MR. W. CROOKES OK REPULSION RESULTING FROM RADIATION. 
267 
be changed to negative by interposing the water screen. The positive rotation was 
much the stronger of the two. 
Exposed to the candle radiation behind a screen of black talc, perfectly opaque but 
diathermous, positive rotation ensued. 
The candle was removed, and a hot glass shade* put over the bulb and fly. 
Negative rotation took place ; on removing the hot shade and letting the instrument 
cool, there was no reversal of rotation (161, 165, 167, 168). 
246. The fly was then removed from the apparatus, and the pith disks replaced by 
others coated on opposite sides with the following powders : — 
No screen. 
Water screen. 
Chromic oxide .... 
. . 71-5 
1-7 
Copper tungstate .... 
. . 51-2 
6'4 
+ 20-3 
-4-7 
When properly exhausted, the instrument was exposed to candle light behind a 
water screen, as in the previous experiment ; quick negative rotation took place, the 
copper tungstate being repelled, Whilst the fly was in full rotation, the water screen 
was drawn away, the candle still remaining ; the negative rotation got slacker and 
slacker till the fly just moved round, but it did not stop entirely, or revolve positively, 
as it should have done according to the indications of the above table. By holding a 
second candle, temporarily, close to the chromium side of a disk, the movement 
could be stopped, but on taking away the second candle the slow negative rotation 
recommenced. So far, the results only half agree with theory : the negative rotation 
behind water is correct, but the positive rotation to the direct light is not seen. The 
candle was then removed, and the fly allowed to come to rest. When quite still and 
cold, the lighted candle was again brought up to the apparatus, when the fly at once 
moved, the chromic oxide being this time repelled, and positive rotation, in accordance 
with theory, keeping up for several revolutions. After a little time, however, the 
positive rotation slackened, then stopped altogether, and finally gave place to a very 
slow negative rotation. 
From these experiments it appears that the superior repulsion, exerted by radiation 
from a naked candle, on chromic oxide over copper tungstate, is due to the former 
being more sensitive to ultra-red heat rays : this, in fact, is in accordance with the 
results behind water. When dark heat rays are thus filtered off, the copper tungstate 
is four times more sensitive to the residual luminous rays than is the chromic oxide. 
But the peculiar behaviour in the last experiment carries the proof a step further. 
The candle, shining on the glass bulb, gradually raises its temperature, and the glass 
begins to radiate dark heat in all directions, on the chromic oxide and copper tungstate 
* A hot glass shade is used as a convenient means of heating the bulb, by immersing it in a hot air bath, 
without the liability of introducing action of rays other than those emitted by hot glass. 
2 M 2 
