376 BELL SYSTEM TECHXICAL JOURNAL 



phony, culminating in the first talk over the Xew York-San Francisco line 

 by Mr. \'ail on July 29, 1914 and in the ofi&cial opening of the line for com- 

 mercial service on January 25, 1915. Many amusing stories are told of 

 the efforts of the engineers to do their final testing on the line without trans- 

 mitting any of their voices from coast to coast, the injunction having gone 

 forth that under no circumstances was anything to happen that would 

 detract from \'airs first talk. 



In general, the present story does not discuss the personal contributions 

 of individual engineers and physicists which were essential to the complete 

 success of the project. Some information on these matters is available in 

 an article, "The Line and The Laboratory," written by John Mills, and 

 published in the Januar}^ 1940 issue of the Bell Telephone Quarterly, along 

 with other articles commemorating a quarter century of transcontinental 

 service. 



In the beginning, the Western Electric Research department and the 

 Transmission Engineering department of the American Company were 

 most actively engaged in the project. In the course of time, these depart- 

 ments were expanded to handle the increasing amount of work and other 

 associated departments in these organizations became involved in the 

 cooperative efforts. The engineers of the Long Lines department, and of 

 the Pacific and Mountain States companies also, did their own very impor- 

 tant jobs, and last but not least, so did the manufacturing organization of 

 the Western Electric Company. 



The Improved Mechanical Repeater 



The 1911 analyses of the then available form of Shreeve repeater, re- 

 ceiver-transmitter mechanical type, indicated the principal defect to be a 

 very marked natural period about midway in the telephone talking range, in 

 which range the amplification was very much greater than at low and high 

 voice frequencies. This caused distortion and tendency to sing well within 

 the audible range. Other serious defects were a variable am.plification with 

 different magnitudes of input energy, the amplification with low levels being 

 markedly less than with high level input, non-linear distortion, and a tend- 

 ency for periodically variable amplification from instant to instant. 

 Inertia of the moving parts was a congenital handicap that could not be 

 completely overcome. The diaphragm of its receiver portion had to vibrate 

 at any and all speech frequencies, and simultaneously drive at the same 

 vibration rates the movable electrode of the carbon-button transmitter. 



The analysis just summarized resulted in design modifications which 

 materially improved the performance characteristics. Specifically, these 

 modifications improved the magnetic circuit, reduced the movable mass, 

 and raised the natural period of the vibrator}- system to the upper part of 



