The reason why these detectors act as they do is not quite understood. One suggested explanation is that the oscillating currents heat the particles and so partially weld them together. Another is that adjacent particles become charged as the plates of a minute condenser, and so are drawn tightly together as the plates in an electrostatic voltmeter are drawn towards each other. Supposing that the original non-conductivity of the loose filings be due to the film of air which may surround them, either of these things would account for the film being broken or squeezed out, resulting in better contact and improved conducting power. But both suggestions seem to be contradicted by the fact that if the pieces in contact be of certain substances the coherer works the opposite way. Under those conditions the conductivity is normally good, but the influence of the incoming waves causes it to become bad.
In 1896 Professor Rutherford, now of Manchester, described some discoveries which he had made as to the magnetic effects of oscillations. A simple little contrivance which he had constructed was operated by the discharge of a coil half-a-mile away, at that time a great performance. This detector was simply an electro-magnet with a steel core instead of the usual soft iron core. The reason the latter is used in the ordinary magnet is that it loses its magnetism the moment the current ceases to pass through the coil with which it is surrounded, while a steel core retains its magnetism. For most purposes a steel core would render an electro-magnet useless, but in this case it was desired that the core should be permanently magnetised. So a current was first passed through the coil to magnetise the core, and then the coil was connected to a simple form of antenna while a swinging magnet was brought near so that the magnetic power of the core would be indicated and any change made apparent. The effect of the discharge half-a-mile away was to demagnetise the core slightly. This was shown by the movement of the swinging magnet, and so the first "magnetic detector" was found.
But here, perhaps, I ought to explain the use of the antenna at the receiving station—its function at the sending end has already been made clear. The electro-magnetic waves, coming from the distant transmitter, strike the receiving antenna and in so doing set up in it oscillations such as those which set them in motion. For every oscillation in the sending antenna there will be another, similar in every respect except that it will be feebler, in the receiving antenna. And the oscillations are here led to the detector, of whatever form it may be, and in it they make their presence felt.
In some few cases a Duddell thermo-galvanometer has been employed as the detector, in which the oscillating currents report themselves directly. In coherers the detector works by causing the oscillating currents to control a continuous current from a battery and it is the latter which actually gives the signal, but there are a number of extremely interesting means which have been invented to detect the oscillating currents by their heating effect.
R. A. Fessenden, for instance, has perfected one which is a marvel of delicate workmanship. He depends upon the heating of a wire by the currents passing through it. Such heating is the result of the electrical force acting against resistance, and the difficulty is that if the resistance be great it will almost entirely kill the faint oscillating forces in the receiving antenna, while if, on the other hand, it be small, the rise in temperature will be inappreciable. So he encloses a fine thread of platinum in a glass bulb from which the air is exhausted. The platinum wire is first of all embedded in a wire of silver: the silver wire is given a core of platinum, in fact. Then the compound wire is drawn down until it is so thin that the platinum core is only one and a half thousandths of an inch in diameter. A short length of this compound wire is then bent into a U-shaped loop and its ends connected to thicker wires. Finally the bottom of the loop is immersed in nitric acid, which eats away the silver at that point and leaves the bare platinum. Thus is produced a very short length (a few millimetres) of exceedingly thin platinum wire supported at its ends by comparatively thick wires.
Being so short, this wire does not offer much resistance, and consequently does not materially check the oscillations. At the same time, since it is so fine, it does offer some resistance, and finally, since what heat is generated will be in an exceedingly small space, it will be appreciable there. A telephone is arranged so that its current also passes through the fine wire, and every slight variation in the temperature of the platinum wire, by varying its resistance, varies the current through the telephone. And exceedingly slight variations can be detected by sound in the telephone. Thus the oscillations generated in the antenna affect the heat in the wire; that affects its resistance; and that again affects the telephone, which, finally, affects the ear of anyone who is listening to it. It must be understood, however, that this is not a wireless telephone, for the sounds heard are not articulate but merely long and short sounds, representing the dots and dashes of the "Morse Code."
Electrolysis provides us with another form of detector. An exceedingly small platinum wire forms one electrode and a large lead plate the other, and both are immersed in dilute acid. The passage of current from a local battery sets up electrolysis, and so stops itself by forming a film of oxygen on the small electrode. This film, however, is broken by the oscillating currents from the antenna, so that as long as they are coming the battery current can flow, but as soon as they cease the battery current stops itself again. Thus the flowing and stopping of the oscillating currents is exactly copied by the current from the battery, which current is led through a telephone or a sensitive galvanometer.
It may occur to readers to inquire why the oscillating currents are not passed direct to a galvanometer. The answer is that because they are oscillating a very sensitive galvanometer is not possible.
True, the Duddell thermo-galvanometer has been mentioned in this connection, but although it is a beautiful instrument it cannot compare for delicacy with the direct-current galvanometers. The latter are easily a hundred thousand times more sensitive. But the trouble can be overcome by "rectifying" the oscillating currents, by passing them through a "unidirectional" conductor—one, that is, which passes current one way only. These remind one of a turnstile as installed at certain public places, which let you out but will not let you in unless you pay. In fact they will not let you in at all. In like manner "rectifiers" will only allow those currents to pass which are flowing in one direction, and so they cut out every alternate oscillation, thus producing something very like continuous current, which can be detected by the very delicate galvanometers which are usable where continuous currents are concerned, or more often by a telephone receiver. The rectifying conductors are in many cases crystals, hence these detectors are called "Crystal Detectors." Carborundum is a favourite for this purpose.
And that brings us to the important question of the secrecy of wireless communication, and the measures taken to prevent confusion from the number of independent messages flying through the air at the same time.