however, is not an important matter. Far more important is it to show, as has been shown, that practically perfect isolation can be achieved if it is desired.
It must be noted, however, that, although the fact that electric circuits have a natural time-period of oscillation of their own is a scientific principle which carries us a considerable way towards a solution of what is called syntonic Hertzian wave telegraphy, it is not in itself alone in every respect an entire solution of the practical problem. The degree to which it is a solution depends to a considerable extent upon the nature of the detecting device, or kumascope, which we are employing. The coherer, or Branly filings tube, has the peculiarity that its passage from a nonconductive to a conductive condition follows immediately when the difference of potential between its ends is made sufficiently great. In other words, if the tube is acted upon by a sufficient electromotive force, it is not necessary that electromotive force should be repeated at intervals to make this particular form of kumascope responsive. Again, if we consider the nature of the oscillations which are sent out from any transmitting aerial, we find that each group of oscillations corresponding to a single spark consists of waves gradually decreasing in amplitude. In other words, the first wave of the group is the strongest, and the decay in amplitude is often very rapid. Supposing, then, we construct a simple receiver consisting of an aerial having inserted in its circuit a sensitive Branly filings tube. Such a receiver is almost entirely non-syntonic ; that is to say, it is affected by any wave passing over it which is sufficiently powerful. We may look upon it that if the first wave of the series is sufficiently powerful to affect the kumascope, the conductive change takes place whether or not the first wave is followed by others. Accordingly, it is perfectly certain that if a transmitter is sending out trains of waves of any period, a simple combination of coherer and aerial will be influenced, if it is placed near enough to the transmitter. On the other hand, it is possible to combine a kumascope of a certain type with a receiving aerial and other circuits in such a manner that when the waves that reach it are feeble it shall not respond at all unless those waves have very nearly a time period of a certain value.
At this stage, it may be perhaps well to explain a little in detail what is meant by an easily responsive circuit, and, on the other hand, by an irresponsive circuit, or, as we may call it, a stiff circuit. Supposing that we consider an aerial consisting of a simple straight wire having small capacity and small inductance, such a circuit admits of being sent into electrical oscillation, not only by waves of its own natural time-period, but by the sudden application of any violent electromotive impulse. If, on the other hand, we bestow upon the circuit in any way considerable inductance, we then obtain what may be called a stiff or irresponsive circuit, which is one in which electrical oscillations
