FOOTNOTES
1 Mr. Gray, in an article inserted in the Telegrapher of October 7, 1876, enters into full details of this mode of transmitting sounds by the tissues of the human body, and he gives the following as the conditions in which it must be placed to obtain a favourable result: 1. The electricity must be of a high tension, in order to have an effect perceptible to the ear.
2. The substance employed to touch the metallic plate must be soft, flexible, and a good conductor, up to the point of contact: it must then interpose a slight resistance, neither too great nor too small.
3. The disk and the hand, or any other tissue, must not only be in contact, but the contact must result from rubbing or gliding over the surface.
4. The parts in contact must be dry, so as to maintain the required degree of resistance.
2 He cites the following names in his account of electric telephony:—Page, Marrian, Beatson, Gassiot, De la Rive, Matteucci, Guillemin, Wertheim, Wartmann, Janniar, Joule, Laborde, Legat, Reiss, Poggendorf, Du Moncel, Delezenne, Gore, &c. Vide Mr. Bell’s paper, in the Journal of the Society of Telegraphic Engineers in London, vol. vi. pp. 390, 391.
3 This statement is disputed by Mr. Elisha Gray, owing, as we shall see, to a misunderstanding as to the word undulatory current.
4 Elisha Gray. Eng. Pat. Spec. No. 2646, Aug. 1874.
5 This property has long been known, but not applied. In 1856, in the second edition of my Exposé des applications d’Electricité, I pointed them out in speaking of the contact-breakers. I also spoke of them in a paper on electro-magnets (published in the Annales télégraphiques, 1865), and in several articles laid before the Académie des Sciences in 1872 and 1875 on the conductivity of filings and conducting powders. M. Clérac, in 1865, also used them to obtain variable resistances.
6 In 1865 I was able to verify this observation when tightening the spirals of an electro-magnet on a naked wire. The greater the number of spirals under pressure, the more definite were the differences of resistance in the magnetising helix.
7 M. Hellesen communicated the plan of his instrument to me on May 3, 1878, and his experiments were made in Copenhagen three weeks earlier.
8 M. M. J. Page had already noticed that if a telephone is placed in the circuit of the primary helix of an induction coil, while the secondary helix of this instrument is placed in the circuit of one of M. Lippmann’s capillary electrometers, a movement of the mercurial column of the electrometer takes place at each word, and this movement is effected towards the capillary end of the tube, in whatever direction the current is sent by the telephone. This is because the mercury always tends to move more rapidly at its capillary end than at the other extremity.
9 Mr. Edison, in a letter written November 25, 1877, writes that he has made two telephones which act with copper diaphragms, based on Arago’s effects of magnetism by rotation. He ascertained that a copper diaphragm might replace the iron plate, if its thickness did not exceed 1/32 of an inch. The effect produced is slight when the copper diaphragm is placed between two corresponding instruments; but when the sender only is furnished with the copper diaphragm, and the receiver is arranged as usual, communication becomes easy.
Mr. Preece repeated these experiments, but he only obtained very slight and indistinct effects: he consequently believes that they are of no practical use, although very interesting in theory.
10 Mr. Bell had previously made a like experiment, which suggested to him that molecular vibrations had as much to do with the action of the telephone as mechanical vibrations.
11 M. Bosscha, who has published in the Archives néerlandaises an interesting paper on the intensity of electric currents in the telephone, says that the minimum intensity of currents necessary to produce a sound in a telephone by the vibration of its diaphragm may be less than 100/1000 of a Daniell element, and the displacement of the centre of the diaphragm would then be invisible. He was unable to measure exactly the range of movements produced in the diaphragm by the influence of the voice, but he believes it to be less than the thousandth part of a millimètre; and from this it follows that, for a sound of 880 vibrations, the intensity of the induced currents developed would be 0·0000792 of the unit of electro-magnetic intensity.
