1879*] Electrical Insulation in High Vacua . 241 
were attached to a glass stem, and the lower part of the bulb 
was drawn out for sealing to a Sprengel pump, as shown at 
Fig. 1. A stick of ebonite excited by fridtion was generally 
used as the source of electricity, but any other source will 
do equally well, provided it is not too powerful. 
No special attention was paid to the action of electricity 
on the leaves in air or at moderate vacua, as it agreed with 
what is already well known. The exhaustion was pushed to 
a very high degree (about the millionth of an atmosphere), 
when it was found that the excited ebonite had a much greater 
effect on the gold leaves than at a lower exhaustion ; for a 
long time, however, I was not able to charge the leaves 
permanently, in consequence of their falling together as soon 
as the source of electricity was removed. 
When a hot substance was brought near the bulb facing a 
gold leaf, so as to warm the glass, molecular repulsion took 
place, and the leaves retreated from the warm spot, standing 
out at an angle of about 45 0 . As the glass cooled the leaves 
resumed their former vertical position. 
While the leaves were repelled from the hot glass the 
excited ebonite had a very powerful adtion on them, and if 
it were brought near hastily the leaves flew off to the 
side of the glass, destroying the apparatus. By careful 
management and repeated trials, however, the ebonite 
could be brought near the warm spot of glass, the leaves 
suddenly extending at an angle to each other. The ap- 
pearance was as if a spark had been able to pass across 
the bridge formed by the line of advancing and retreating 
molecules connecting the hot glass with the gold leaves. 
On the ebonite being removed and the glass allowed to 
cool, it was found that the repulsion of the leaves was 
permanent. The rubbed ebonite would attradt and repel 
them as it was moved to and fro, but the angle formed 
by the leaves with one another remained unchanged. A 
warm body brought near the glass opposite one leaf would 
repel the pair as a whole ; on then warming the opposite 
side of the glass repulsion on that side took place, the angle 
of the leaves being somewhat diminished, but on cooling the 
leaves opened again to their former extent. 
When the glass bulb was strongly heated by a spirit flame 
the leaves suddenly discharged and fell together. 
Another bulb (Fig. 2) was prepared, containing a plate of 
mica, a, which could be suddenly placed between the gold 
leaves, b b. The plate of mica was longer and wider than 
the gold leaves, and was connected with a small piece of iron 
wire, capable of moving up and down a tube sealed into the 
VOL. ix. (n.s.) R 
