"My first improvement consists in automatically tapping or disturbing the powder in the sensitive tube, or in shaking the imperfect contact, so that immediately the electrical stimulus from the transmitter has ceased the tube or imperfect contact regains its ordinary non-conductive state. This part of my invention is illustrated in Fig. 2.   .   .   .   In the apparatus I have made I have found that the relay n should be one possessing small self-induction, and wound to a resistance of about 1,000 ohms. It should preferably be able to work regularly with a current of a milliampere or less. The trembler or tapper p on the circuit of the relay n is similar in construction to that of a small electric bell, but having a shorter arm. I have used a trembler wound to 1,000 ohms resistance, having a core of good soft iron hollow and split lengthways, like most electromagnets used in telegraph instruments. The trembler must be carefully adjusted. Preferably the blows should be directed slightly upwards so as to prevent the filings from getting caked. In place of tapping the tube the powder can be disturbed by slightly moving outwards and inwards one or both of the stops of the sensitive tube (see Fig. 5  j¹ j²) the trembler p (Fig. 2) being replaced by a small electromagnet or magnets or vibrator whose armature is connected to the stop. I ordinarily work the receiving instrument h, which may be of any description by a derivation as shown, from the circuit which works the trembler p. It can also, however, be worked in series with the trembler. It is desirable that the receiving instrument if on a derivation of the circuit which includes the trembler or tapper should preferably have a resistance equal to the resistance of the trembler p."

"I prefer to use a mixture of two or more different metals. I find hard nickel to be the best metal, and I prefer to add to the nickel filings about four per cent. of hard silver filings, which increase greatly the sensitiveness of the tube to electric oscillations. By increasing the proportion of silver powder or grains the sensitiveness of the tube also increases, but it is better for ordinary work not to use a tube of too great sensitiveness as it might be influenced by atmospheric or other electricity. The sensitiveness can also be increased by adding a very small amount of mercury to the filings and mixing up until the mercury is absorbed. The mercury must not be in such a quantity as to clot or cake the filings, and almost imperceptible globule is sufficient for a tube. Instead of mixing the mercury with the powder one can obtain the same effects by slightly amalgamating the inner surfaces of the plugs which are to be in contact with the filings. Very little mercury must be used, just sufficient to brighten the surface of the metallic plugs without showing any free mercury or globules. The size of the tube and the distance between the two metallic stops or plugs may vary under certain limits, the greater the space allowed for the powder the larger or coarser ought to be the filings or grains. I prefer to make my sensitive tubes of the following size :--The tube j is 1½in. long and 1/10th or 1/12th of an inch internal diameter. The lengths of the stops j² is about 1/5th of an inch, and the distance between the stops or plugs j² j² is about 1/30th of an inch. I find that the smaller or narrower the space is between the plugs in the tube the more sensitive it proves, but the space cannot under ordinary circumstances be excessively shortened without injuring the fidelity of the transmission. Care must be taken that the plugs j² j² fit the tube exactly, as otherwise the filings might escape from the space between the stops which would soon destroy the action of the sensitive tube. The metallic powders ought not to be fine but rather coarse, as can be produced by a large and rough file. The powder should preferably be of uniform grain or thickness. All the very fine powder or the excessively coarse powder ought to be removed from it by blowing or sifting. It is also desirable that the powder or grains should be dry and free from grease or dirt, and the files used in producing the same ought to be frequently washed and dried and used when warm. The powder ought not to be compressed between the plugs but rather loose, and in such a condition that when the tube is tapped the powder may be seen to move freely."

"The tube j may be sealed, but a vacuum inside it is not essential except perhaps the slight vacuum which results from having heated it while sealing it. Care should also be taken not to heat the tube too much in the centre when sealing it as it would oxidise the surfaces of the silver stops and also the powder which would diminish its sensitiveness. I have used in sealing the tubes a hydrogen and air flame. A vacuum is however desirable, and I have used one of about 1/1000th of an atmosphere obtained by a mercury pump."

