to encircle the armature. Again to
improve, the secondary coil is made to overlap partly the primary, so
that it cannot free itself from a strong inductive action of the
latter, repel its lines as it may. Once more to improve, the proper
difference of phase is obtained between the primary and secondary
currents by a condenser, self-induction, resistance or equivalent
windings.

I had discovered, however, that rotation is produced by means of a
single coil and core; my explanation of the phenomenon, and leading
thought in trying the experiment, being that there must be a true time
lag in the magnetization of the core. I remember the pleasure I had
when, in the writings of Professor Ayrton, which came later to my
hand, I found the idea of the time lag advocated. Whether there is a
true time lag, or whether the retardation is due to eddy currents
circulating in minute paths, must remain an open question, but the
fact is that a coil wound upon an iron core and traversed by an
alternating current creates a moving field of force, capable of
setting an armature in rotation. It is of some interest, in
conjunction with the historical Arago experiment, to mention that in
lag or phase motors I have produced rotation in the opposite direction
to the moving field, which means that in that experiment the magnet
may not rotate, or may even rotate in the opposite direction to the
moving disc. Here, then, is a motor (diagrammatically illustrated in
Fig. 17), comprising a coil and iron core, and a freely movable copper
disc in proximity to the latter.

[Illustration: FIG. 17.--SINGLE WIRE AND "NO-WIRE" MOTOR.]

To demonstrate a novel and interesting feature, I have, for a reason
which I will explain, selected this type of motor. When the ends of
the coil are connected to the terminals of an alternator the disc is
set in rotation. But it is not this experiment, now well known, which
I desire to perform. What I wish to show you is that this motor
rotates with _one single_ connection between it and the generator;
that is to say, one terminal of the motor is connected to one terminal
of the generator--in this case the secondary of a high-tension
induction coil--the other terminals of motor and generator being
insulated in space. To produce rotation it is generally (but not
absolutely) necessary to connect the free end of the motor coil to an
insulated body of some size. The experimenter's body is more than
sufficient. If he touches the