368 Prof. J. A. McClelland on the Comparison 
(d) A careful experiment was now made to find how small 
a capacity could be detected and measured by this method. 
To do this a condenser of the following type (fig. 2) was 
arranged. 3 
Fig. 2. 
B 


a i ee le a 
Pee eee A TE 
c aa 
A 
AB is a long wide tube, 7°90 cms. internal diameter ; ab is 
another tube fixed, as shown, to be coaxial with AB. In ab 
a third cylinder cd slides, fitting closely into ab, the external 
diameter of cd being 1°94 cm. ADB is joined to earth, and ab 
(and cd) connected to the Kelvin electrometer. The capacity 
of the electrometer, the condenser as arranged (fig. 2), and 
joining wires is determined by comparing it with a standard 
capacity of -001 microfarad. 
A careful series of observations is then taken, with ed in its 
above position, using a suitable quantity of uranium nitrate. 
Then the tube cd is moved 8°01 cms. further into AB, care 
being taken not to displace AB or ab. A vernier was attached 
to ed working in a slot in ab, so that the distance through 
which cd was displaced could be accurately measured. A 
second series of observations was then made with cd in the 
new position, keeping the same uranium as before. We have 
thus the data for deducing from the experiments the increase 
of capacity produced by the movement of cd. This increase 
of capacity can also be accurately calculated from the formula 
atic 
2 log 

. 
Je 
ry 
i) 
since the effects of the ends are eliminated by the arrange- 
ment used, / being the distance cd is moved, and 7, and T2 
the radii of AB and ed respectively. We can thus estimate 
the value of the method for measuring very small capacities. 
The numbers observed were as follows :— 
(1) Finding the capacity C made up of ie condenser 
at (fig. 2), the electrometer, and connexions. 
C. 001+C: 
100 divisions in 39°7 seconds 450 divisions in 117°8 seconds 
100 divisions in 39°6 __,, 50 divisions in 1181 __,, 
100 divisions in 395 __,, 50 divisions in 117°8 
2) 
Mean giosu oe, Mean... 117°9 
29 
