1879.] 
Electrical Insulation in High Vacua . 
243 
excited ebonite. The mica plate was now carefully lowered. 
As it came between the gold leaves they diverged farther 
apart, and kept so as long as the mica plate was between 
them. On removing the plate the leaves re-assumed their 
former divergence. This could be repeated any number of 
times. 
A similar piece of apparatus (Fig. 3) was made, only in- 
stead of a mica plate coming between the leaves, a mica 
cylinder, a , capable of being raised and lowered outside the 
divergent leaves, was employed. I was not able to get entirely 
concordant results with this, owing to the friction of the mica 
developing eleCfricity on the inner surface of the glass tube ; 
but in all cases, when the cylinder was raised until it covered 
the electrified leaves, it had the effeCl of diminishing the 
angle which they formed with each other. 
The following experiments were also tried : — The leaves 
being separated about 160°, as at Fig. 4, A, one side of the 
tube was slightly heated by a spirit-flame. The leaf on that 
side fell to a vertical position, and remained so when all was 
Fig. 4. 
cold, the other leaf sticking out as before, as at b. This 
would seem to show that the divergence of the leaves in this 
case was not so much due to their mutual repulsion as to an 
attraction exerted on each of them by the inner surface of 
the glass tube. The remaining divergent leaf could be 
slightly lowered when the glass tube above it was warme^ 
with a bunch of cotton-wool dipped in hot water. On coolin^ 
the leaf rose again to its original position. When this sid^ 
of the tube was also heated with a lamp, the leaf was re _ 
pelled down, but not so readily as the other had been, and 