"in order to keep the sensitive tube j in good working order it is desirable but not absolutely necessary not to allow more than one milliampere to flow through it when active. If a stronger current is necessary several tubes may be put in parallel provided they all get shaken by the tapper or trembler, but this arrangement is not always quite as satisfactory as the single tube. It is preferable when using sensitive tubes of the type I have described not to insert in the circuit with it more than one cell of the Leclanché type as a higher electromotive force than 1·5 volts is apt to pass a current through the tube even when no oscillations are transmitted. I can, however, construct sensitive tubes capable of working with a higher electromotive force. Fig. 5A shows one of these tubes. In this tube, instead of one space or gap filled with filings, there are several spaces j¹ j¹ separated by plugs of tight-fitting silver wire. A tube thus constructed, observing also the rules of construction of my tubes in general, will work satisfactorily if the electromotive force of the battery in circuit with the tube is equal to about 1·2 volts multiplied by the number of gaps. With this tube, also, it is well not to allow a current of more than one milliampere to pass through it."

"This I effect by introducing into the circuits at the places marked p¹ p² q h¹ in Fig. 2, high resistances having as little self-induction as possible."

"in parallel across the terminals of the relay (i.e., corresponding to the circuit worked by the relay) it is well to have a liquid resistance s constituted of a series of tubes   .   .   .   partially filled with water acidulated with sulphuric acid. The number of these tubes in series across the said terminals ought to be about ten for a circuit of 15 volts, so as to prevent in consequence of their counter electromotive force, the current of the local battery from passing through them, but allowing the high-tension jerk of current generated at the opening of the circuit in the relay to pass smoothly across them without producing perturbating sparks at the movable contact of the relay. A double-wound platinoid resistance may be used instead of the water resistance, provided its resistance be about 20,000 ohms   .   .   .  Condensers of suitable capacity may be submitted to the above-mentioned coils, but I prefer using coils of water resistances."

"This I effect by connecting the battery wires to the sensitive tube or contact, or to the plates attached to the tube through small coils (see k¹ in the figures) possessing self-induction, which may be called choking coils, formed by winding in the ordinary manner a short length (about a yard) of thin and well-insulated wire round a core (preferably containing iron) two or three inches long."

