This inventive period extended over a number of years, during which time he took out forty-six patents on stock-printing instruments and devices, two of such patents being issued to Edison and Pope as joint inventors.

These various inventions were mostly in the line of development of the art as it progressed during those early years, but out of it all came the Edison universal printer, which entered into very extensive use, and which is still used throughout the United States and in some foreign countries to a considerable extent at this very day.

Edison's inventive work on stock printers has left its mark upon the art as it exists at the present time. In his earlier work he directed his attention to the employment of a single-circuit system, in which only one wire was required, the two operations of setting the type-wheels and of printing being controlled by separate electromagnets which were actuated through polarized relays, as occasion required, one polarity energizing the electromagnet controlling the type-wheels, and the opposite polarity energizing the electromagnet controlling the printing.

Later on, however, he changed over to a two-wire circuit, such as shown in Fig. 2 of this article in connection with the universal stock printer. In the earliest days of the stock printer, Edison realized the vital commercial importance of having all instruments recording precisely alike at the same moment, and it was he who first devised (in 1869) the "unison stop," by means of which all connected instruments could at any moment be brought to zero from the central transmitting station, and thus be made to work in correspondence with the central instrument and with one another. He also originated the idea of using only one inking-pad and shifting it from side to side to ink the type-wheels. It was also in Edison's stock printer that the principle of shifting type-wheels was first employed. Hence it will be seen that, as in many other arts, he made a lasting impression in this one by the intrinsic merits of the improvements resulting from his work therein.

We shall not attempt to digest the forty-six patents above named, nor to follow Edison through the progressive steps which led to the completion of his universal printer, but shall simply present a sketch of the instrument itself, and follow with a very brief and general explanation of its theory. The Edison universal printer, as it virtually appears in practice, is ill.u.s.trated in Fig. 1 below, from which it will be seen that the most prominent parts are the two type-wheels, the inking-pad, and the paper tape feeding from the reel, all appropriately placed in a substantial framework.

The electromagnets and other actuating mechanism cannot be seen plainly in this figure, but are produced diagrammatically in Fig. 2, and somewhat enlarged for convenience of explanation.

It will be seen that there are two electromagnets, one of which, TM, is known as the "type-magnet," and the other, PM, as the "press-magnet,"

the former having to do with the operation of the type-wheels, and the latter with the pressing of the paper tape against them. As will be seen from the diagram, the armature, A, of the type-magnet has an extension arm, on the end of which is an escapement engaging with a toothed wheel placed at the extremity of the shaft carrying the type-wheels. This extension arm is pivoted at B. Hence, as the armature is alternately attracted when current pa.s.ses around its electromagnet, and drawn up by the spring on cessation of current, it moves up and down, thus actuating the escapement and causing a rotation of the toothed wheel in the direction of the arrow. This, in turn, brings any desired letters or figures on the type-wheels to a central point, where they may be impressed upon the paper tape. One type-wheel carries letters, and the other one figures. These two wheels are mounted rigidly on a sleeve carried by the wheel-shaft. As it is desired to print from only one type-wheel at a time, it becomes necessary to shift them back and forth from time to time, in order to bring the desired characters in line with the paper tape. This is accomplished through the movements of a three-arm rocking-lever attached to the wheel-sleeve at the end of the shaft. This lever is actuated through the agency of two small pins carried by an arm projecting from the press-lever, PL. As the latter moves up and down the pins play upon the under side of the lower arm of the rocking-lever, thus canting it and pushing the type-wheels to the right or left, as the case may be. The operation of shifting the type-wheels will be given further on.

The press-lever is actuated by the press-magnet. From the diagram it will be seen that the armature of the latter has a long, pivoted extension arm, or platen, trough-like in shape, in which the paper tape runs. It has already been noted that the object of the press-lever is to press this tape against that character of the type-wheel centrally located above it at the moment. It will at once be perceived that this action takes place when current flows through the electromagnet and its armature is attracted downward, the platen again dropping away from the type-wheel as the armature is released upon cessation of current. The paper "feed" is shown at the end of the press-lever, and consists of a push "dog," or pawl, which operates to urge the paper forward as the press-lever descends.

