134 
POPULAR SCIENCE REVIEW. 
the candle flame could make it, gave out heat instead of receiv- 
ing it. I then tried the converse experiment. After counting 
the number of u black repelling ” revolutions produced in one 
minute by the candle when the radiometer had obtained equal 
temperature with the surrounding air, I chilled the air in the 
top of the globe with 66 ether spray.” The heat the internal 
air received from the black discs was thus quickly disposed of, 
and the candle was able to produce greater effect, which was 
apparent by the increase in the velocity of the rotation. 
Wishing to somewhat vary this experiment, I fastened cotton 
wool tightly round A, until it was encased in a thickness of 
nearly an inch, leaving a part of the globe uncovered about the 
size of a half-crown piece, thus preventing to a great extent any 
radiation of heat from any part of the radiometer. I then 
covered it with a large glass jar, and on one side placed a solu- 
tion of alum about half an inch thick, which prevented the 
radiant heat from nine candles warming the globe. These nine 
candles were thus placed about 1 6 inches from the radiometer ; 
their light having to pass through the alum solution, the two 
glass sides of the vessel containing it, the glass jar and the 
glass of the globe, before it could affect the discs. They were 
of such a height that the more they burnt away the greater 
-effect they ought to produce, as the more light could then pass 
through the different glasses and the solution, and through the 
opening in the cotton wool placed nearly opposite to them. 
This opening was so situated that the light should fall almost 
entirely on the black discs. Soon after the nine candles were 
lighted I noticed that they caused the black discs to be repelled 
with a velocity equal to 4^ complete revolutions per minute. 
After burning for 15 minutes, the velocity had reduced to 3£ 
revolutions, and in time, I have no doubt, would have almost 
ceased. Then suddenly taking off the cotton wool, the velocity 
increased to 5J revolutions, at which rate it remained without 
change. Four candles produced at first 2f revolutions, de- 
creasing in 15 minutes to 1^, and when the woc-1 was removed 
immediately quickened to nearly 3. The great difficulty in 
these experiments is to prevent the loss of heat from the globe 
by radiation. If all is preserved, rotation, sooner or later, 
would probably cease. 
I have reason to believe that moonlight is capable of repel- 
ling the black discs, and that without the use of a condensing 
lens or telescope. I find the experiment attended by many 
difficulties; but if the radiant heat from the observer’s 
body at a distance of 8 feet from the radiometer is 
incapable of affecting the discs, when the temperature of the 
surrounding air is 10° below freezing point, then 
moonlight will cause the black discs to be slowly repelled at a 
