MR. W. CROOKES ON REPULSION RESULTING FROM RADIATION. 
109 
shown on the same figure, so I shall refer to the positions there given to illustrate the 
results here obtained. 
426. The mica screen was in the first experiment adjusted in the position shown 
at fig. 8 A, or as close as possible to the convex surfaces of the cups. The rare¬ 
faction was brought to 12 M. Electric contact being made, the wire ring became 
red hot, heating the blacked mica disk which rested on it. The fly commenced to 
revolve in the negative direction Th, at first slowly, then quicker, and afterwards 
slowly again, finally coming to rest after it had completed 34 revolutions. It now 
oscillated a little, and then remained quiet, although the battery contact was kept 
down, and the wire was glowing red. After remaining thus for a little time 
battery contact was broken. Instantly the fly acted as if released from a state of 
tension, and rotated in the positive direction A/', coming to rest only after a con¬ 
siderable time, 
427. The screens were now adjusted midway between the cups, as shown at fig. 8, E, 
and the foregoing experiment was repeated. As soon as the wire was made hot the 
fly commenced to rotate negatively, and after making 20^ revolutions it stopped. Now, 
however, instead of remaining quiet as in the last experiment, the fly commenced 
rotating in the opposite direction 9/, and kept moving at an uniform speed of 50 
revolutions a minute, as long as battery contact was maintained. On breaking contact 
the fly considerably increased its speed for some time, then gradually got slower, and 
finally came to rest. 
428. In this experiment the screens were brought close to the concave faces of the 
cups, as in fig. 8, G. The fly made 15 revolutions in the negative direction Ip, it 
then stopped and rotated positively at a speed of 86 revolutions a minute, keeping 
at this rate as long as the wire remained hot. On allowing the wire to cool, the 
speed got much faster at first, and then declined till the whole apparatus was cold, 
when all movement ceased. 
429. Experiments were now tried with the screens in various positions intermediate 
between those described in the foregoing paragraphs. The results differed in no 
material respect from those given above, the number of preliminary negative revolutions 
and the speed of the continuous positive revolutions being the chief variations. 
430. An alteration was made in the fly, the clear mica screens being lampblacked 
on the sides facing the concavity of the cups. On repeating the experiments described 
in pars. 426 to 428, it was found that the blackening had caused a change in the order 
of the phenomena. When the screens and cups were in position A, fig. 8, and no 
lampblack was on the screens, the action of the hot wire was to cause the fly to remain 
at rest, after having revolved a certain time negatively (426). Now, however, with 
the mica screen lampblacked, and in the same position, i.e., close to the convex surface 
of the cups, the permanent rotation was at the rate of 30 a minute positively, after 
having made 22 preliminary negative revolutions. Again, when the blacked mica 