The worm-gear which appears in the diagram on the shaft, near the toothed wheel, forms part of the unison stop above referred to, but this device is not shown in full, in order to avoid unnecessary complications of the drawing.

At the right-hand side of the diagram (Fig. 2) is shown a portion of the transmitting apparatus at a central office. Generally speaking, this consists of a motor-driven cylinder having metallic pins placed at intervals, and arranged spirally, around its periphery. These pins correspond in number to the characters on the type-wheels. A keyboard (not shown) is arranged above the cylinder, having keys lettered and numbered corresponding to the letters and figures on the type-wheels.

Upon depressing any one of these keys the motion of the cylinder is arrested when one of its pins is caught and held by the depressed key.

When the key is released the cylinder continues in motion. Hence, it is evident that the revolution of the cylinder may be interrupted as often as desired by manipulation of the various keys in transmitting the letters and figures which are to be recorded by the printing instrument.

The method of transmission will presently appear.

In the sketch (Fig. 2) there will be seen, mounted upon the cylinder shaft, two wheels made up of metallic segments insulated from each other, and upon the hubs of these wheels are two brushes which connect with the main battery. Resting upon the periphery of these two segmental wheels there are two brushes to which are connected the wires which carry the battery current to the type-magnet and press-magnet, respectively, as the brushes make circuit by coming in contact with the metallic segments. It will be remembered that upon the cylinder there are as many pins as there are characters on the type-wheels of the ticker, and one of the segmental wheels, W, has a like number of metallic segments, while upon the other wheel, W', there are only one-half that number. The wheel W controls the supply of current to the press-magnet, and the wheel W' to the type-magnet. The type-magnet advances the letter and figure wheels one step when the magnet is energized, and a succeeding step when the circuit is broken. Hence, the metallic contact surfaces on wheel W' are, as stated, only half as many as on the wheel W, which controls the press-magnet.

It should be borne in mind, however, that the contact surfaces and insulated surfaces on wheel W' are together equal in number to the characters on the type-wheels, but the retractile spring of TM does half the work of operating the escapement. On the other hand, the wheel W has the full number of contact surfaces, because it must provide for the operative closure of the press-magnet circuit whether the brush B' is in engagement with a metallic segment or an insulated segment of the wheel W'. As the cylinder revolves, the wheels are carried around with its shaft and current impulses flow through the wires to the magnets as the brushes make contact with the metallic segments of these wheels.

One example will be sufficient to convey to the reader an idea of the operation of the apparatus. a.s.suming, for instance, that it is desired to send out the letters AM to the printer, let us suppose that the pin corresponding to the letter A is at one end of the cylinder and near the upper part of its periphery, and that the letter M is about the centre of the cylinder and near the lower part of its periphery. The operator at the keyboard would depress the letter A, whereupon the cylinder would in its revolution bring the first-named pin against the key. During the rotation of the cylinder a current would pa.s.s through wheel W' and actuate TM, drawing down the armature and operating the escapement, which would bring the type-wheel to a point where the letter A would be central as regards the paper tape When the cylinder came to rest, current would flow through the brush of wheel W to PM, and its armature would be attracted, causing the platen to be lifted and thus bringing the paper tape in contact with the type-wheel and printing the letter A.

The operator next sends the letter M by depressing the appropriate key.

