[Illustration: FIG. 1. FIG. 2.]

All measurements are comparative. We measure weights or forces by
comparison with some generally known and accepted unit standard
weights, lengths, areas, and volumes, by comparison with a unit length,
resistance by a standard ohm, and so forth. In the same way currents
could be measured by comparison with a standard current: but this
would be a troublesome process, not only on account of the apparatus
necessary, but also because it would be a matter of some difficulty to
have a standard current always ready for use. In general, measurement
by direct comparison with a standard unit is discarded for the more
indirect method of measuring not the current itself, but its chemical,
mechanical, or magnetic effect. The chemical method is very accurate if
a proper density of current through the surface of the electrodes be
used,[1] but since it requires a considerable time, and, above all, an
absolutely constant current, its use is almost entirely restricted
to laboratory work and to the calibration of other instruments. For
practical ready use, instruments employing the mechanical or magnetic
effect of the current are alone suitable. We weigh, so to speak, the
current against the force of a magnet, of a spring, or of gravity.
The measurement will be exact if the thing against which we weigh or
counterbalance the current itself retains its original standard value.
Where permanent magnets or springs are used as a balancing force, this
condition of constancy in our weights and measures is not always fully
maintained, and to make matters worse, there is no visible sign by which
a change, should it have occurred, can be readily detected. A spring
may have been overstrained or a steel magnet may have become weakened
without showing the least alteration in outward appearance. To overcome
this difficulty, the obvious remedy is not to use springs or steel