Dec. 28, 1876] 



NATURE 



l8f 



pensating " circuit. Both of these wires were connected 

 at their other ends to a signalhng key or sending instru- 

 ment, K, having duplicate points. Two batteries, B and 

 ]?', one large and the other small, were also connected to 

 this key in such a manner that on making contact the 

 current from the larger battery B passed through the long 

 coil of the instrument into the hne, while at the same 

 instant the current from the smaller battery B', passed 

 through the short coil and compensating circuit. These 

 two currents were so adjusted as to balance each other in 

 their effects on the indicator of the receiving instrument, 

 and no signal was therefore made by the outgoing or 

 signalling current from the battery B. Such was Gintl's 

 arrangement at both stations, s and s'. Confining our 

 attention to any one of these stations, say s, we see that 

 as long as only S is sending, the receiving instrument at 

 that station is unaffected, but when s' is also sending at 

 the same time, the currents in the line from s' must inter- 

 fere with the currents in the line from s, either aiding or 

 opposing these according to the poles of the battery which 

 are applied, and thereby disturbing the balance of currents 

 on the receiving instrument at S, causing it to make 

 signals. And we see also that these signals are entirely 

 under the control of s' — that when s' applies his battery 

 to the line, s's instrument will make a signal, and that 

 when s applies his battery to the line, s"s instrument will 

 make a signal, or that when either put " earth " to the line, 

 that corresponding " spaces" will be recorded. The cur- 

 rents do not cross each other in the line, as was some- 

 time thought, but they interfere with each other in such 

 a way as to disturb the electric balance which independ- 

 ently exists at either end. 



One objection, however, to this plan of Gintl's was that 

 the lever of the key during sending interrupted the line 

 circuit when it was between the earth and battery con- 

 tacts ; and another more serious objection lay in main- 

 taining the equivalence of the two currents, as the smaller 

 battery working through the shorter circuit lost power 

 more rapidly than the signalling battery. 



Gintl, however, in the following year (1854) obtained 

 better success with a Bain's chemical printing instrument. 

 Here he made the two currents oppose each other upon 

 the chemically prepared paper and no stain was produced 

 by the outgoing current. The sending current from the 

 other station neutralised to a certain extent the outgoing 

 current and the local current then overmastering it 

 stained the paper and produced signals. 



Early in 1854, while Gintl was still engaged in experi- 

 ^nenting with the Bain's instrument, a great improvement 

 was effected on his plan by Herr Carl Frischen, a tele- 

 graph engineer of Hanover. Frischen dispensed with 

 the second battery b' and split up the signalling current 

 itself, causing one part to pass round one coil of the 

 receiving instrument into the line and the other part to 

 pass round the other coil of the instrument into the local 

 or compensating circuit. Instead of making the two 

 wires of the instrument dissimilar, like Gintl, he made 

 them both alike, so that the instrument was practically 

 wound by equal wires in opposite directions and thereby 

 rendered differential. He inserted into the compensating 

 circuit a rheostat or set of resistance coils whose resist- 

 ance could be adjusted to equal the electric resistance of 

 the line. His arrangement at either station is shown in 

 Fig. 2. Here R i is the differentially wound receiving 

 instrument, B is the battery, and R is the rheostat or arti- 

 ficial line as it was called, because of its being intended 

 to imitate the actual line. K and E are as before, the 

 sending key and earth plates respectively. In the act of 

 sending a message, on making contact betweea the key 

 and battery at a the current flowing into the apparatus 

 divides itself at the point b, and part passes through the 

 right coil of the instrument into the line, while part 

 passes through the left coil into the local circuit. But 

 itie resistance of the local circuit being made equal to 



that of the line and the two coils of the instrument bein j 

 electrically equal one //^/^ of the current will flow through 

 one coil and the other half through the other coil. The 

 coils being oppositely wound these currents will neutralise 

 each other in their effect on the indicator or needle of the 

 instrument and no signal will be produced as long as they 

 flow freely. We have seen, however, that the sending at 

 the remote station may either oppose or assist that part 

 of the current entering the line. Thus the balance is 

 disturbed and signals will be made on the receiving in- 

 strument of one kind or another according as the line 

 current or the local current overbalance each other. 



In sending by this method of Frischen's the line cir- 

 cuit is never completely interrupted, but it will be observed 

 that according to the position of the key the line is either 

 applied " to earth" direct or through the battery, or 

 through the resistance R, and this leads to a troublesome 

 variation in the signals. 



Frischen's method was re-invented a few months sub- 

 sequently by Messrs. Siemens and Halske of Berlin, who 

 patented it in England, where, however, they were fore- 

 stalled by a week by Mr. Stirling Newall, of Gateshead, 

 whose patent bears date October 30, 1854. Newall de- 

 scribes substantially the method of Frischen-Siemens, 

 and in 1855-56, successful experiments were made under 

 his patent on the Manchester and Altringham line ; but 

 when he came to make trials upon the longer line between 

 Birmingham and London, the static charge and discharge 

 of the line was found to make false signals. Condensers 

 or accumulators of electricity were applied to correct this 

 disturbance with promising results, but the expense of 

 constructing large condensers was objected to and their 

 use was prematurely abandoned. 



At this time there was a great deal of activity in the 

 direction of duplex working displayed both at home and 

 abroad. Besides the names mentioned, Messrs. W. H. 

 Preece, Highton, De Sauty, and others in this country, Ed- 

 lund, Bosscha, Kramer, in Germany, and Farmer, in Ame- 

 rica, were all more or less engaged in the work. In 1856 

 a good many lines were worked on the duplex principle in 

 Prussia, but the time was not yet ripe for its successful 

 introduction, and it gradually fell into disuse again. 

 Some desultory activity was still shown here and there, 

 however, showing that the idea had not been lost sight 

 of. In 1862 Mr. William Hinckling Burnett patented a 

 method of working two or even three distinct systems of 

 telegraphy by the use of currents of different degrees of 

 force, and in 1863 M. Maron, of Berlin, appears to have 

 first placed the receiving instrument in the diagonal of 

 a Wheatstone's balance. This has been called the Wheat- 

 stone bridge method in the recent revival of duplex, and 

 it may be shown as follows. Fig. 3. 



Here r and ;' are two of the branch resistances of a 

 Wheatstone's balance ; the line and artificial line, or 

 rheostat, R, forming the other two branches. The receiv- 

 ing instrument, R I, is inserted in the " bridge " wire. In 

 sending, the current from the battery B splits at b, and 

 part passes by r into the actual line, while part passes by 

 r' into the artificial line. If r be equal to r', and the 

 actual line be equal to the artificial line, the current will 

 divide itself equally, and the potentials, or, as it may be 

 called, the electric levels, at c and d, will be equal, and 

 there will be no tendency for a current to flow through 

 the cross channel or '• bridge wire," and the receiving in- 

 strument will not, therefore, be affec:ted. When, however, 

 the line currents are interfered with by the sending cur- 

 rents of the distant station, this balance will be disturbed, 

 currents will flow through the " bridge wire," c d, and the 

 receiving instrument will signalise. The resistance, iv, 

 may with some slight advantage be made equal to the 

 battery resistance. It will be seen that in the balance 

 method, as in the differential method, the principle con- 

 sists in dividing the signalling current between two cir- 

 cuits whose electrical properties are practically the same-, 