On account of the position of the corresponding pin, the cylinder would make nearly half a revolution before bringing the pin to the key. During this half revolution the segmental wheels have also been turning, and the brushes have transmitted a number of current impulses to TM, which have caused it to operate the escapement a corresponding number of times, thus turning the type-wheels around to the letter M. When the cylinder stops, current once more goes to the press-magnet, and the operation of lifting and printing is repeated. As a matter of fact, current flows over both circuits as the cylinder is rotated, but the press-magnet is purposely made to be comparatively "sluggish" and the narrowness of the segments on wheel W tends to diminish the flow of current in the press circuit until the cylinder comes to rest, when the current continuously flows over that circuit without interruption and fully energizes the press-magnet. The shifting of the type-wheels is brought about as follows: On the keyboard of the transmitter there are two characters known as "dots"--namely, the letter dot and the figure dot. If the operator presses one of these dot keys, it is engaged by an appropriate pin on the revolving cylinder. Meanwhile the type-wheels are rotating, carrying with them the rocking-lever, and current is pulsating over both circuits. When the type-wheels have arrived at the proper point the rocking-lever has been carried to a position where its lower arm is directly over one of the pins on the arm extending from the platen of the press-lever. The cylinder stops, and current operates the sluggish press-magnet, causing its armature to be attracted, thus lifting the platen and its projecting arm. As the arm lifts upward, the pin moves along the under side of the lower arm of the rocking-lever, thus causing it to cant and shift the type-wheels to the right or left, as desired. The principles of operation of this apparatus have been confined to a very brief and general description, but it is believed to be sufficient for the scope of this article.

NOTE.--The ill.u.s.trations in this article are reproduced from American Telegraphy and Encyclopedia of the Telegraph, by William Maver, Jr., by permission of Maver Publishing Company, New York.

II. THE QUADRUPLEX AND PHONOPLEX

EDISON'S work in stock printers and telegraphy had marked him as a rising man in the electrical art of the period but his invention of quadruplex telegraphy in 1874 was what brought him very prominently before the notice of the public. Duplex telegraphy, or the sending of two separate messages in opposite directions at the same time over one line was known and practiced previous to this time, but quadruplex telegraphy, or the simultaneous sending of four separate messages, two in each direction, over a single line had not been successfully accomplished, although it had been the subject of many an inventor's dream and the object of anxious efforts for many long years.

In the early part of 1873, and for some time afterward, the system invented by Joseph Stearns was the duplex in practical use. In April of that year, however, Edison took up the study of the subject and filed two applications for patents. One of these applications [23] embraced an invention by which two messages could be sent not only duplex, or in opposite directions as above explained, but could also be sent "diplex"--that is to say, in one direction, simultaneously, as separate and distinct messages, over the one line. Thus there was introduced a new feature into the art of multiplex telegraphy, for, whereas duplexing (accomplished by varying the strength of the current) permitted messages to be sent simultaneously from opposite stations, diplexing (achieved by also varying the direction of the current) permitted the simultaneous transmission of two messages from the same station and their separate reception at the distant station.

[Footnote 23: Afterward issued as Patent No. 162,633, April 27, 1875.]

The quadruplex was the tempting goal toward which Edison now constantly turned, and after more than a year's strenuous work he filed a number of applications for patents in the late summer of 1874. Among them was one which was issued some years afterward as Patent No. 480,567, covering his well-known quadruplex. He had improved his own diplex, combined it with the Stearns duplex and thereby produced a system by means of which four messages could be sent over a single line at the same time, two in each direction.

As the reader will probably be interested to learn something of the theoretical principles of this fascinating invention, we shall endeavor to offer a brief and condensed explanation thereof with as little technicality as the subject will permit. This explanation will necessarily be of somewhat elementary character for the benefit of the lay reader, whose indulgence is asked for an occasional reiteration introduced for the sake of clearness of comprehension. While the apparatus and the circuits are seemingly very intricate, the principles are really quite simple, and the difficulty of comprehension is more apparent than real if the underlying phenomena are studied attentively.

At the root of all systems of telegraphy, including multiplex systems, there lies the single basic principle upon which their performance depends--namely, the obtaining of a slight mechanical movement at the more or less distant end of a telegraph line. This is accomplished through the utilization of the phenomena of electromagnetism. These phenomena are easy of comprehension and demonstration. If a rod of soft iron be wound around with a number of turns of insulated wire, and a current of electricity be sent through the wire, the rod will be instantly magnetized and will remain a magnet as long as the current flows; but when the current is cut off the magnetic effect instantly ceases. This device is known as an electromagnet, and the charging and discharging of such a magnet may, of course, be repeated indefinitely.

