ELECTROMAGNETIC UNIT OF ELECTRICITY TO THE ELECTROSTATIC UNIT. 609 
complete vibi’ation of the fork, the vibrations set up by the fork were sufficient to 
make tbe contacts uncertain. As finally fixed, the apparatus worked extremely 
satisfactorily, and would run for several hours without either the brushes or the 
springs requiring any adjustment. 
The commutator was driven from a water motor which was supplied from a tank at 
the top of the laboratory to secure a constant head of water. It was driven by a band 
of fine fishing line joined with a “ long splice ”; any rougher method of joining 
produced a joint, the effect of whose passage over the pulley of the commutator was 
plainly seen by the observer at the gah'anometer. A second band went from a small 
pulley on the motor to a pulley fixed within easy reach of the observer stationed at 
the tuning fork. The regulation of the speed was done by letting the auxiliary band 
run through the fingers, and slightly pressing it. This was found to be a much better 
plan than regulating by the band driving the commutator. But, in spite of this, the 
speed of the commutator as judged by the steadiness of the pattern seen through the 
slits of the tuning fork was subject to incessant small agitations, and it required con¬ 
siderable vigilance on the part of the observer at the fork to keep the pattern quite 
at rest. A heavier disc on the commutator would no doubt have made this easier. 
The supply of water was so adjusted that it was able to drive the commutator 
slightly faster then the speed required. The necessary fine adjustment was made by 
slightly pressing the regulating band. 
Determination of the Speed of the Commutator. 
To ascertain the speed at which the commutator was being driven its stroboscopic 
disc was observed through a pair of narrow slits fastened to the prongs of an electri¬ 
cally maintained fork. This fork made approximately 64 complete vibrations per 
second. The disc was provided with circles containing 4, 5, 6, 7, 8 spots at equal 
intervals, so that when a distinct pattern was observed through the slits on the fork 
the commutator made one of the following numbers of revolutions per second :— 
16-0, 18-3, 21'35, 25-6, 32-0, 36-6 427, 48-0, 
51-2, 54-9, 64, 73-2, 76-9, 80-1, 85-4. 
These numbers are respectively i, f, i, f, f f, |, f, f, f, 1, f, f, f, and | of 64. 
Any one of these speeds could be obtained by simply regulating the supply of water 
to the motor. Higher speeds could, of course, have been observed, but the motor 
would not drive the commutator much faster than 80 revolutions per second. 
Experiments at most of these speeds will be found below. 
It will be observed that this method gives a great choice of speeds, all of wdiich can 
be determined with the same accuracy, and whose relation one to another is known 
with absolute accuracy. 
The standard to which the speed was referred during the experiments was the 
MDCCCXC.—A. 4 I 
