EXCURSION'S OF THE CAPILLARY ELECTROMETER. 
83 
is, as it were, a reservoir on each side of the electrometer, which acts as a spring, 
permitting a certain oscillation of the charge. A far more severe test is to preserve 
the electrometer circuit unbroken, and to remove the source of electromotive force, 
by breaking the primary circuit of the rheochord. This I have done, and the result¬ 
ing photographs show the effect of overshooting to be, for the same velocity of the 
meniscus, nearly twice as great. But it can still be overcome by external resistance, 
and is only noticeable when the difference of potential is two or three times as great 
as it ever is in physiological experiments, and even then I have not found it exceed 
2 per cent, of the length of the complete excursion, with no external resistance. 
It remains true, therefore, that with an ordinarily quick electrometer, under the 
conditions which obtain in practical work, the meniscus ceases to move the instant the 
source of electromotive force is withdrawn. 
The velocity of the movement is also greatest at the first, and there is no portion 
of the normal photographic curve, even at its commencement, concave to the 
asymptote, except in those cases in which overshooting may occur. Under such 
circumstances, slight signs of increase of the velocity are sometimes just discernible. 
But this was only detected in three cases, when the currents used were as great as 
the electrometer would bear without injury. 
These facts, implying the absence of acquired “ momentum,” seemed to indicate 
that in ordinary working the velocity of the meniscus at any moment must be some 
function of the accelerating force at that moment, in the sense that it is independent 
of any previous motion, and can contribute nothing—or practically nothing—to the 
velocity with which it moves during the next interval. 
In order to determine experimentally the form of this function, I arranged a 
rheotome in the derived circuit of a rheochord in such a way that any desired difference 
of potential could be introduced into the electrometer circuit for an accurately measured 
period of from '005 sec. to about '6 sec. This rheotome, which was upon the principle 
of a drop-shutter, acted by first snatching open a short-circuiting key, and then after 
the required interval, breaking the circuit on both sides of the electrometer simul¬ 
taneously—this arrangement being necessary, in order that the meniscus might stop 
suddenly without running back to zero. A specially sensitive electrometer was used, 
having a long range, and with no tendency to “ creep.” It was less rapid in its action 
than those usually employed in the laboratory for physiological work, being selected 
on this account so that the time-measurements might be more accurately made. The 
insulation was so good that no appreciable return of the meniscus took place during 
10 minutes after the circuit was broken. 
The experiments were conducted as follows :—Having found the total length of the 
excursion produced by a difference of potential derived from 150 millims. of the rheo¬ 
chord wire to be 126 divisions of the eye-piece micrometer, the rheotome was set so 
as to check the meniscus exactly in the middle of the excursion, namely, at 63 divisions 
from zero. This required a closure of '414 sec. The rider of the rheochord was then 
M 2 
