Assumption of a Solar Electric Potential. 169 
prodigiously powerful Leyden jar, by which action at a 
distance of the electricity of the conducting nucleus of the 
sun would be in a great measure excluded. This cannot be 
assumed at once, since the conductivity of flame has quite 
different causes, directly connected with the combustion-process 
itself, than that of conductive bodies not in chemical action ; 
so that an analogy between the two is hardly to be inferred 
with respect to their capacity for electric distribution. I have 
therefore made some experiments on the question whether a 
flame is subject to the action of induction in the same manner 
as other conductors; and these experiments have confirmed 
the hypothesis. According to this, two flames insulated from 
each other can act as the coatings of a charged Leyden jar, 
in like manner as other conductors*. According to this it 
must be assumed that the seat of the sun's electricity is chiefly 
to be sought in the photosphere, and not in the body itself of 
the sun. The electric properties of flame are still very 
obscure, notwithstanding all the experiments hitherto made 
on the subject. In particular, it is not yet decisively deter- 
mined whether a difference of potentials, spontaneously pro- 
duced, between the different zones of the flame, especially 
between that where the combustion commences and that 
where it goes out, exists or not. If this were the case, as 
appears probable from some experiments by Eiess and others, 
the cause of the sun's electricity might be sought therein, 
considering the prodigious dimensions of the sea of flame 
* The experiment was made thus : — A ring-shaped gas-burner was in- 
sulated. On opening the cock a cylindrical flame of about 2 centim. 
diameter rose above it to the height of about 15 centim. The flame 
passed through a metallic cylinder of about 8 centim. diameter, placed so 
as to be insulated and to surround it concentrically. To produce con- 
ducting connexion wi th the flame, an insulated platinum wire, bent into 
a circular shape, was placed in the lower part of the flame. The charge 
produced between this platinum wire and the cylinder by a galvanic 
series of fifty Daniell cells was now measured, by means of my rapidly 
oscillating electromagnetic switch, with the gascock alternately nearly 
closed and quite open. The difference between the deflections of the 
mirror-galvanometer was then a measure of the capacity of the Leyden 
jar formed by the flame and the cylinder. The results obtained are 
collected in the following table : — 
Difference of deflection 
Number of oscillations of 
between low and 
Amount of a 
the switch per minute. 
high flame. 
discharge. 
310 
8 
96 
600 
6 
100 
700 
8 
115 
1000 
12 
120 
The increasing numbers of the last column show that with slow 
oscillations a part of the charge was lost by conduction. 
