the principal valuable positive conclusions have been derived from linguistic researches. In the dispute as to the relative merits of the zoological and philological methods in ethnology, I accordingly side with the advocates of the latter; and, in regard to the special subject of this paper, I say with Quatrefages, in the words of Virgil, " Ne crede colori."

THE CARBON BUTTON.*

By E. A. ENGLER, A. M.

ALTHOUGH the telephone seems to have sprung up at which

very suddenly, there have been steps in its development which show that the difficulties encountered in devising a means for the transmission of articulate speech have not been overcome altogether by a single stroke of individual genius, but singly by the patient and, for the most part, unrewarded labor of many. Each stage of its development was the outgrowth of suggestions obtained from previous experiments. Of the instruments which served their purpose in the discovery of the properties of the carbon button, a brief description will be given in this paper.

Sound is known to be produced by vibrations, generally of air; differences between sounds are due to differences in vibration. There are but three essential characteristics to be noted, all dependent upon the vibrations of the air: 1. The pitch, by virtue of which a sound is called high or low, and which depends upon the number or rapidity of the vibrations; 2. The intensity or loudness, which is determined by the amplitude of the vibration; 3. The quality by which we distinguish the corresponding tones of different instruments, and which depends on the form of the vibration. In order to obtain an exact reproduction of any sound, its pitch, intensity, and quality must be exactly reproduced; and, to render this possible, the rapidity, amplitude, and form of the vibration must be exactly reproduced.

For producing sound at a distant place two methods suggest themselves 1. Actually to transmit the sound vibrations through the air; this is the method employed in the speaking tube. 2. To reproduce the sound vibrations at the distant station; this is the method employed in the telephone. The previous development of the telegraph naturally suggested electricity as the agent to carry the vibrations from one place to another. It thus became necessary to convert sound waves into electric waves and vice versa, and experiments look

This paper, at first intended for a special occasion, has been published at the suggestion of several friends. In its preparation, use has been made of information to be found in George B. Prescott's work on the telephone, and in the journals of science. Most of the illustrations are from Prescott's work.

ing to the accomplishment of that end were begun nearly twenty years ago.

The first successful experiments were made by Philip Reis, of Fredericksdorf, Germany, in 1861. He argued that if it could be found practicable to convert sound pulsations into electric pulsations, and then convert these pulsations back again into sound pulsations, the same effect would be produced as if the vibrations had been actually transmitted through the air. In his instruments a membrane rigidly secured on the sides was caused to vibrate in the center by the motion of the air produced by any sound. In the center of this membrane was a delicate circuit-breaker so arranged as to break the circuit of an electric telegraph line at every vibration, thus successively magnetizing and demagnetizing an electro-magnet at the receiving station, and causing its armature to vibrate in accordance with the vibrations of the membrane at the transmitting station. The vibrations of this armature, properly mounted on a sounding-board, set into vibration the surrounding air, which carried the sound to the ear. His first instrument is represented in Fig. 1. A is the transmitting and B the receiving instrument, supposed to be placed at different stations and connected with each other by a metallic conductor. A conical tube, a b, six inches long, four inches in diameter at the larger, and one and a half inch in diameter at the smaller end, is closed at b by a collodion membrane o, against the center of which rests one end, c, of the lever c d. This lever has electric connection with the wire of the line joining the two stations at its point of support, e. The end d of the lever rests against the flat spring g, which can be properly adjusted by means of the screw h, and which, through the metal standard ƒ, is connected with the battery C. At station B the conducting wire passed around the electro-magnet m, which is mounted on a sounding-box W; thence to the ground. Attached to the armature at the pole of the magnet is a thin plate i, which is hung on an horizontal axis projecting from the upright k; the motion of the plate can be regulated by the screw and the spring s. The best dimensions and most suitable adjustments of the instrument were determined by experiment. Its operation is as follows: When at rest the small spring n keeps the lever c d in contact at g, the circuit is closed, and the magnet m attracts the armature i. But, when by speaking into the tube a b, the air in the tube and therefore the membrane o is set into vibration, the contact at g is alternately broken and closed, and consequently the magnet at B is demagnetized and magnetized, alternately releasing and attracting the armature i. It is evident that the vibrations of i correspond in number and interval to the vibrations of the membrane o; so that the sound which enters the tube a b is reproduced at B so far as its pitch is concerned. But as the strength of the current is constant, neither the intensity nor the quality of the sound is reproduced.

