and the current diminishes. Hence suitable resistances are, first, a
high resistance for diminishing the current, and consequently, the
sparking at making and breaking of of the circuit; and, secondly, one
or more low resistances for varying the speed of the car. If the form
of _f_(C) be known, as is the case with a Siemens machine, equations 2
and 3 can be completely solved for _w_ and C, giving the current and
speed in terms of L, E, and R. The expressions so obtained are not
without interest, and agree with the results of experiment.

It may be observed that an arc light presents the converse case to a
motor. The E.M.F. of the arc is approximately constant, whatever the
intensity of the current passing between the carbons; and the current
depends entirely on the resistance in circuit. Hence the instability
of an arc produced by machines of low internal resistance, unless
compensated by considerable resistance in the leads.

The following experiment shows in a striking form the principles just
considered: An Edison lamp is placed in parallel circuit with a small
dynamo machine, used as a motor. The Prony brake on the pulley of the
dynamo is quite slack, allowing it to revolve freely. Now let the lamp
and dynamo be coupled to the generator running at full speed. First,
the lamp glows, in a moment it again becomes dark, then, as the dynamo
gets up speed, glows again. If the brake be screwed up tight, the lamp
once more becomes dark. The explanation is simple. Owing to the
coefficient of self-induction of the dynamo machine being
considerable, it takes a finite time for the current to obtain an
appreciable intensity, but the lamp having no self-induction, the
current at once passes through it, and causes it to glow. Secondly,
the electrical inertia of the dynamo being overcome, it must draw a
large current to produce the kinetic energy of rotation, i.e., to