12 Mr. Warwick describes his experiments as follows: ‘The magnets employed were nearly of the usual size, 1½ inch in diameter, and nearly eight times as long. At first I employed iron disks, but I found them to be unnecessary. When I had discarded them, I tried several substances: first a thin disk of iron, which answered perfectly both for sender and receiver. A disk of sheet iron, about ⅒ of an inch in thickness, did not act so well, but all that was said was quite understood. In making experiments with the disks, I simply placed them above the instrument, without fixing them in any way: the wooden cover and the conical cavity were also laid aside, because the transmission and reception could be effected as well without them. This part of the instrument seems to be superfluous, since, when the disk is simply placed level to the ear, the sound seems to be increased by being brought nearer. Although iron acts better than anything, it appears that iron disks are not absolutely necessary, and that diamagnetic substances also act perfectly. I wished that my assistant, who was at some distance, and could not hear any direct sound, should continue his calculations. I took away the iron disk and placed across the instrument a wide iron bar, an inch thick. On applying my ear to it, I could hear every sound distinctly, but somewhat more faintly. A piece of copper, three inches square, was substituted for it: although the sound was still distinct, it was fainter than before. Thick pieces of lead, zinc, and steel were alternately tried. The steel acted in almost the same way as the iron, and, as in the other cases, each word was heard faintly but distinctly. Some of these metals are diamagnetic, and yet the action took place. Some non-metallic substances were next tried; first a piece of window-glass, which acted very well. The action was faint with a piece of a wooden match-box; but on using pieces of gradually increasing thickness the sound was sensibly increased, and with a piece of solid wood, 1½ inch in thickness, the sound was perfectly distinct. I next replaced it by an empty wooden box, which acted very well. A piece of cork, ½ inch thick, acted, but somewhat faintly. A block of razor-stone, 2 inches thick, was placed upon the instrument; and, on applying the ear to it, it was quite easy to follow the speaker. I then tried to hear without the insertion of any substance, and, on applying my ear quite close to the coil and magnet, I heard a faint sound, and on listening attentively I understood all that was said. In all these experiments the sounds were perceived, but the sounds transmitted or attempted did not act precisely alike. The sound of a tuning-fork, placed on the iron disk itself or on the case of the instrument, was clearly heard: thin iron disks were more effective for articulate speech. With other substances, stone, solid wood, glass, zinc, &c., the sound of the tuning-fork was heard, whether it rested upon them, or the vibrating fork was held above them. These substances were not adapted for transmitting the sound of the voice. These were all laid aside, and the sounding instrument was held directly above the pole of the magnet: the sound was clearly heard, although there was nothing but air between the end of the magnet and the tuning-fork. The sound was perhaps less intense when the tuning-fork was held directly above the pole, than when it was at the end of the magnet. I next tried if my voice could be heard with this arrangement. The result was rather doubtful, but I think that some action must have taken place, for the tuning-fork was heard when it was simply vibrated near the pole. The effect of the voice can only have differed in the degree of intensity: it was too faint to be heard at the other extremity. I repeated these effects; I assured myself of them, and I succeeded in transmitting sounds distinctly on the pole without a disk, and, on the other hand, by applying my ear to the instrument, I was able to hear distinctly all that was said, although there was no disk.’
13 These are his own words: ‘The articulation produced from the instrument was remarkably clear, but its great defect consisted in the fact that it could not be used as a sending instrument, and thus two telephones were required at each station, one for transmitting and one for receiving spoken messages.’
14 These carbons are made by heating, in a temperature gradually raised to white heat, fragments of deal of a close fibre, which is enclosed in an iron tube or box hermetically sealed.
15 Mr. Willoughby Smith varied this experiment by placing a packet of silk threads coated with copper on the disconnected ends of the circuit, which were arranged at right angles with each other. Under these conditions the instrument became so sensitive, that the current of air produced by a lamp placed above the system, caused a decided crackling noise in the telephone.
16 Mr. Hughes observes on this subject that carbon is a valuable material for such purposes, since it does not oxidise, and its effects are greater when combined with mercury. He takes the prepared charcoal used by artists, brings it to a white heat, and suddenly plunges it in a bath of mercury, of which the globules instantly penetrate the pores of charcoal, and may be said to metallise it. He also tried charcoal coated with a deposit of platinum, or impregnated with chloride of platinum, but this was not more successful than the former method. If the charcoal of fir-wood is brought to a white heat in an iron tube, containing tin and zinc, or any other metal which readily evaporates, it is metallised, and is adapted for use if the metal is subdivided in the pores of charcoal and not combined with it. When iron is introduced into carbon in this way it is one of the most effective metals. The charcoal of fir-wood, in itself a bad conductor, may thus acquire great conducting power.
17 Mr. Hughes remarks that the vibrations which affect the microphone, even in speaking at a distance from the instrument, do not proceed from the direct action of the sound waves on the contacts of the microphone, but from the molecular vibrations produced by it on the board which serves to support the instrument; he shows, in fact, that the intensity of sounds produced by the microphone is in proportion to the size of the surface of this board, and when the sending microphone is enclosed in a cylindrical case, its sensitiveness is not much diminished if the surface of the box enclosing the whole is sufficiently large. From this point of view he has sought to increase the sensitiveness of his instruments by fixing the frame on which the moveable parts of the sender and receiver revolve on a spring plate.
18 Helmholtz’s resonator is based upon the principle that a volume of air contained in an open vase emits a certain note when placed in vibration, and that the height of the note depends on the size of the vase and of its opening. Helmholtz makes use of a globe with a large opening on one side and a small one on the other, and the small one is applied to the ear. If a series of musical notes take place in the air, the one which is in harmony with the fundamental note of the globe is intensified, and can be distinguished from the rest. The same effect takes place when, on singing to a piano accompaniment, some strings are heard to vibrate more strongly than others, namely, those which vibrate in unison with the sounds emitted. The resonators are made in various ways; those most generally used are cases of different lengths which also serve as sounding-boxes.
19 I give the text of M. Cros’ sealed paper, opened by his request, at the Académie des Sciences, December 3, 1877:—‘Speaking generally, my process consists in obtaining traces of the movement to and fro of a vibrating membrane, and in using this tracing to reproduce the same movements, with their intrinsic relations of duration and intensity, either on the same membrane or on one adapted to give out the sounds which result from this series of movements.
