136 
POPULAR SCIENCE REVIEW. 
into a cooler part of the globe enough heat (obtained by its 
absorptive power) to enable it to be further repelled, as in the 
way already described, the next black disc is thus brought under 
the influence of the warm glass, and the operation is repeated. 
If, when the white disc be first repelled, and no rotation pro- 
duced, the finger be removed, and the globe chilled, the black 
gradually approaches the heated part of the globe, absorbing and 
radiating the heat, and so contracting the air, and making room 
for itself, as mentioned in the description of u reverse motion.” 
I noticed that the black surfaces on my radiometer B were 
gradually becoming more grey every day ; in fact, the black 
(lamp black, I believe) was coming offlittle by little — in places 
it had worn almost white. This was not the case with A, 
although it had done much more work. I at once thought that 
this was caused by the reverse motion, the black particles 
having been sucked off by the contracting air. I therefore 
placed B on its side, and so collected the black dust in one 
spot in the bottom of the stem of the globe, and carefully 
placed it upright again. I then heated it, and so caused the 
black discs to be repelled with great velocity, carefully noticing 
at the time, by the help of a magnifying lens, whether any 
black particles fell on to that side of the stem that I had 
cleared of them. Not one- appeared, until the reverse motion 
began. I was then soon able to count a score or more. I 
repeated this experiment two or three times, and always with 
the same effect. It must not be tried too often, as every 
time it makes the black discs less sensitive to the action of 
heat and light for further experiments. 
I think one can compare the residuum of air within a 
radiometer to a piece of elastic stretched almost to its utmost, 
but far too strong to be broken by any “ Sprengel air-pump.” 
I have sometimes used the word “ heat,” and sometimes 
“ light ” in describing these experiments. They cannot be 
considered as two separate agencies. All that is necessary to 
produce 66 black repulsion ” in a radiometer is that one should 
use some part of the spectrum between the ultra red or heat 
rays, wherever they may begin, and the extreme luminous part. 
The action decreases rapidly from the red towards the blue. 
If the ultra red (or “ active heat ”) rays be cut off by the ■ j 
intervention of an alum solution, the red rays (the others giving 
but very little help) are absorbed by the black surfaces, and a 
little heat is produced, which it seems causes the repulsion. 
If a source of light be used which is very deficient in red rays, 
the light, however intense, is hardly capable of moving the 
discs. I have used the electric spark from a 44 6-inch ” induc- 
tion coil, intensified by introducing Leyden jars into the 
secondary circuit, and allowed the continuous stream of sparks 
