MR. W. CROOKES ON REPULSION RESULTING FROM RADIATION. 119 
whole instrument was floated in water, and one of the outer terminal wires was caused 
to dip into a mercury cup at the bottom of the water, wdiilst the other terminal dipped 
into a ring-shaped trough of mercury surrounding the upper part of the bulb. Bat¬ 
tery wires were connected with the two mercury cups, when the fly rotated rapidly. 
On arresting the movement of the fly by means of a magnet, the bulb was seen 
to rotate in the opposite direction. The movement, however, was feeble, and with 
difficulty could be kept up for any length of time, owing to the friction of the 
platinum wires in the mercury cups. Having demonstrated the fact of rotation, 
further experiments seemed unnecessary. 
448. The fixed vanes in the apparatus just described appearing to act in a very 
decided manner, further experiments were tried in this direction, as it was thought 
that the results might throw further light on the theory of' the movement. 
Fig. 23. 
A radiometer was furnished with a fly, the four vanes of which were cut from thin 
transparent mica, and were mounted symmetrically with the axis of rotation, not being 
favourably presented. At the side of the bulb, in a vertical plane, a plate of mica was 
fastened in such a position that each clear vane in rotating should pass it, clearing it by 
about a millimetre. The mica screen was cut away in the middle to allow the vanes to 
pass, as shown in fig. 23. The screen is a double one, formed of two plates of mica 
about a millimetre apart, the outer surface of one being lampblackecl, the other being 
clear. On bringing a candle near, and allowing the light to shine on the clear side of 
the mica screen, no effect is produced. If by means of a shade the light is allowed 
only to shine on the clear mica vanes, there is still no action ; but if the light shines 
on the blacked side of the mica screen, the fly rotates rapidly as if it were blown 
round by a wind issuing from the black surface, and keeps on moving as long as the 
light is near. 
A similar instrument contained a mica screen, blacked on both sides. The fly carried 
four bright aluminium cups (318). On shading the light of a candle from the blacked 
mica screen, and allowing it to shine only on the cups, the fly rotates positively, the 
convex sides retreating from the light. If the candle is allowed to shine only on the 
