272 
Mr Sears, On the Longitudinal Impact 
C 2 represents the standard condenser used in calibrating the 
galvanometer. In using this instrument, it was necessary to throw 
it in or out of the circuit instantaneously, owing to the leakage 
effects, which persisted in spite of the most careful insulation. 
For this reason the short-circuiting key, K. 2 , was inserted. On 
depressing this key, the galvanometer receives its true impulse, 
the whole charge on the condenser passing through it before any 
leakage currents have time to produce an appreciable effect. The 
result of the calibration was to give an exact straight line over the 
range in which the galvanometer was used. This line, however, 
does not pass quite through the zero, and this effect was finally 
traced to the condenser ; a similar calibration, in which the 
condenser was discharged , instead of charged, giving a straight line, 
parallel to the first, but passing on the other side of the origin. 
The error appears quite insignificant on plotting, but it is sufficient 
to account for the fact that the flings recorded later on, as the 
result of charging the smaller capacities to higher potentials, 
apparently do not quite agree with the value we should expect 
from the galvanometer constant taken from the straight line. This 
constant is found to be: q = 2'945 x 10 -8 8 where q is in coulombs, 
and 8 in scale divisions. 
[With regard to the rest of the connections, 0 1 is a condenser 
inserted in order to ensure the E.M.F. keeping constant during the 
sudden discharge, as any variation occupying so short a time 
would have no effect on the voltmeter. It is not essential. The 
resistance, was so inserted as to be always in circuit with the 
galvanometer, even when the latter was shorted, as it was found 
that the violent damping effect caused by direct shorting 
continually altered the zero of the galvanometer. IF is a thick 
copper wire, of negligible resistance, to any point of which the 
lead, w u could be clamped, and F 1 is a fuse wire to provide against 
accidental short-circuits.] 
We have lastly to consider the working of the circuit with a 
view to determining the necessary corrections. In the normal 
experiment, the rods were adjusted, the galvanometer key 
depressed, and, when the spot had come to rest at zero, the fuse 
was blown and a reading taken. The contact-resistance between 
the rods being negligible compared with the whole resistance of 
the circuit, it follows that when the rods first impinge, the resist- 
ance falls to its final value, Rj (= 5000) practically instantaneously. 
We may therefore treat the circuit as having this constant resist- 
ance during the growth of the current at the beginning of the 
impact. The equation for this growth then becomes : — 
L i t +E ‘ i+v =° or 