Inasmuch as a magnet has the power of attracting to itself pieces of iron or steel, the basic importance of an electromagnet in telegraphy will be at once apparent when we consider the sounder, whose clicks are familiar to every ear. This instrument consists essentially of an electro-magnet of horseshoe form with its two poles close together, and with its armature, a bar of iron, maintained in close proximity to the poles, but kept normally in a retracted position by a spring. When the distant operator presses down his key the circuit is closed and a current pa.s.ses along the line and through the (generally two) coils of the electromagnet, thus magnetizing the iron core. Its attractive power draws the armature toward the poles. When the operator releases the pressure on his key the circuit is broken, current does not flow, the magnetic effect ceases, and the armature is drawn back by its spring.

These movements give rise to the clicking sounds which represent the dots and dashes of the Morse or other alphabet as transmitted by the operator. Similar movements, produced in like manner, are availed of in another instrument known as the relay, whose office is to act practically as an automatic transmitter key, repeating the messages received in its coils, and sending them on to the next section of the line, equipped with its own battery; or, when the message is intended for its own station, sending the message to an adjacent sounder included in a local battery circuit. With a simple circuit, therefore, between two stations and where an intermediate battery is not necessary, a relay is not used.

Pa.s.sing on to the consideration of another phase of the phenomena of electromagnetism, the reader's attention is called to Fig. 1, in which will be seen on the left a simple form of electromagnet consisting of a bar of soft iron wound around with insulated wire, through which a current is flowing from a battery. The arrows indicate the direction of flow.

All magnets have two poles, north and south. A permanent magnet (made of steel, which, as distinguished from soft iron, retains its magnetism for long periods) is so called because it is permanently magnetized and its polarity remains fixed. In an electromagnet the magnetism exists only as long as current is flowing through the wire, and the polarity of the soft-iron bar is determined by the DIRECTION of flow of current around it for the time being. If the direction is reversed, the polarity will also be reversed. a.s.suming, for instance, the bar to be end-on toward the observer, that end will be a south pole if the current is flowing from left to right, clockwise, around the bar; or a north pole if flowing in the other direction, as ill.u.s.trated at the right of the figure. It is immaterial which way the wire is wound around the bar, the determining factor of polarity being the DIRECTION of the current. It will be clear, therefore, that if two EQUAL currents be pa.s.sed around a bar in opposite directions (Fig. 3) they will tend to produce exactly opposite polarities and thus neutralize each other. Hence, the bar would remain non-magnetic.

As the path to the quadruplex pa.s.ses through the duplex, let us consider the Stearns system, after noting one other principle--namely, that if more than one path is presented in which an electric current may complete its circuit, it divides in proportion to the resistance of each path. Hence, if we connect one pole of a battery with the earth, and from the other pole run to the earth two wires of equal resistance as ill.u.s.trated in Fig. 2, equal currents will traverse the wires.

The above principles were employed in the Stearns differential duplex system in the following manner: Referring to Fig. 3, suppose a wire, A, is led from a battery around a bar of soft iron from left to right, and another wire of equal resistance and equal number of turns, B, around from right to left. The flow of current will cause two equal opposing actions to be set up in the bar; one will exactly offset the other, and no magnetic effect will be produced. A relay thus wound is known as a differential relay--more generally called a neutral relay.

The non-technical reader may wonder what use can possibly be made of an apparently non-operative piece of apparatus. It must be borne in mind, however, in considering a duplex system, that a differential relay is used AT EACH END of the line and forms part of the circuit; and that while each relay must be absolutely unresponsive to the signals SENT OUT FROM ITS HOME OFFICE, it must respond to signals transmitted by a DISTANT OFFICE. Hence, the next figure (4), with its accompanying explanation, will probably make the matter clear. If another battery, D, be introduced at the distant end of the wire A the differential or neutral relay becomes actively operative as follows: Battery C supplies wires A and B with an equal current, but battery D doubles the strength of the current traversing wire A. This is sufficient to not only neutralize the magnetism which the current in wire B would tend to set up, but also--by reason of the excess of current in wire A--to make the bar a magnet whose polarity would be determined by the direction of the flow of current around it.

