280 
MR. W. CROOKES OK REPULSION RESULTING PROM RADIATION. 
shading the light- from the black side the bright side was repelled, causing positive 
rotation, and on shading the light from the bright side the black was repelled, causing 
negative rotation ; but the positive repulsion was rather stronger than the negative 
repulsion, and consequently when both sides were illuminated the force was only that 
due to the difference of these repulsions.* 
A hot glass shade inverted over the bulb (245, note ) produced negative rotation, 
changing to positive on cooling. Both these rotations were stronger than that given 
by the candle. 
267. Instead of 4 millims. of the sides being turned up, 6 millims. were turned up, 
as shown in fig. 8. The candle gave moderately good positive rotation. On shading 
Pig. 8. 
the light from the black surface by means of a screen, and letting it shine only on the 
bright face, there was no movement. On shading the light from the bright side, 
and letting it shine on the black side, there was apparent attraction f with positive 
rotation, the speed being about half what it was when the light shone on both vanes 
simultaneously. A hot shade gave negative rotation, that is, in the opposite direction 
to that caused by light, changing to positive as the vanes cooled. 
268. The plates were now folded across the vertical diagonal, as shown in fig. 9. 
The candle gave strong positive rotation in the direction of the arrows. When a 
screen was interposed so as to allow the light to shine only on the black side, the 
positive rotation was kept up with scarcely diminished speed, the direction being that 
of attraction. When the light shone only on the bright surface no rotation took 
place. A hot glass shade gave good negative rotation, changing to the positive on 
cooling. 
* A radiometer similar to this, but not turned up to quite the same extent, is insensitive to light shining 
on it in the ordinary way. 
t I use the word attraction in these cases for convenience of expression. W hat appears to be attraction 
is really due to a vis a tergo. See par. 312. 
