528 
ME. W. CEOOKES ON EEPULSION EESULTINGr EEOM EADIATION. 
heat to pass through was proved with 
a thermopile. Solution of sulphate of 
copper was the most opaque to heat. 
98. Another form of apparatus is 
shown in fig. 5. Two bulbs were 
blown one in the other, and they 
were fused together at the necks; 
to the neck a small tube was fused 
for connecting with the Sprengel 
pump. The space between the two 
bulbs was then perfectly exhausted, 
and the small tube sealed up. I 
thus possessed what might be called 
a spherical shell of vacuum sur- 
rounding a bulb open to the air. 
In this inner bulb was suspended a 
pith ball on the end of a glass arm 
balanced by a knob of glass on to the 
other end, the suspending fibre being 
protected by a glass tube fitting into 
the neck of the inner bulb with a cork. 
It was found that heat applied to any 
part of the outer bulb passed across 
the vacuum, and attracted the pith 
ball (suspended in air). The sphe- 
rical shell of vacuum across which 
the heat passed, therefore, produced 
no change of action, but simply 
behaved like an extra thick glass 
bulb. This experiment bears upon 
the speculation in par. 81 of my 
former paper on this subject. 
99. Having succeeded in proving 
the fact of repulsion resulting from 
radiation, I was desirous of getting 
some quantitative estimations of the 
forces under examination. A pen- 
dulum-apparatus was constructed as 
shown in fig. 6. A wide glass tube 
(a b) has fused to it a narrower tube 
(c d), about 40 inches long; e is a 
Fig. 6. 
