In this extract, Trouton proposed that one method for generating electromagnetic radiation might be by running an electrical alternator at very high speeds. At this time Tesla's alternator was capable of operating at up to 20,000 cycles-per-second, but this was far too slow to generate light. However, the author thought that with improvements the device might someday achieve the required higher speeds.
 
The Electrician (London), January 22, 1892, page 302:

RADIATION  OF  ELECTRICAL  ENERGY.*

    What we do here is apparently to give the molecules of these substances a great shake every now and then, they continuing to vibrate between each disturbance after their own fashion as little electric radiators, thus affording us light. Experiments lately made by Prof. Tesla, of America, seem to show that the more frequently they are disturbed the better. An ideal, of course, would be to give them a help each oscillation by a synchronous alternating current. Prof. Tesla has built an alternating dynamo, which afforded currents alternating 20,000 times a second; but this is far and away short of what would be required, if in truth such a thing is possible, as a circuit carrying a current alternating in periods comparable with that of light. The speed of rotation and the number of coils which can fit on a rotating disc seems nearly reached in his machine; but by combining the principle of the transformer with that of the dynamo, we could push the rate of alternation beyond these mechanical limits: for instance, by passing an alternating current through the field-magnets of an alternating dynamo instead of the usual continuous current.
    We might suppose this done by having two machines on the same shaft, the first arranged as usual, while the field-magnets of the second derive their current from the armature of the first, and in such a way that when the armature coils and field coils are closest the current in the field coils should be zero. The current from the armature of the second machine would thus be of double the rate, for we have superimposed on the geometrical zero position of the armature coils which occur half way between two field coils, zero position situated at the field coils. The two machines might, of course, be combined, only in that case the field-magnets of our second machine rotate, and are in fact the armature coils of the first, while the second set of coils to form the second armature would remain at rest. Unfortunately, we cannot go on in the same fashion doubling the rate, only, indeed, adding subsidiary vibrations by further machines.
_______________
    * The conclusion of a series of lectures delivered in Inverness, December, 1891, under the auspices of the "Ettle's Trust."