shown at e ; it consists of a i 
fine curved brass wire with a needle-point soldered to the centre, Ijj 
and having a very thin disk of pith, half an inch in diameter, 
cemented on to each end. Each disk of pith is lampblacked on one side and plain 
white on the other, and they are fastened on so that one black and one white surface is 
always visible. The movable arms are balanced so that they turn easily to the slightest 
Fig. 5. 
SI 
