430 
low this the point had its envelopes and the blue light, above 
which was a convex haze reaching to the edge of the glass, 
and streaming round it in a pink cone to the brass nut which 
secures the glass plate. With two Groves the strata were 
permanent, but cannot be called menisci, as in this arrange- 
ment the decrease of thickness at the edge scarcely occurred. 
“7. This, with hydrogen vacuum 0:-05 and two Groves, 
gave far less light than in the air vacuum. There was no 
fluorescence with sulphate of quinine, buta little with platino- 
cyanide of potassium.* With the ball positive, the spherical 
shells were well formed, but only through half the column of 
light: below that was luminous haze down to the dark inter- 
val near the point. When the point was positive, the shells 
were perfect hemispheres, with the glass disc as their diameter, 
and reaching to its edge: above that they were lesser seg- 
ments, which did not extend to the glass of the receiver. 
*¢8, Substituting for the balla glass disc similar to that 
below, and depressing the points 0'*05 below the surface of the 
glass, air vacuum 0:08, four Groves, this seemed to interfere 
with the production of the strata, which were scarcely percep- 
tible, and the stream of light, though intense, was narrow. 
Making the lower point to project 0:10, as before, they were 
much plainer when it was positive, but scarcely to be seen 
when it was negative. Replacing the ball, they were, as 
usual, spreading out into the faint envelope. When the in- 
duction circuit was not continuous, but completed by drop- 
ping sparks on the bind-screw of the pump-plate, or on a 
globule of mercury placed there, no change was produced in 
their appearance. When the contact breaker was surrounded 
by alcohol (which reduced the spark in air from 0'-42 to 0+:34) 
there was no alteration, except what arose from the diminished 
current. 
* For this I am indebted to G. G. Stokes, Esq. Itis the most sensitive 
test of fluorescent rays that I know. 
