of Metal Rods with rounded ends. 
259 
from the time when they first meet until they again separate, 
being measured on a ballistic galvanometer. Neglecting, for the 
present, the corrections due to self-induction, we get the formula : — 
q = ^T,or.;.T=pq (1) 
where q is the quantity of electricity ; V the voltage ; R 1 the 
resistance ; and T the time of impact. The general arrangement 
of the circuit is shown in Fig. 1. The galvanometer G was 
an instrument of the Ayrton-Mather form, by R. W. Paul, and 
had a coil of 600 turns (but of small self-induction) carried on 
a phosphor-bronze suspension. It was calibrated by placing it in 
circuit with a standard condenser and charging to a known voltage. 
This is the only method of calibration in which the galvanometer 
works under the same conditions as in testing the rods — i.e. 
making its fling on open circuit. Several attempts were made to 
check this calibration by the steady deflexion method, but were all 
frustrated by fatigue effects in the suspension when under the 
action of steady currents. The same voltmeter, V (a Siemens 
milli- volt-ammeter with variable series resistance), was used both 
in calibrating the galvanometer and in the actual testing of the 
rods. It was used in the same part of its scale for both purposes, 
so that no errors can arise from this source. In order that the 
varying contact-resistance between the rods, r lt r», might be safely 
neglected, and also with a view to diminishing self-induction 
effects, the whole resistance of the circuit was made up to 5,000 
ohms by means of the box R x . 
The corrections for self-induction are two in number : — 
(a) The current, instead of rising instantaneously to its full 
value when the rods meet, grows exponentially. 