"the oil or insulating liquid between the spheres e e increases the power of the radiation, and also enables one to obtain constant effects, which are not easily obtained if the oil is omitted. The balls d and e (Fig. 6) are preferably of solid brass or copper, and the distance they should be apart depends on the quantity and electromotive force of the electricity employed, the effect increasing with the distance (especially by increasing the distance between the spheres d and the spheres e) so long as the discharge passes freely. With an induction oil giving an ordinary 8in. spark, the distance between e and e should be from 1/25th to 1/30th of an inch, and the distance between d and e about 1in.   .   .   .  Other conditions being equal, the larger the balls the greater is the distance at which it is possible to communicate. I have generally used balls of solid brass of 4in. diameter, giving oscillations of 10in. length of wave. Instead of spheres, cylinders or ellipsoids, &c., maybe employed. Preferably the reflector applied to the transmitter ought to be in length, and opening the double at least of the length of wave emitted from the oscillator. If these conditions are satisfied, and with a suitable receiver, a transmitter furnished with spheres of 4in. diameter connected to an induction coil giving a 10in. spark will transmit signals to two miles or more. If a very powerful source of electricity giving a very long spark be employed it is preferable to divide the spark gap between the central balls of the oscillator into several smaller gaps in series. This may be done by introducing between the big balls smaller ones (of about ½in diameter) held in position by ebonite frames."
    "A further improvement consists in causing one of the contacts of the vibrating brake applied to the induction coil to revolve rapidly. This improvement has for its object to maintain the platinum contacts of the interrupter in good working order, and to prevent them sticking, &c. This part of my invention is illustrated in Fig. 3 (c² c³ c⁴). I obtain this result by having a revolvable central core c² (Fig. 3 and Fig. 13) in the ordinary screw c³, which is in communication with the platinum contacts. I cause the said central core with one of the platinum contacts attached to it to revolve by coupling it to a small electric motor, c⁴. This motor can be worked by the same circuit that works the coil, or, if necessary, by a separate circuit--the connections are not shown in the drawing. By this means the regularity and power of the discharge of an ordinary induction coil with a trembler brake is greatly improved."
    "A further improvement has for its object to facilitate the focussing of the electric rays. The oscillator in this case being different from the one previously described, because, instead of being constituted of two spheres it is made of two hemispheres, separated by a small space filled with oil or other dielectric. The spark between the hemispheres takes place in the dielectric from small projections at the centres of the hemispheres."
    "Fig. 9 shows another modified form of transmitter, with which one can transmit signals to considerable distances without using reflectors. In Fig. 9 t t are two poles connected by a rope t¹ to which are suspended by means of insulating suspenders two metallic plates t² t² connected to the spheres e (in oil or other dielectric as before) and to the other balls t³ in proximity to the spheres c¹ which are in communication with the coil or transformer c. The ball t³ are not absolutely necessary as the plates t² may be made to communicate with the coil or transformer by means of thin insulated wires. The receiver I adopt with this transmitter is similar to it, except that the spheres e are replaced by the sensitive tube or imperfect contact j (Fig. 5). whilst the spheres t³ may be replaced by the choking coils k¹ in communication with the local circuit. If a circular tuned receiver of large size be employed the plates t² may be omitted from the receiver.
    "I have observed that other conditions being equal the larger the plates at the transmitter and receiver, and the higher they are from earth and to a certain extent the further apart they are the greater is the distance at which correspondence is possible.
    "The permanent installations it is convenient to replace the plates by metallic cylinders closed at one end, and put over the pole like a hat and resting on insulators. By this arrangement no wet can come to the insulators, and the effects obtainable are better in wet weather. A cone or hemisphere may be used in place of a cylinder. The pole employed ought preferably to be dry and tarred.
    "Where obstacles such as many houses or a hill or mountains intervene between the transmitter and the receiver, I have devised and adopt the arrangement shown in Figs. 10 and 11. In the transmitting instrument (Fig. 10) I connect one of the spheres d to earth E preferably by a thick wire and the other to a plate or conductor u which may be suspended on a pole v and insulated from earth. Or the spheres d may be omitted and one of the spheres e connected to earth and the other to a plate or conductor u. At the receiving station, Fig. 11, I connect one terminal of the sensitive tube or imperfect contact j to earth E, preferably also by a thick wire, and the other to a plate or conductor w preferably similar to u. The plate w may be suspended on a pole, x, and should be insulated from earth. The larger the plates of the receiver and transmitter, and the higher from the earth the plates are suspended, the greater is the distance at which it is possible to communicate at parity of other conditions.  .  .

    "At the receiver it is possible to pickup the oscillations from the earth or water without having the plate w. This may be done by connecting the terminals of the sensitive tube j to two earths preferably at a certain distance from each other and in a line with the direction from which the oscillations are coming. These connections must not be entirely conductive, but must contain a condenser of suitable capacity, say of one square yard surface (paraffined paper as dielectric). Balloons can also be used instead of plates on poles, provided they carry up a plate or are themselves made conductive by being covered with tin foil. As the height to which they may be sent is great, the distance at which communication is possible becomes greatly multiplied. Kites may also be successfully employed if made conductive by means of tin foil. When working the described apparatus it is necessary either that the local transmitter and receiver at each station should be at a considerable distance from each other or that they should be screened from each other by metal plates. It is sufficient to have all the telegraphic apparatus in a metal box (except the reading instrument) and any exposed part of the circuit of the receiver enclosed in metallic tubes which are in electrical communication with the box (of course the part of the apparatus which has to receive the radiation from the distant station must not be enclosed, but possibly screened from the local transmitting instrument by means of metallic sheets). When the apparatus is connected to the earth or water the receiver must be switched out of circuit when the local transmitter is at work, and this may also be done when the apparatus is not earthed."

• United States Early Radio History > Period Overview > Radio Development > The Early Days of Radio in America