MR. W. CROOKES ON REPULSION RESULTING EROM RADIATION. 
527 
succeeded in the very difficult feat of sealing up some of these tubes under an internal 
pressure of 1^ atmosphere. 
96. To carry this experiment a step further bulbs containing a suspended ivory or 
mica index were filled with carbonic acid gas, water, carbonic disulphide, ether, alcohol, 
and other liquids. The index in carbonic acid behaved as if it were in air of somewhat 
higher density than the atmosphere ; movements were also obtained when the liquids 
were present, but they were so obviously due, in whole or in greater part, to currents, that 
they proved nothing of importance. 
97. Two other forms of the bulb -apparatus require mentioning. A thin glass bulb 
was blown 2J inches in diameter (fig. 4). Inside this another bulb was blown 2 inches 
in diameter, at the end of a glass tube 12 inches long. In this a light glass index with 
pith terminals was suspended, and the whole was perfectly exhausted. Fig. 4 shows 
the complete arrangement. In the space between the two bulbs various liquids were 
enclosed, such as water, solutions of sulphate of copper, alum, perchloride of iron, 
sulphate of iron, bichromate of potash, sulphate of nickel, &c. These were selected 
in the hope that amongst them one would be found which would sift out the heat-rays, 
and so allow me to obtain an action due to light. They, however, only affect the dark 
or extreme red heat-rays, and do not affect the luminous rays which also have a heating- 
Fig. 4. 
0 
Fig. 5. 
effect. By throwing a beam of sunlight on one of the pith disks powerful repulsion was 
obtained, whatever was the surrounding shell of liquid. That all these liquids allowed 