In 1874 Elisha Gray, of Chicago, accomplished the reproduction of

intensity and quality as well as pitch of sound by means of an instrument in which the strength of the current could be varied in exact accordance with the tone to be transmitted, and was thus enabled to

reproduce any number of tones simultaneously without losing their specific character-a thing plainly impossible with the Reis instrument. The device used is shown in Fig. 2. T, is a mouthpiece into which the person transmitting sounds speaks. D, is a tense thin diaphragm connected with the line joining the two stations. To the center of the diaphragm is fastened one end of a metal rod N, whose other end dips into a jar J containing acidulated water. A metal plug

p at the bottom of this jar is connected with one wire of the battery E, the other going to the ground. At the receiving station the wire simply passes over an electro-magnet H, thence to the ground. Close to H is placed the diaphragm D, properly provided at its center with a metal plate which serves as armature for the electro-magnet, and fastened at its circumference in the holder T. The action of the in

strument is as follows: The person sending the message speaks into the mouthpiece T1, thus causing the diaphragm D1, with the plunger N, to vibrate. The greater the amplitude of vibration the deeper the rod N descends into the liquid, and therefore the thinner the stratum of liquid through which the current will have to pass; thus the resistance to the passage of the current is varied inversely as the intensity of the sound. At the receiving station the current magnetizes the electromagnet H, and thus reproduces in the diaphragm D the vibrations of the diaphragm D1.

A number of telephones have since been invented, differing from each other in method of application and details of construction, but all embodying the scientific principle used by Gray.

Using the instrument invented by Reis, and the suggestions which Gray's experiments afforded, Thomas A. Edison began his attempts to construct a new form of telephone. Inasmuch as his experiments in this direction" cover many thousand pages of manuscript," only a few of the more characteristic ones will be given.

In the Reis transmitter a platinum screw was made to face the diaphragm, and a drop of water was put between them. The only result, however, was the decomposition of the water and the deposit of a sediment on the platinum. Two disks of platinum, one on the diaphragm and the other on the screw, so placed as to hold several drops of water by capillary attraction, were then tried. Acidulated solutions were substituted for water; paper and other materials, saturated with various solutions, were tried; sharp edges were substituted for disks. The result of all these experiments was complete failure, on

account of the decomposition of the fluids. These were therefore abandoned and the attempt was made to vary the strength of the current by the use of platinum points, springs, and other devices. The number of these points which was to be brought into the electric circuit was to be dependent upon the amplitude of the vibration, and thus the resistance of the circuit was to be varied inversely as the intensity of the sound producing the vibration. All of these contrivances were of no avail. Subsequently plumbago and white Arkansas oil-stone were tried on account of their great resistance, and with these fair success was attained. Various expedients were used to make the portion of the material employed in the circuit proportional to the amplitude of vibration, but the confusion introduced by the devices themselves rendered the apparatus practically useless. All these experiments were conducted before the close of the year 1876.

In January of the next year the idea occurred to Mr. Edison to make use of the fact that semi-conductors vary their resistance with the pressure to which they are subjected-a thing which he had accidentally discovered while constructing some apparatus for artificial cables about four years before. He immediately set to work to construct an instrument. A diaphragm carrying at its center a spring faced with platinum was placed opposite to a small cup containing the semi-conductor to be tried. The adjustment was secured by means of a screw fastened to the cup. The vibrations of the diaphragm produced by the tones of the voice determined the pressure of the spring upon the semi-conductor. The materials first experimented upon were crude plumbago mixed with dry powders of different kinds. The results obtained were encouraging, the volume of sound being great, but the articulation so poor that some practice was necessary before the peculiar sound of the instrument could be caught with ease. An improvement was effected when, after much experimenting, solid materials were abandoned and tufts of gloss silk coated with semi-conductors were substituted. But, with all the improvement that could be devised, the instrument was still very inferior to the magneto-telephone of Professor Bell, and required such frequent adjustment as to make it very objectionable. Experiment developed the fact that the change in resistance in the semi-conductor, due to the impact of sound-vibrations, was very small, and, in order to make this change of resistance as important a factor as possible, Mr. Edison determined to make the resistance of his circuit very small to that end he tried the primary circuit of an induction-coil, but the experiment failed. The cause of failure was at first only a matter of conjecture; but, by trying one thing after another as they suggested themselves, without any very definite purpose, conjecture finally condensed into the belief that the resistance of the semi-conductor was too great to be used with the primary circuit of an induction-coil. The effort then was to reduce the resistance of the semi-conductor to a few ohms and still be able to