ring that instant
they express their objection by an e. m. f. of self-induction which
opposes the generator voltage. For an instant, then, the voltage of the
oscillating tube is lowered and its alternating-current output is
smaller.

[Illustration: Fig 125]

For the purpose of bringing about such threatened changes in current,
and hence such e. m. f.'s of self-induction, the carbon transmitter is
not suitable because it has too small a resistance and too small a
current carrying ability. The plate circuit of a vacuum tube will serve
admirably. You know from the audion characteristic that without changing
the plate voltage we can, by applying a voltage to the grid, change the
current through the plate circuit. Now if it was a wire resistance with
which we were dealing and we should be able to obtain a change in
current without changing the voltage acting on this wire we would say
that we had changed the resistance. We can say, therefore, that the
internal resistance of the plate circuit of a vacuum tube can be changed
by what we do to the grid.

In Fig. 125 I have substituted the plate circuit of an audion for the
transmitter of Fig. 124 and arranged to vary its resistance by changing
the potential of the grid. This we do by impressing upon the grid the e.
m. f. developed in the secondary of a transformer, to the primary of
which is connected a battery and a carbon transmitter. The current
through the primary varies in accordance with the sounds spoken into the
transmitter. And for all the reasons which we have just finished
studying there are similar variations in the output current of the
oscillating tube in the transmitting set of Fig. 125.

In this latter figure you will notice a small air-core coil,
_L_{R}_, between the oscillator and the modulator tube. This coil
has a small inductance but it is enough to offer a large impedance to
radio-frequency currents. The result is, it does not let the alternating
currents of the oscillating tube flow into the modulator. These currents
are confined to their own circuit, where they are useful in establishing
similar currents in the antenna. On the other hand, the coil _L_{R}_
doesn't seriously impede low-frequency currents and therefore it does
not prevent variations in the current which are at audio-frequency. It
does not interfere with the changes in current which accompany the
variations in the resistance of the plate circuit of the modulator.
That is, it has too littl