298 
MR. W. CROOKES ON REPULSION RESULTING FROM RADIATION. 
325. Series IV. — Aluminium Cups, blacked on both sides. 
Experiment a. 
No screen interposed. 
33 revolutions a minute. 
Fig. 27. 
.Experiment b. 
Concave screened off. 
12 revolutions a minute. 
Experiment c. 
Convex screened off. 
8 revolutions a minute ; 
apparent attraction. 
326. It will be observed that in Series I., experiment c, Series II., experiment c, 
and Series IV., experiment c, I get the same apparent attraction as with the 
bright aluminium cones (309, 310). The explanation given in par. 312 covers all 
these cases, and also accounts for the absence of rotation in Series II., experiment b, 
and Series III., experiment c. As the different behaviour of the blacked and plain 
cups affords a further proof of the correctness of the theory there advanced, it will be 
useful to consider these cases seriatim. 
327. The case of the cups bright on both sides (Series I,, par. 322) is covered by 
that of the bright aluminium cones already discussed (309 to 312). In Series II. (323), 
the lampblack is applied to the concave side of the cups ; the absorption of light on 
that side is consequently increased, and the temperature of the cups rises more 
rapidly than when both sides are bright. Lampblack is not only a good absorber 
of light and heat, but it rapidly gives up its heat to the gaseous molecules, and is 
thereby a most powerful generator of molecular pressure, whilst the bright side, 
giving up heat less easily, produces less molecular pressure. An excess of molecular 
pressure is therefore generated on the concave side by virtue of its black surface, and 
a less amount of pressure produced on the bright convex side. Were the presentation 
of these two sides to the glass bulb equally favourable, the excess of pressure on the 
black side would overcome the other, and the black would retreat ; but as shown in 
fig. 21 (312), most of the molecular disturbance from the concave surface is dissipated 
before it gets to the glass, whilst a great part of that from the convex surface is 
active. It follows that the active pressure from the black is not sufficient to overcome 
that from the convex surface, and the excess determines the direction of rotation. 
The influence of the black is apparent in diminishing the speed from 18 to 11 revolu- 
tions a minute. 
328. In Series II., experiment b, in which the concave black is screened off and the 
light shines only on the bright convex side, there is no rotation but only slight 
