pointed out the existence of a submarine plateau be- 

 tween Ireland and Newfoundland. This plateau 

 promised a natural advantage for laying an Atlantic 

 cable. In 1856 Field enlisted Charles Bright and 

 John Brett to join him in organizing the Atlantic 

 Telegraph Company for the purpose of undertaking 

 to lay a cable between Newfoundland and Ireland. 

 By 1857 some 2,500 miles of armored gutta-percha 

 cable had been manufactured, and the cable-laying 

 ship started from Ireland at the end of the summer. 

 However, the cable broke about 330 miles away from 

 the starting point, and the project had to be postponed 

 to the following year. It was not until the second 

 attempt in 1858 that the cable was successfully laid; 

 signals were sent through it for a few weeks, but then 

 it failed. 



In addition to the considerable mechanical problem 

 of contriving a submarine cable for the Atlantic 

 Ocean and of successfully laying it along the bottom, 

 there was also the electrical problem of the invention 

 of a new kind of telegraph receiver. Obviously none 

 of the commercial instruments of the time were able 

 to work through a line thousands of miles long, and 

 it was impossible to insert relays into the circuit of a 

 line at the floor of the sea. Moreover, even if sensitive 

 receivers were used, the line acted like a huge Leyden 

 jar and smeared out the signal. 



The problem of designing a new receiver for the 

 Atlantic cable was solved by Prof. William Thomson — 

 later Lord Kelvin.^' In 1855 Thomson had published 

 an article on signaling through submarine cables, in 

 which he pointed out some of the problems that would 

 have to be met. After joining the Atlantic Telegraph 

 Company as one of the directors, Thomson turned his 

 attention to finding a practical method of eliminating 

 the difficulties in the detection of oceanic cable signals. 

 The most sensitive detector known at that time was the 

 needle galvanometer ^* as provided with a mirror, and 

 it was upon this instrument that Thomson, in 1858, 

 based a new telegraph receiver with a "speaking" 

 galvanometer (fig. 40) that could be used on ship- 

 board without being influenced by the rolling motion 

 of the sea. C. F. Varley (1862) and C. W. Smith 

 (1866) independently showed how the addition of a 

 condenser to each end of the cable would insulate the 

 cable and sharpen the signal, and thus counterbalance 

 the loss of signal definition resulting from its passage 



Figure 41. — Movement of Thomson's siphon 

 recorder of 1871. From G. B. Prescott, 

 Electricity and the Electric Telegraph, New York, 

 1859, pp. 459> 461- 



through the cable. Later, in 1867,'^ Thomson 

 patented the siphon recorder, which was able to furnish 

 a permanent record of the message. In this siphon 

 recorder, a mobile coil of wire about one pole of a 

 stationary permanent magnet replaced the magnetic 

 needle and fixed coil used in conventional galva- 

 nometers. Thomson's final (1871) form of gal- 

 vanometer ^^ was so sensitive that the current from 

 a chemical cell (made from a silver thimble) could be 

 detected after being sent across the ocean and back 

 again. By the time these improvements in the cable 

 system were worked out, two successful Atlantic cables 

 had been laid. 



In spite of the considerable monetary loss resulting 

 from the breakdown of the first Atlantic cable, Gyrus 

 Field did not become discouraged and in 1864 he was 

 able to organize another company, which first at- 

 tempted to lay a cable in 1 86 5 . This cable came within 

 660 miles of Newfoundland when it broke. By then it 

 was too late in the year to undertake a relaying; how- 

 ever, in the summer of 1866 Field's company was 

 successful. Moreover, the company was able to find 

 and use the cable that had been laid the previous 

 year, so that the net result was two cables under the 

 Atlantic Ocean. These cables remained in use for a 



33 Silvanus P. Thompson, Life of William Thomson: Baron 

 Kelvin of Largs, London, 1910, 2 vols. 

 3« William Thomson, British patent 329 (February 20, 1 858). 



35 British patent 2147 Quly 23, 1867). 

 38 British patent 252 (January 31, 1871). 



306 



BULLETIN 228: CONTRIBUTIONS FROM THE MUSEUM OF HISTORY AND TECHNOLOGY 



