sent over a single line either simultaneously or in 

 rapid succession. A system by which two messages 

 could be sent simultaneously over the same wire was 

 called duplex, and one over which four could be sent 

 at the same time over the same wire (two in each 

 direction) was called quadruplex. A system that 

 could transmit a number of messages in rapid succes- 

 sion was originally called a multiplex system (fig. 46) . 

 Since receivers always were more sluggish than trans- 

 mitters, a multiplex system changed the "dead" inter- 

 vals of the receiver to an advantage. Various 

 electrical and mechanical techniques were applied to 

 establish a practical multiplex system that could be 

 duplexed or quadruplexed. 



The earliest duplex circuit was based upon a bridge 

 circuit in which currents in the transmitting station 

 would divide and then come together again in 

 such a manner as to cancel one another out and not 

 actuate the local receiver. Currents from the other 

 station instead of canceling would add so as to operate 

 the local receiver. Such a bridge circuit could be 

 applied to telegraph systems if the impedance of 

 the telegraph line joining the stations was related in 

 a certain way to the impedance of the stations. The 

 first attempts to work out such a bridge circuit were 

 made in 1853 by Wilhelm Gintl, director of the 

 Austrian State Telegraph, and in 1857 by Carl 

 Frischen, a Hanoverian telegraph inspector. How- 

 ever, the circuits of Gintl and of Frischen and those 

 of certain subsequent inventors tried to match the 

 impedance of the telegraph line with resistances only. 

 Since a real telegraph line has a capacity as well as 

 a resistance, no satisfactory duplex method was 

 devised until both these features were taken into 

 account. Joseph B. Stearns of Boston succeeded in 

 creating a duplex telegraph by adding a capacitance 

 to the circuit in the proper place (fig. 47). 



Thomas A. Edison ^* was the first designer of 

 a practical quadruplex system. In 1874 he showed 

 how two duplex circuits could be superimposed by 

 using a reversal of current to signal in one circuit, 

 and an increase and decrease of current away from 

 the reference level to signal in the other circuit. 

 While the message-carrying capacity of telegraph lines 

 was increased by such duplex and quadruplex methods 



(fig. 48), they reduced the speed of transmission for 

 each station; consequently other methods of multiple 

 telegraphy were sought. 



Multiplexing a line was another method by which 

 a single line could be used for transmitting messages 

 between a number of stations. Multiplexing was 

 originally a mechanical method that used a commu- 

 tator to switch rapidly among several pairs of trans- 

 mitters and receivers. It is obvious that by this 

 method many messages could be sent over a single 

 line but that they could not be transmitted simul- 

 taneously. The multiplex system was first suggested 

 in 1852 by Moses G. Farmer, whose idea was to set 

 up a commutator, similar to that in the distributor 

 of an automobile, at each end of a telegraph line 

 in such a manner that the motion of the brushes in 

 the commutator was synchronized so as to join 

 corresponding stations at each end of the line. How- 

 ever, it was some time after Farmer conceived this 

 idea that it was reduced to practice. 



After Patrick B. Delany and Bernhard Meyer 

 had made initial attempts to create a practical 

 multiplex system, between 1872 and 1878 J. M. E. 

 Baudot'' managed to devise a workable multiplex 

 system. While the Morse system could send up to 

 25 dispatches per hour and the European Hughes 

 machine could send 60 dispatches per hour, the 

 duplex process enabled them to transmit 45 and 110 

 dispatches per hour, respectively. The quadruplex 

 process as applied to the Wheatstone automatic tele- 

 graph could send 90 dispatches per hour and 160 

 dispatches per hour if the system was duplexed again. 

 A hundred dispatches per hour could be sent by 

 the Meyer multiplex system, and 160 by the Baudot 

 system and almost double that if duplexed. Use 

 of the Baudot system spread in France in the 1880's, 

 and in the late 1890's it was introduced into England. 

 Further improvements in the Baudot system and its 



38 Thomas A. Edison, U.S. patents 207723 and 207724 

 (September 3, 1878), and 209241 (October 22, 1878); F. L. 

 Dyer and T. C. Martin, Edison: His Life and Inventions, New 

 York and London, 1910, 2 vols.; Matthew Josephson, Edison: 

 A Biography, New York, 1959. 



39 French patents 103898 (June 17, 1874), 11719 (March 

 2, 1876), and 146716 (January 6, 1882); Theodose du Moncel, 

 "Systemes telegraphiques imprimeurs i transmissions mul- 

 tiples," La Lumiere electrique, 1880, vol. 2, pp. 61-66, 81-84; 

 "J. P.," "Etude sur le systeme de transmission multiple et le 

 telegraphe imprimeur de M. Baudot," La Lumiere electrique, 

 1881, vol. 4, pp. 378-380, vol. 5, pp. 53-57; 1882, vol. 6, pp. 

 55-60, 79-82, 1 27-1 31 , 1 77-184, 1 98-202 ; Theodose du Moncel, 

 "Rapport . . . sur le telegraphe multiple de M. Baudot," 

 Bulletin de la Societe d' Encouragement pour I'Industrie Nationale, 1 883, 

 vol. 10, pp. 149-154; A. C. Booth, "The Baudot Telegraph 

 System Duplexed," Post Office Electrical Engineers' Journal, 1911, 

 vol. 3, pp. 336-339, and "Progress of the Baudot System," 

 1914, vol. 6, pp. 324-336. 



308 



BULLETIN 228: CONTRIBUTIONS FROM THE MUSEUM OF HISTORY AND TECHNOLOGY 



