Thus, at the end of 1893, a few physicists clearly recognised that a new means had been given to us for detecting those invisible ether waves, the chief properties of which Hertz had unravelled with surpassing skill six years before, by means of a detector consisting of a ring of wire having a small spark-gap in it.
In June, 1894, Sir Oliver Lodge delivered a discourse at the Royal Institution, entitled "The Work of Hertz," and at this lecture use was made of the Branly tube as a Hertz wave detector. The chief object of the lecture was to describe the properties of Hertzian waves and their reflection, absorption and transmission, and many brilliant quasi-optical experiments were exhibited. Although a Branly tube, or imperfect metallic contact, then named by him a coherer, was employed by Sir Oliver Lodge to detect an electric wave generated in another room, there was no mention in this lecture of any use of the instrument for telegraphic purposes.[32]
As we are here concerned only with the applications in telegraphy, we shall not spend any more time discussing the purely scientific work done with laboratory forms of this wave detector.
Without attempting to touch the very delicate question as to the precise point at which laboratory research passed into technical application, we shall briefly describe the forms of kumascope which have been devised with special reference to Hertzian wave telegraphic work. A very exact classification is at present impossible, but we may say that telegraphic kumascopes may be roughly divided into six classes. The first class includes all those that depend for their action on the "coherer principle" or the reduction of the resistance of a metallic microphone by the action of electromotive force. As they depend upon an imperfect contact, they may be called contact kumascopes. This class is furthermore subdivided into the self-restoring and the non-self-restoring varieties. The second class comprises the magnetic kumascopes which depend upon the action of an electrical oscillation as a magnetising or demagnetising agency. The third class comprises the electrolytic responders, in which the action of electric oscillations either promotes or destroys the results of electrolysis. The fourth class consists of the electrothermal detectors, in which the power of an electrical oscillation as a high frequency electric current to heat a conductor is utilised. The fifth class comprises the electromagnetic or electro-dynamic instruments, which are virtually very sensitive alternating-current ammeters, adapted for immensely high frequency. The sixth class must be made to contain all those which cannot be well fitted at present into any of the others, such as the sensitive responder of Schäfer, the action of which is not very clearly made out.
We may proceed briefly to describe the construction of the principal forms of kumascope coming under the above headings. In the first place, let us consider those which are commonly called the "coherers" or, as the writer prefers to call them, the contact kumascopes. The simplest of these is the crossed needle or single contact, which originated with Professor E. Branly.[33] The pressure of the point of a steel needle against an aluminium plate was subsequently found by Sir Oliver Lodge to be a very sensitive arrangement when so adjusted that a single cell sends little or no current through the contact.[34] When an electric wave passes over it, good conducting contact ensues. The point is, in fact, welded to the plate, and can only be detached by giving the plate or needle a light shock or vibration. A variation of the above form is a pair of crossed needles, one resting on the other.
Professor Branly found, in 1891, that if a pair of slightly-oxidised copper wires rest across one another the contact-resistance may fall from 8,000 to 7 ohms by the impact of an electric wave. He has recently devised a tripod arrangement, in which a light metal stool with three slightly-oxidised legs stands on a polished plate of steel. The contact points must be oxidised, but not too heavily, and the stool makes a bad electrical contact until a wave falls upon it.[35] The decoherence is effected by giving the stool a tilt by means of an electromagnet.
These single or multiple-point kumascopes labour under the disadvantage that only a very small current can be passed through the variable contact when used as a relay arrangement, without welding them together so much that a considerable mechanical shock is required to break the contact and reset the trap.
The logical development of the single contact is, therefore, the infinite number of contacts existing in the tube of metallic filings, which has been the form of kumascope most used for many years. In its typical form it consists of a tube of insulating material with metallic plugs at each end, and between them a mass of metallic powder, filings, borings, granules or small spheres, lightly touching one another. Imperfect contact must be arranged by light pressure, and in the majority of cases the resistance is very large until an electric wave falls upon the tube, when it drops suddenly to a small value and remains there until the tube is given a slight shake or the granules disturbed in any way, when the resistance suddenly rises again. This type of responder is a non-restoring kumascope, and requires the continual operation of some external agency to keep it in a condition in which it is receptive or sensitive to electric waves.
Much discussion and considerable research have taken place in connection with the action and improvement of these metallic powder kumascopes. As regards materials, the magnetic metals, nickel, iron and cobalt, in the order named, appear to give the best results. The noble metals, gold, silver and platinum, are too sensitive, and the very oxidisable metals too insensitive, for telegraphic work, but an admixture may be advantageously made.
Omitting the intermediate developments of invention, it may be said that Mr. Marconi was the first to recognise that to secure great sensibility in an electric wave detector of this type the following conditions must be fulfilled: An exceedingly small mass of metallic filings must be placed in a very narrow gap between two plugs, the whole being contained in a vessel which is wholly or partly exhausted of its air. Mr. Marconi devoted himself with great success to the development of this instrument, and in a very short time succeeded in transforming it from an uncertain laboratory appliance, capable of yielding results only in very skilled hands, into an instrument certain and simple in its operations as an ordinary telegraphic relay. He did this, partly by reducing its size, and partly by a most judicious selection of materials for its construction. As made at present, the Marconi metallic filings tube consists of a small glass tube, the interior diameter of which is not much more than one-eighth of an inch, which has in it two silver plugs which are bevelled off obliquely. These are placed opposite to each other, so as to form a wedge-shaped gap, about a millimetre in width at the bottom and two, or at most three, millimetres in width at the top (see Fig. 16). The silver plugs exactly fill the aperture of the tube, and are connected to platinum wires sealed through the glass. The tube has a lateral glass tube fused into it, by which the exhaustion is made, which is afterwards sealed off, and this tube projects on the side of the wider portion of the gap between the silver plugs. The sensitive material consists of a mixture of metallic filings, five per cent. silver and ninety-five per cent. nickel, being carefully mixed and sifted to a certain standard fineness. In the manufacture of these tubes, great care is taken to make them as far as possible absolutely identical. Each tube when finished is exhausted, but not to a very high vacuum. The tube so finished is attached to a bone holder, by which it can be held in a horizontal position. The object of bevelling off the plugs in the Marconi tube is to enable the sensitiveness of the tube to be varied by turning it round, so that the small quantity of filings lie in between a wider or narrower part of the gap.[36]