MR. W. CROOKES ON REPULSION RESULTING PROM RADIATION. 
131 
With the candle 1 foot off, there was at first repulsion 10 divisions, immediately 
followed by attraction 100 divisions. 
With the candle 2 feet off, there were 10 divisions repulsion, and 85 divisions 
attraction. 
With the candle 4 feet off, the first movement was repulsion 35 divisions, and then 
attraction 10 divisions. 
At 6 feet off the candle only gave repulsion 50 divisions, and no attraction. 
At 8 feet off the candle gave repulsion 10 divisions, with no subsequent attraction. 
The bar of the photometor was of pith, one-half plain, and the other half lamp- 
blacked. The repulsion is much stronger on the black than on the white half. The 
warm bulb attracts much more powerfully than radiation repels, therefore by increasing 
the distance of the candle from the bulb, heating is avoided and the action of radiation 
becomes apparent. 
477. The candle was now placed 10 inches from the bulb of the photometer. A 
glass cell full of water, and having parallel sides, was interposed, near the candle, and 
a thick sheet of plate glass was put close to the bulb. The light was screened off till 
an experiment was to be tried, and the exposure was continued only as long as was 
requisite to get the result. The index ray of light and the scale were the same as 
before. 
Tried in air (pressure 745 millims.) there were 10 divisions repulsion with no sub¬ 
sequent attraction, showing that the water and glass screens prevented the heating of 
the bulb. 
478. The bulb was then exhausted with the Sprengel pump, observations being 
taken from time to time. 
At 330 millims. pressure there was repulsion =12 
225 
115 
85 
80 
72 
68 
60 
45 
40 
35 
25 
10 
2 
1 
11 
9 
5 
3 
no movement 
repulsion 3° 
no movement 
repulsion 2' 
25° 
•n t-iO 
In this series I obtained no attraction. The gradual diminution of action to 
72 millims. pressure, its absence (with one exception) till the gauge was within 2 
millims. of barometric height, and its rapid increase after that, may, however, be 
s 2 
