294 
MR. W. CROOKES OK REPULSION RESULTING FROM RADIATION. 
When silver-flake mica is used to make the cones, the action is different. Here 
some of the light and heat are arrested, with a consequent rise of temperature. 
But silver-flake mica being a very bad conductor of heat, most of the molecular 
pressure is exerted from the surface on which the light falls, and very little is 
generated on the other side. In position C, fig. 20, when the light shines only on 
the concave surface no motion is produced, as no pressure is generated on the surface 
away from the light ; whilst in positions A and B, where the light shines on the 
concave surface, the movement is nearly equal in velocity. No effect being produced 
by the light shining on the concave surface, no retardation is caused when the light is 
screened from it. 
314. The influence which a variation of shape of the moving fly has on the direction 
and amount of its repulsion depends on favourable presentation to the surrounding 
bulb. A very little alteration in angle of presentation has more effect than a great 
difference of colour. It remains to be seen, however, what angle or inclination of the 
vanes to the glass bulb is the most favourable. Cones being inconvenient in shape, 
I employed portions of cylinders wherewith to shape the vanes. Four sets of 
cylindrical aluminium vanes were mounted on arms, and pivoted on needle points, in 
four bulbs. They were made at the same time and were alike in all respects, except 
in the radii of curvature of the cylinders on which the vanes were curved. The vanes 
were of thin aluminium, and the bulbs were exhausted simultaneously by being sealed 
on to four arms of a five-tube mercury pump, so as to ensure the exhaustion being 
identical in the four bulbs. Fig. 22, A, B, C, D, shows the four sets of vanes drawn 
Fig. 22. 
the full size. The radius of curvature of A is 5 millims., of B 10 millims., of C 
20 millims., and of D 30 millims. Each vane is 10 millims. high and 10 millims. 
across the chord of the arc ; consequently there is more metal in the vanes of deeper 
than in those of shallower curvature. This was thought to afford a fairer means of 
