SUBMARINE TELEGRAPH CABLES 429 



American Telephone and Telegraph Company and the Western 

 Electric Company in 1912. The vacuum tube amplifier appeared 

 to ofifer important advantages for use on submarine cables because of 

 its lack of distorting effects which are a function of the frequency of 

 the current amplified, and also because of the ease with which signal 

 distortion correcting circuits could be associated with the vacuum 

 tubes. The initial studies on amplifier circuits suited to currents of 

 the low frequencies involved in submarine cable telegraphy were made 

 by Mr. R. V. L. Hartley and Mr. B. W. Kendall. « One of the 

 difficulties which loomed quite large at that time was that most cables 

 were operated duplex and the connection of an amplifier to a duplex 

 circuit would involve the insertion of a transformer which promised 

 to introduce distortion''. A suitable distortionless amplifier was first 

 tried and subsequently distortion correcting networks were introduced 

 between its stages. It was found that this permitted the use of more 

 correcting elements than had been feasible in previous practice and 

 thus indicated the possibility of attaining higher speeds than were 

 usual at that time. The development of the shaping circuits employed 

 was at first based on the principle of producing the various derivatives 

 of the arriving current wave and adding them in proper phase relation 

 to the arriving wave. This principle and methods of applying it had 

 been developed mathematically by Mr. J. R. Carson of the American 

 Telephone and Telegraph Company.^ Mr. R. C. Mathes who con- 

 ducted the experimental investigation beginning in 1916 simplified his 

 work somewhat by recognizing that this principle was equivalent to a 

 statement that the received signal would be satisfactory if the attenu- 

 ation and phase distortion of the entire system of cable and amplifier for 

 steady state alternating currents were corrected by the shaping net- 

 works over a range of frequencies from nearly zero to approximately the 

 nominal signalling frequency. By the middle of August 1918 the 

 employment of improved shaping methods made speeds of 22 cycles 

 possible in simplex working on an artificial cable having a KR. of 2.7. 

 The then standard cable apparatus would have permitted a speed of 

 not more than 9 cycles, on a cable subject to interference of the mag- 

 nitude usually encountered. 



6B. W. Kendall, U. S. Patent No. 1,491,349, April 22, 1924. 



^ Expedients for avoiding distortion of this nature were suggested by Dr. H. W. 

 Nichols of these Laboratories and by Mr. Lloyd Espenschied of the A. T. & T. Co. 

 Their plans contemplated the modulation of an alternating current of relatively high 

 frequency by the incoming signal, the amplification of the modulated current by 

 suitable apparatus and its subsequent demodulation for obtaining the amplified low 

 frequency signal (H. W. Nichols U. S. Patent No. 1,257,381, February 16, 1918; 

 Lloyd Espenschied, U. S. Patent No. 1,428,156, September 5, 1922). 



s'See U. S. Patents No. 1,315,539, September 9, 1919, No. 1,450,969, April 10, 

 1923, No. 1,516,518, November 25, 1924 and No. 1,532,172, April 7, 1925. See also 

 article by Dr. K. W. Wagner, Electriche Nachrichten-Technik, October 1924. 