‘It is therefore necessary that an extremely delicate tracing, such as may be obtained by passing a needle over a surface blackened by fire, should be transformed into a tracing, capable of sufficient resistance to guide an index which will transmit its movements to the membrane of sound.
‘A light index is fastened to the centre of a vibrating membrane; it terminates in a point (a metallic wire or tip of a feather) which rests on a surface which has been blackened by fire. This surface forms part of a disk, to which the double action of rotation and rectilinear progression has been given. If the membrane is at rest, the point will trace a simple spiral; if the membrane vibrates, there will be undulations in the spiral, and these undulations will represent the precise movements of the membrane in their duration and intensity.
‘By a well-known photographic process a transparent tracing of the undulations of the spiral can be represented by a line of similar dimensions on some resisting substance, tempered steel for example.
‘When this is done, this resisting surface is placed in a turning machine which causes it to revolve and advance with a velocity and motion similar to those by which the registering surface was actuated. A metallic point if the tracing is concave, or a grooved index if it is in relief, is kept upon the tracing by a spring, and the index which supports this point is connected with the centre of the membrane which produces the sounds. Under these conditions, the membrane will be actuated not by the vibrating air, but by the tracing which guides the index, and the impulses will be precisely similar in duration and intensity to those to which the registering membrane was subjected.
‘The spiral tracing represents equal successions of time by increasing or decreasing lengths. There is no inconvenience in this, since the turns of the spiral are very close together, if only the circumference of the turning circle is used; but then the central surface is lost.
‘In all cases the tracing of the helix on a cylinder is much more satisfactory, and I am now trying to make this idea practicable.’
20 Never make a contact between the stylus and the cylinder until the latter is covered with the tinfoil. Do not begin to turn the cylinder until assured that everything is in its place. Take care, when the stylus returns to the point of departure, to bring the mouthpiece forward. Always leave a margin of from five to ten millimètres on the left and at the beginning of the sheet of tinfoil; for if the stylus describes the curve on the extreme edge of the cylinder, it may tear the sheet or come out of the groove. Be careful not to detach the spring of the caoutchouc pad.
To fix the tinfoil, apply varnish to the end with a paint-brush; take this end between the finger and thumb of the left hand, with the sticky part towards the cylinder; raise it with the right hand and apply it quite smoothly to the cylinder; bring round the sticky end, and join them firmly.
To adjust the stylus and place it in the centre of the groove, bring the cylinder to the right, so as to place the stylus opposite the left extremity of the tinfoil; bring forward the cylinder gently and by degrees, until the stylus touches the tinfoil with force enough to imprint a mark. Observe if this mark is quite in the centre of the groove (in order to do this, make a mark with the nail across the cylinder), and if it is not, adjust the stylus to the right or left by means of the little screw placed above the mouthpiece. The depth of the impression made by the stylus should be ⅓ millimètre, just enough for it to leave a slight tracing, whatever the range of vibrations may be.
To reproduce the words, the winch must be turned with the same velocity as when they were inscribed. The average velocity should be about eighty turns a minute.
In speaking, the lips must touch the mouthpiece, and deep guttural sounds are better heard than those which are shrill. In reproducing, the tightening screw must be loosened and brought in front of the mouthpiece, the cylinder must be brought back to its point of departure, the contact between the stylus and the foil must be renewed, and the cylinder must again be turned in the same direction as when the sentence was spoken.
To increase the volume of reproduced sound, a tube of cardboard, wood, or horn may be applied to the mouthpiece; it must be of a conical form, and its lower end should be rather larger than the opening of the mouthpiece.
The stylus consists of a No. 9 needle, somewhat flattened on its two sides by friction on an oiled stone. The caoutchouc pad which connects the plate with the disk serves to weaken the vibrations of the plate. If this pad should come off, heat the head of a small nail, apply it to the wax which fastens the pad to the plate or to the spring, so as to soften it; then press the caoutchouc lightly, until it adheres to the place from which it was detached. The pads must be renewed from time to time, as they lose their elasticity. Care must be taken in replacing them not to injure the vibrating plate, either by too strong a pressure or by grazing it with the instrument employed to fix the pad.
The first experiments should be with single words or very short sentences, which can be extended as the ear becomes accustomed to the instrument’s peculiar tone.
The tone is varied by accelerating or slackening the rotatory movement of the cylinder. The cries of animals may be imitated. Instrumental music may be reproduced by placing a cardboard tube before the mouthpiece. The airs should be played in rapid time, since, when there is no system of clockwork, they will be more perfectly reproduced than those which are played slowly.
21 We confess that we find it difficult to believe in this property of the phonograph, from which we have only heard the harsh and unpleasant voice of Punch.
22 The action of this pedal is effected by two little rockers, so connected that the upper damper is lowered a little before the lower damper is raised—a condition necessary to produce the quivering motion of the plate which furnishes the rolling r.
23 The arrangement of this part of the instrument is remarkable in this particular, that in the case of certain letters the air is ejected with more or less force through the pipe I, while in the case of other letters the air is drawn into the same tube. Since I was unable to see the internal arrangement of these cavities, I can only give an imperfect account of the mechanism at work.