Suppose we have, as shown in Fig. 124, a d-c generator supplying current to a vacuum tube oscillator. The current from the generator passes through an iron-cored choke coil, marked _L_{a}_ in the figure. Between this coil and the plate circuit we connect across the line a telephone transmitter. To make a system which will work efficiently we shall have to suppose that this transmitter has a high resistance, say about the same as the internal resistance, _R_{p}_, of the tube and also that it can carry as large a current. Of the current which comes from the generator about one-half goes to the tube and the rest to the transmitter. If the resistance of the transmitter is increased it can't take as much current. The coil, _L_{a}_, however, because of its inductance, tends to keep the same amount of current flowing through itself. For just an instant then the current in _L_{a}_ keeps steady even though the transmitter doesn't take its share. The result is more current for the oscillating tube. On the other hand if the transmitter takes more current, because its resistance is decreased, the choke coil, _L_{a}_, will momentarily tend to keep the current steady so that what the transmitter takes must be at the expense of the oscillating tube. That's one way of looking at what happens. We know, however, from Fig. 123 that to get an increase in the amplitude of the current in the oscillating tube we must apply an increased voltage to its plate circuit. That is what really happens when the transmitter increases in resistance and so doesn't take its full share of the current. The reason is this: When the transmitter resistance is increased the current in the transmitter decreases. Just for a moment it looks as though the current in _L_{a}_ is going to decrease. That's the way it looks to the electrons; and you know what electrons do in an inductive circuit when they think they shall have to stop. They induce each other to keep on for a moment. For a moment they act just as if there was some extra e. m. f. which was acting to keep them going. We say, therefore, that there is an extra e. m. f., and we call this an e. m. f. of self-induction. All this time there has been active on the plate circuit of the tube the e. m. f. of the generator. To this there is added at the instant when the transmitter resistance increases, the e. m. f. of self-induction in the coil, _L_{a}_ and so the total e. m. f. applied to the tube is momentarily increased. This increased e. m. f., of course, results in an increased amplitude for the alternating current which the oscillator is supplying to the transmitting antenna.