In fig. 10 the secondary system is again illustrated diagrammatically in symmetrical arrangement. At one end of the coil is the aerial wire, and at the other a perfectly equivalent symmetry wire.

At the symmetrical centre Z

α

we have a potential node, and
at a and b current nodes.
These symmetrical arrange-
ments, on which theoretical
calculations are based with
regard to the duration of oscil-
lation of the entire system, can
rarely be realised in practice.
Moreover, the symmetry wire
has proved unfavourable, in
consequence of increased damp-
ing through the generation of
heat in accordance with Joule's
law, and through radiation.
Clearly enough, however, it
cannot be simply omitted, but
must be replaced by a more efficient substitute. The importance
of the symmetry wire, and the best substitute for the same, namely,
metallic surfaces, were first theoretically elucidated by J. Zenneck.
This observer formed the conclusion that the point of attack of the
inductive excitation must be at a place of electrical movement
(belly of the current), in order to excite oscillations of maximum
intensity. This certainly seems quite obvious.

FIG. 10.-Secondary System with Symmetrical
Attachments.

Hence the relative position of the coupled systems must be such as is shown in fig. 11. On the other hand, fig. 12 indicates a false relative position, since in this case the primary coil is situated opposite a current node in the secondary system. If the symmetry wire be cut off in this fashion, it must be supplemented by metallic surfaces, so that the belly of the current may return to the right position.

Fig. 13 gives the arrangement then obtained; and it follows— not merely from the foregoing, but also from purely theoretical considerations advanced by P. Drude-that we have to deal with well-defined conditions demonstrating the error of assuming that "earthing" is equivalent to a compensating counter-capacity. It would be more accurate to say that, in case of need, recourse may be

had to earth connection also; but this is quite impracticable when the soil is of badly conducting material. Earthing has also the drawback

FIG. 11.-Correct Relative Position of the Coupled Systems.

of causing considerable disturbance, by introducing atmospheric discharges.

I have determined experimentally the dimensions of the com

pensating metallic surfaces, the results coinciding almost exactly with the formula (see pp. 50, 55) deduced theoretically by Drude nearly a year later. This circumstance, and other experiments made in the same connection, lead me to consider it improbable that the

earth plays any important part in wave transmission. On account of "earth resistance" in this sense, many at present hold the idea that the metal surfaces should be as large as possible; but according to my researches, with the growing size of the compensating surfaces the theoretical size is at first very quickly and then gradually approached until we reach a corresponding maximum, which in practice shows itself to be the optimum.

If the aerial wire be grounded, it must of course be borne in mind, as already mentioned, that the vibration is thereby lowered; and besides as L. Mandelstam theoretically deduced-the coupling becomes closer.

CHAPTER IV.

THE RECEIVER.

We will now turn to a brief preliminary consideration of the receiving instrument and its development.

Coherer

Aerial Wire

Relay

The important component for detecting the electrical impulse is the coherer. This consists, as shown in figs. 14 and 15, of loose metallic powder or granules, packed into a small space between the terminal surfaces. of two metallic electrodes m, the whole being enclosed in a glass or ebonite tube. One electrode is connected to the aerial wire

collecting the electrical impulses, the other to the earth. In its ordinary condition, the coherer forms an imperfect contact, the high resistance of which in the circuit of the element E prevents the passage of the current. As soon, however, as the electrical impulses are received, the resistance of the coherer sinks to a very low value. A current then passes, and, by means of a relay, actuates a more powerful circuit (Battery B), as shown in fig. 15. In addition to a Morse instrument, this circuit includes a tapper for the purpose of gently shaking the coherer into the non-conducting state again after each exposure to radiation, and thus making it sensitive to new impulses. Through shorter and longer radiation we obtain in this manner the dots and dashes of the Morse alphabet, and thus wireless messages are received.

Earth

FIG. 14.-Arrangement of the Coherer.

Opinions differ widely as to the actual manner in which the coherer acts, and correspondingly numerous theories have been advanced. It is hardly necessary to repeat them, beyond giving a bibliographical list in the Appendix, to which those who are interested in the matter may refer.

Of importance, however, are the properties of the coherer, which have now been definitely established and may be summarised as follows. The coherer reacts on fluctuations in potential, even when

m r m

Coherer

Relay

Morse

Tapper

FIG. 15.—Arrangement of Receiving Instrument, with Coherer, Relay,
Morse, and Tapper.

the amounts of energy are exceedingly minute and transitory. Its sensitiveness to the rapidly alternating differences of potential of the vibrations pulsating in the receiver is so enormous that the most sensitive galvanometer is inferior by comparison.

In its ordinary condition the coherer must be regarded as high resistance and low capacity; and it is just these two properties that have rendered possible the evolution of the modern receiver, which-omitting intermediate stages of development is illustrated diagrammatically in fig. 16.

In this case, also, coupled systems are employed. A primary circuit with the capacity C and the self-induction L is connected