In the arrangement shown in Fig. 4 the batteries are so connected that current flow is in the same direction, thus doubling the amount of current flowing through wire A. But suppose the batteries were so connected that the current from each set flowed in an opposite direction? The result would be that these currents would oppose and neutralize each other, and, therefore, none would flow in wire A.

Inasmuch, however, as there is nothing to hinder, current would flow from battery C through wire B, and the bar would therefore be magnetized. Hence, a.s.suming that the relay is to be actuated from the distant end, D, it is in a sense immaterial whether the batteries connected with wire A a.s.sist or oppose each other, as, in either case, the bar would be magnetized only through the operation of the distant key.

A slight elaboration of Fig. 4 will further ill.u.s.trate the principle of the differential duplex. In Fig. 5 are two stations, A the home end, and B the distant station to which a message is to be sent. The relay at each end has two coils, 1 and 2, No. 1 in each case being known as the "main-line coil" and 2 as the "artificial-line coil." The latter, in each case, has in its circuit a resistance, R, to compensate for the resistance of the main line, so that there shall be no inequalities in the circuits. The artificial line, as well as that to which the two coils are joined, are connected to earth. There is a battery, C, and a key, K. When the key is depressed, current flows through the relay coils at A, but no magnetism is produced, as they oppose each other. The current, however, flows out through the main-line coil over the line and through the main-line coil 1 at B, completing its circuit to earth and magnetizing the bar of the relay, thus causing its armature to be attracted. On releasing the key the circuit is broken and magnetism instantly ceases.

It will be evident, therefore, that the operator at A may cause the relay at B to act without affecting his own relay. Similar effects would be produced from B to A if the battery and key were placed at the B end.

If, therefore, like instruments are placed at each end of the line, as in Fig. 6, we have a differential duplex arrangement by means of which two operators may actuate relays at the ends distant from them, without causing the operation of the relays at their home ends. In practice this is done by means of a special instrument known as a continuity preserving transmitter, or, usually, as a transmitter. This consists of an electromagnet, T, operated by a key, K, and separate battery. The armature lever, L, is long, pivoted in the centre, and is bent over at the end. At a point a little beyond its centre is a small piece of insulating material to which is screwed a strip of spring metal, S.

Conveniently placed with reference to the end of the lever is a bent metallic piece, P, having a contact screw in its upper horizontal arm, and attached to the lower end of this bent piece is a post, or standard, to which the main battery is electrically connected. The relay coils are connected by wire to the spring piece, S, and the armature lever is connected to earth. If the key is depressed, the armature is attracted and its bent end is moved upward, depressing the spring which makes contact with the upper screw, which places the battery to the line, and simultaneously breaks the ground connection between the spring and the upturned end of the lever, as shown at the left. When the key is released the battery is again connected to earth. The compensating resistances and condensers necessary for a duplex arrangement are shown in the diagram.

In Fig. 6 one transmitter is shown as closed, at A, while the other one is open. From our previous ill.u.s.trations and explanations it will be readily seen that, with the transmitter closed at station A, current flows via post P, through S, and to both relay coils at A, thence over the main line to main-line coil at B, and down to earth through S and the armature lever with its grounded wire. The relay at A would be unresponsive, but the core of the relay at B would be magnetized and its armature respond to signals from A. In like manner, if the transmitter at B be closed, current would flow through similar parts and thus cause the relay at A to respond. If both transmitters be closed simultaneously, both batteries will be placed to the line, which would practically result in doubling the current in each of the main-line coils, in consequence of which both relays are energized and their armatures attracted through the operation of the keys at the distant ends. Hence, two messages can be sent in opposite directions over the same line simultaneously.

The reader will undoubtedly see quite clearly from the above system, which rests upon varying the STRENGTH of the current, that two messages could not be sent in the same direction over the one line at the same time. To accomplish this object Edison introduced another and distinct feature--namely, the using of the same current, but ALSO varying its DIRECTION of flow; that is to say, alternately reversing the POLARITY of the batteries as applied to the line and thus producing corresponding changes in the polarity of another specially constructed type of relay, called a polarized relay. To afford the reader a clear conception of such a relay we would refer again to Fig. 1 and its explanation, from which it appears that the polarity of a soft-iron bar is determined not by the strength of the current flowing around it but by the direction thereof.

