ELECTRIC DISCHARGE WITH THE CHLORIDE OF SILVER BATTERY. 
75 
make the outer cylinder positive and negative alternately ; in the majority of cases the 
longest spark was obtained with the outer cylinder positive. For the first 1200 cells 
the striking distance between two conoentric cylinders is less, but after that number 
greater than between nearly flat surfaces, as will be seen by comparing the numbers 
given in Table YI. with those in Table VIII. * 
From curve XI., Plate 7, the following ratios for a given number of volts were 
obtained. 
* M. Gaugain, Annales de Chim. et de Physique, 4 a serie, T. VIII., pp. 115 to 118, has investigated 
the striking distance between two concentric cylinders, using, however, much higher potentials than 
we have employed. His experiments show that if the inner cylinder is kept constant and the number 
of elements and the diameter of the outer cylinder are increased or diminished, the density required to 
produce a spark does not change at the moment of the spark jumping. If the inner cylinder is increased 
or diminished in diameter, then the density is lessened or increased. Our results calculated in accordance 
with the formula 
N 
where fi = density, 
N = number of cells, 
R = radius of outer cylinder, 
r = radius of inner cylinder, 
give the following numbers : 
No. of Cells. 
Length of 
Spark. 
Density /x. 
1200 
Inches. 
0-0089 
3163 
2400 
0-0264 
2217 
3600 
0-0492 
1889 
4800 
0-0646 
2002 
5880 
0-0746 
2190 
6960 
0-0859 
2330 
8040 
0-0998 
2438 
From these results, and from our own with spherical and plane surfaces, as well as those of Sir W. 
Thomson, it is evident that the density is not constant for short lengths of spark (1200 cells) ; after that 
number it is pretty uniform, except in the case of 3600 cells, where it reaches a minimum. 
