110 
ME. W. CROOKES ON REPULSION RESULTING EROM RADIATION. 
screen was brought close to the concave surface of the cups, as at fig. 8, G, the preli¬ 
minary negative revolutions were too rapid to count; after a time they slackened and 
ceased, and the fly then refused to move till the current was turned off, when it 
seemed to be suddenly released from constraint, and revolved rapidly in the positive 
direction. A reference to par. 428 will show that when no lampblack was on the mica 
screens there was permanent rotation in this position at a somewhat rapid rate, after 
the first preliminary turns had taken place. 
431. The fly and screens were now altered in the several ways described in pars. 403, 
404, 405, 406, 408, and numerous experiments were tried with them, the relative 
positions of cups and screens being altered each time. To recapitulate the several 
results would be wearisome. I will therefore shortly say that they confirm what has 
already been said as to the effect of the hot wire. In all cases whilst the fly was 
receiving heat from the wire the direction of motion was negative. When the fly had 
ceased to increase in temperature the continuous rotation was positive, but occasionally 
a position of fly and screens was met with at which no movement took place (426, 430). 
On breaking battery contact, and allowing the whole instrument to cool, the first effect 
was to cause the fly to move at a greater speed positively, gradually slackening till it 
had cooled down to the ordinary temperature, when all movement ceased. 
Besides these I tried two other forms of fly, which yielded results of sufficient 
interest to be worth giving more in detail. 
432. The first fly consists of a single arm. supported in the centre by a glass cap 
working on the needle point, and having at one extremity a small metal counterpoise, 
and at the other a single aluminium cup. The cup is supported on the horizontal arm, 
which forms an axis traversing the centre of gravity of the cup through holes pierced 
in the sides. 
Fig. 18 shows this arrangement of fly; a is the counterpoise balancing the cup h ; 
Fig. 18. 
6' S' S *>" 
3 E) ^ 
J 2 3 i 5 
the axis passes through holes shown at d, on the side view of the cup L". Stops c c 
on the axis keep the cup in position, and just sufficient friction is secured to retain the 
cup in any desired position, shown at b", 1, 2, 3, 4, 5, into which it may be brought by 
tapping the apparatus. The remainder of the apparatus, the black mica disk, the 
platinum wire ring, &c., is the same as before. 
When the cup was set horizontally with the concavity upwards, as shown at //, and 
