A later invention, the Siphon recorder, also due to Sir Wm. Thomson (Lord Kelvin) over comes this objection. This instrument, out lined in Fig. 15, consists of a coil of fine copper wire, A, which is suspended between the poles of a powerful magnet, M, in such a manner that when a current passes through the coil it, like the needles in the mirror galvanometer, tends to place itself at right angles to the lines of force of the magnetic field. When no cur rent is flowing in the coil, two small weights, w, suspended from the lower end of the coil, hold it in the plane of the magnetic lines of force of the magnet. A siphon C consisting of a very small glass tube is attached to the coil by a fine wire. The lower and bent end of the siphon is placed directly over the centre, or imaginary zero, of a paper ribbon P. The double pole-changing key K sets up momen tary pulsations of positive and negative cur rents through the coil the movements of which cause the lower end of the siphon to move to one side or other of the zero line of the paper. The upper end of the siphon dips in an inkwell B. and by an ingenious electrical arrangement, not shown in the figure, the ink is caused to spurt out at the lower end of the siphon upon the paper strip as a succession of very fine dots. Thus it is not needful that the siphon should touch the paper and in this way friction is avoided. A specimen of the signals as recorded by the siphon recorder is shown at the top of the figure.
For reasons not well known the earth's electric potential varies at different parts of the earth. Hence there is usually a difference of potential between the terminals of a long cable. This would tend to produce a current in a sensitive receiving instrument, which would deflect it to the detriment of signaling. To avoid this difficulty a condenser p is placed be tween the cable and the earth. This, so to speak, breaks the continuity of the circuit for slow changes of potential. When the condenser is once charged with the earth's potential, cur rent ceases to flow into the cable and the re ceiving instrument comes to zero. The changes in the earth's potential occur comparatively slowly— from maximum positive to negative in perhaps five minutes. Consequently these changes are scarcely perceptible on the receiv ing instrument. The quick changes of poten tial due to the transmitter, however, produce currents of charge and discharge that operate that instrument.
Long submarine cables are usually du plexed, the "bridge' method of preventing the effect of the home transmitting key upon the home receiving instruments, being employed. An artificial cable consisting of strips of tin foil arranged to give resistance and capacity equal to the actual cable is employed instead of the rheostats and condenser used for the artificial line in overland duplex telegraphy.
Signals are now usually transmitted over long submarine cables by automatic trans mitters somewhat similar to the Wheatstone transmitter; the messages being perforated in advance on a strip of paper. A speed of about 40 words per minute in each direction is reached on some of the newer Atlantic cables.
Simultaneous Telegraphy and Telephony. —This is sometimes erroneously termed 'learn posite telegraphy.° The art relates to telegraph ing and telephoning over one wire at the one time.
When a telephone receiver is inserted in an ordinary telegraph circuit the operation of the Morse keys deflects the diaphragm abruptly and produces loud noises in the tele phone which render the reception of speech nearly impossible. Van Rysselberghe, of Belgium, discovered that if the rise and fall of the telegraph currents were made gradual no disturbing sounds would be heard in the telephone. To bring about this result he in troduced into the telegraph circuit a combi nation of electromagnets and condensers, which by retarding the rise and prolonging the fall of the telegraph currents merely in flect the diaphragm of the telephone, but do not produce any sound thereby. When this
result is obtained the telephone currents may then be superposed upon the telegraph cur rents without impairing the efficiency of the telegraph signals. Simultaneous telegraphy and telephony is now in successful use in the United States on a large scale; two telegraph circuits being operated as one telephone metallic (two wire) circuit, on circuits up to 400 miles in length.
Synchronous Multiple: Telegraphy.— It is known that 500 pulsations of electricity per second can be transmitted on an overhead wire of moderate length. A telegraph oper ator at his best speed is not capable of trans mitting more than an average of 10 dots per second. Hence it was thought that if means were devised whereby a number of operators should consecutively be given exclusive con trol of a wire for brief intervals of time the same wire might be utilized to transmit four, six or more messages at practically the same time. In order that this might be done satis factorily it was evident that the correspond ing transmitting and receiving instruments at the near and distant stations should be placed in connection with the wire at identical instants. This entailed the construction of de vices for obtaining exact synchronism; hence the name of the system.
The apparatus for obtaining synchronism and for apportioning the wire among a num ber of operators consists of a revolving wheel at each end of the telegraph line, the wheels revolving as nearly as possible at a uniform rate. Each wheel is driven by an electric motor whose motion is controlled by a vibrating reed at each station, which reeds are attuned to the same rate of vibration, as closely as pos sible. This wheel is supported on a vertical shaft. Below it is a stationary circular disc made up chiefly of a large number of metal segments (84) radiating from near the centre of the disc, and insulated from one another by suitable material. The shaft supporting the wheel passes through the centre of the disc. The shaft carries by a suitable projection a brush or trailer, which, as the wheel revolves, is swept over the segments in rapid succession. As the trailer makes three revolutions per sec ond it comes in contact with each segment three times per second. If it is desired to transmit six messages at once, 72 segments are set apart for the purpose, and each of six desks at each end of the wire are allotted 12 segments. That is, starting from any given point on the disc, the first segment will be given to desk No. 1; the second segment to desk No. 2, and so on to the sixth segment, where two segments are skipped, being re served for synchronizing purposes. A second series of six segments is then connected to desks Nos. 1, 2, 3, and so on around the disc. The line wire is connected to the shaft or trailer of the revolving wheel at each station, and, consequently, as the trailers make three rev olutions per second, each desk of the six is placed in contact with the line, and with its corresponding desk at the distant station, 36 times per second. As an operator cannot make a dot in less than the one-twelfth of a second it follows that in that time the trailer will have given him contact with the line thrice. Hence each one of six operators may transmit messages as though he had entire control of the line. As each character received at any one desk is formed of a number of pulsations an ar rangement of relays with a contact on the ((back-stop° is employed which delivers the signals virtually unbroken at the receiving end. Synchronism is maintained by means of the two segments referred to which send ucor reefing') impulses that retard or accelerate the speed of the wheel that carries the trailer.