With this idea clearly in mind, the theory of the polarized relay, generally called "polar" relay, as presented in the diagram (Fig. 7), will be readily understood.

A is a bar of soft iron, bent as shown, and wound around with insulated copper wire, the ends of which are connected with a battery, B, thus forming an electromagnet. An essential part of this relay consists of a swinging PERMANENT magnet, C, whose polarity remains fixed, that end between the terminals of the electromagnet being a north pole. Inasmuch as unlike poles of magnets are attracted to each other and like poles repelled, it follows that this north pole will be repelled by the north pole of the electromagnet, but will swing over and be attracted by its south pole. If the direction of flow of current be reversed, by reversing the battery, the electromagnetic polarity also reverses and the end of the permanent magnet swings over to the other side. This is shown in the two figures of Fig. 7. This device being a relay, its purpose is to repeat transmitted signals into a local circuit, as before explained. For this purpose there are provided at D and E a contact and a back stop, the former of which is opened and closed by the swinging permanent magnet, thus opening and closing the local circuit.

Manifestly there must be provided some convenient way for rapidly transposing the direction of the current flow if such a device as the polar relay is to be used for the reception of telegraph messages, and this is accomplished by means of an instrument called a pole-changer, which consists essentially of a movable contact piece connected permanently to the earth, or grounded, and arranged to connect one or the other pole of a battery to the line and simultaneously ground the other pole. This action of the pole-changer is effected by movements of the armature of an electromagnet through the manipulation of an ordinary telegraph key by an operator at the home station, as in the operation of the "transmitter," above referred to.

By a combination of the neutral relay and the polar relay two operators, by manipulating two telegraph keys in the ordinary way, can simultaneously send two messages over one line in the SAME direction with the SAME current, one operator varying its strength and the other operator varying its polarity or direction of flow. This principle was covered by Edison's Patent No. 162,633, and was known as the "diplex"

system, although, in the patent referred to, Edison showed and claimed the adaptation of the principle to duplex telegraphy. Indeed, as a matter of fact, it was found that by winding the polar relay differentially and arranging the circuits and collateral appliances appropriately, the polar duplex system was more highly efficient than the neutral system, and it is extensively used to the present day.

Thus far we have referred to two systems, one the neutral or differential duplex, and the other the combination of the neutral and polar relays, making a diplex system. By one of these two systems a single wire could be used for sending two messages in opposite directions, and by the other in the same direction or in opposite directions. Edison followed up his work on the diplex and combined the two systems into the quadruplex, by means of which FOUR messages could be sent and received simultaneously over the one wire, two in each direction, thus employing eight operators--four at each end--two sending and two receiving. The general principles of quadruplex telegraphy are based upon the phenomena which we have briefly outlined in connection with the neutral relay and the polar relay. The equipment of such a system at each end of the line consists of these two instruments, together with the special form of transmitter and the pole-changer and their keys for actuating the neutral and polar relays at the other, or distant, end. Besides these there are the compensating resistances and condensers. All of these will be seen in the diagram (Fig. 8). It will be understood, of course, that the polar relay, as used in the quadruplex system, is wound differentially, and therefore its operation is somewhat similar in principle to that of the differentially wound neutral relay, in that it does not respond to the operation of the key at the home office, but only operates in response to the movements of the distant key.

Our explanation has merely aimed to show the underlying phenomena and principles in broad outline without entering into more detail than was deemed absolutely necessary. It should be stated, however, that between the outline and the filling in of the details there was an enormous amount of hard work, study, patient plodding, and endless experiments before Edison finally perfected his quadruplex system in the year 1874.

