ELECTRIC DISCHARGE WITH THE CHLORIDE OF SILVER BATTERY. 115 
With two spherical surfaces, each of 3 inches radius of curvature and 1*5 inch 
diameter, the following results were obtained with 8040 cells : — 
Pressure. 
Fraction of an 
Length of spark. 
Ratio to length at 
Ratio of length of spark 
millims. 
atmosphere. 
inch. 
1 atmosphere. 
to dilatation. 
760 
T 
0-079 
1-00 
i = 1-000 
602 
1*2 6 2 
o-ioo 
1-26 
i?nV = 0-999 
4147 
1*833 
0-200 
2-52 
VS7S = 1‘875 
299-5 
2*537 
0-400 
5-04 
= 1'986 
141-5 
1 
5*3 7 0 
0-800 
10-08 
= 1-876 
With spherical surfaces the spark is longer in the last three cases than it would be 
if in the inverse ratio of the pressure ; nearly, indeed, twice as long when the pressure 
is reduced to about two-fifths. 
The comparison between the striking distances with spherical surfaces, and with the 
point and disc, points to the probability of the distances coinciding at a certain degree 
of rarefaction. 
Note A. 
“ I should attribute the difference of adhesion of the paper to the plate according as the point was 
positive (P) or negative (N) to the different facility with which electricity passes by discontinuous, and 
not too rapidly succeeding, discharges from points P and N ; the facility being much greater from the 
former. The points I have in view are not the metallic point alone, but also those ends of the fibres of 
which the paper consists which happen to be at the surfaces. The dried paper is a very bad conductor. 
Its upper surface is electrified by the electrified air whether the metallic point is P or N, and the electri- 
city passes down through the bad conductor so slowly that the supply of electricity from the electrified 
air suffices to keep up a good charge on it. When the metallic point and consequently the paper is P, 
the ends of the fibres and other little roughnesses on the under face of the paper, by the property of 
P points, part with their electricity readily to the plate, so that the under portion of the stratum of paper 
is pretty well free from charge. When it is N, the negative electricity of the paper has much greater 
difficulty in passing into the plate, and therefore there is more charge in the paper, and that too is 
situated (in the mean) further down, so that the paper and plate are equivalent to a condenser thinner 
than the paper itself, and therefore capable of receiving a greater charge than a metallic plate whose 
lower surface is at the height of the upper surface of the paper, and which is opposed to the actual plate, 
the paper being supposed removed. 
“ When the paper is damp, it conducts so much better that the electrification by the electrified air 
cannot keep pace with the loss by transfer to the plate, and there is small charge, and therefore small 
attraction. When the upper electrode is greatly curved there is little electrification of the air, and 
therefore the charge of the paper resting on the lower electrode could not be kept up.” 
Q 
