711 
UNITS IN THE ELECTROMAGNETIC UNIT OP ELECTRICITY. 
Secondly, the comparison of the capacity of this guard ring condenser with that of 
a condenser without a guard ring. 
Thirdly, the determination by the method just described of the electromagnetic 
measure of the capacity of the condenser without a guard ring. 
I shall describe these parts separately. 
Part I. 
The calculation of the electrostatic capacity of the guard ring condenser. 
The condenser, which was designed by Lord Rayleigh, is represented in section in 
fig. 2, and in plan in fig. 3. 
Fig. 2. 
A 
I 
M 
Fig. 3. 
BH PD is a thick ebonite board placed in an approximately horizontal position, 
in this board two concentric circular grooves are cut. A cylindrical brass ring, H P, 
whose external diameter is about 23 centims., and whose height is about 10 centims., 
fits into the smaller of these grooves. Three pieces of ebonite carefully ground 
down to the same thickness (about 3 centims.), are placed at equal intervals on the 
top of this ring. On these the brass cylinder F G M N is placed ; this cylinder is of 
exactly the same diameter as the cylindrical ring H P, and is about GO centims. long. 
The cylinders, G F M N and H P, are placed so that their axes are coincident; this 
is tested by placing a straight-edge against the sides of the cylinder. On the top of 
this cylinder three pieces of -ebonite are placed, and upon the top of these a cylindrical 
ring E L, similar to the one at the bottom ; another brass cylinder, A B D C, made 
m three pieces, two rings similar in dimensions to the rings El P E L, and a long 
middle piece of the same length as the cylinder F G M N is then fitted over the 
other cylinders, the bottom ring fitting into the outer groove in the ebonite board ; 
the internal diameter of this cylinder is about 25 centims. The distance between 
the cylinders at the top is tested by observing how far a wedge, whose vertical angle 
is very small, sinks down between the cylinders. When the system is properly 
adjusted, the variation in the distance is only a small percentage of its mean value. 
4 y 2 