If it were attempted to offer here a detailed explanation of the varied and numerous operations of the quadruplex, this article would a.s.sume the proportions of a treatise. An idea of their complexity may be gathered from the following, which is quoted from American Telegraphy and Encyclopedia of the Telegraph, by William Maver, Jr.:

"It may well be doubted whether in the whole range of applied electricity there occur such beautiful combinations, so quickly made, broken up, and others reformed, as in the operation of the Edison quadruplex. For example, it is quite demonstrable that during the making of a simple dash of the Morse alphabet by the neutral relay at the home station the distant pole-changer may reverse its battery several times; the home pole-changer may do likewise, and the home transmitter may increase and decrease the electromotive force of the home battery repeatedly. Simultaneously, and, of course, as a consequence of the foregoing actions, the home neutral relay itself may have had its magnetism reversed several times, and the SIGNAL, that is, the dash, will have been made, partly by the home battery, partly by the distant and home batteries combined, partly by current on the main line, partly by current on the artificial line, partly by the main-line 'static'

current, partly by the condenser static current, and yet, on a well-adjusted circuit the dash will have been produced on the quadruplex sounder as clearly as any dash on an ordinary single-wire sounder."

We present a diagrammatic ill.u.s.tration of the Edison quadruplex, battery key system, in Fig. 8, and refer the reader to the above or other text-books if he desires to make a close study of its intricate operations. Before finally dismissing the quadruplex, and for the benefit of the inquiring reader who may vainly puzzle over the intricacies of the circuits shown in Fig. 8, a hint as to an essential difference between the neutral relay, as used in the duplex and as used in the quadruplex, may be given. With the duplex, as we have seen, the current on the main line is changed in strength only when both keys at OPPOSITE stations are closed together, so that a current due to both batteries flows over the main line. When a single message is sent from one station to the other, or when both stations are sending messages that do not conflict, only one battery or the other is connected to the main line; but with the quadruplex, suppose one of the operators, in New York for instance, is sending reversals of current to Chicago; we can readily see how these changes in polarity will operate the polar relay at the distant station, but why will they not also operate the neutral relay at the distant station as well? This difficulty was solved by dividing the battery at each station into two unequal parts, the smaller battery being always in circuit with the pole-changer ready to have its polarity reversed on the main line to operate the distant polar relay, but the spring retracting the armature of the neutral relay is made so stiff as to resist these weak currents. If, however, the transmitter is operated at the same end, the entire battery is connected to the main line, and the strength of this current is sufficient to operate the neutral relay. Whether the part or all the battery is alternately connected to or disconnected from the main line by the transmitter, the current so varied in strength is subject to reversal of polarity by the pole-changer; but the variations in strength have no effect upon the distant polar relay, because that relay being responsive to changes in polarity of a weak current is obviously responsive to corresponding changes in polarity of a powerful current. With this distinction before him, the reader will have no difficulty in following the circuits of Fig. 8, bearing always in mind that by reason of the differential winding of the polar and neutral relays, neither of the relays at one station will respond to the home battery, and can only respond to the distant battery--the polar relay responding when the polarity of the current is reversed, whether the current be strong or weak, and the neutral relay responding when the line-current is increased, regardless of its polarity. It should be added that besides the system ill.u.s.trated in Fig. 8, which is known as the differential principle, the quadruplex was also arranged to operate on the Wheatstone bridge principle; but it is not deemed necessary to enter into its details. The underlying phenomena were similar, the difference consisting largely in the arrangement of the circuits and apparatus. [24]

[Footnote 24: Many of the ill.u.s.trations in this article are reproduced from American Telegraphy and Encyclopedia of the Telegraph, by William Maver, Jr., by permission of Maver Publishing Company, New York.]

Edison made another notable contribution to multiplex telegraphy some years later in the Phonoplex. The name suggests the use of the telephone, and such indeed is the case. The necessity for this invention arose out of the problem of increasing the capacity of telegraph lines employed in "through" and "way" service, such as upon railroads. In a railroad system there are usually two terminal stations and a number of way stations. There is naturally much intercommunication, which would be greatly curtailed by a system having the capacity of only a single message at a time. The duplexes above described could not be used on a railroad telegraph system, because of the necessity of electrically balancing the line, which, while entirely feasible on a through line, would not be practicable between a number of intercommunicating points.

Edison's phonoplex normally doubled the capacity of telegraph lines, whether employed on way business or through traffic, but in actual practice made it possible to obtain more than double service. It has been in practical use for many years on some of the leading railroads of the United States.