554 BELL SYSTEM TECHNICAL JOURNAL 



long telephone circuit, should diminish rather than increase the 

 attenuation. Campbell followed this suggestion and developed a 

 theory of loading, but in his case there occurred one of those coinci- 

 dences — fortunately rare in the history of science — of two investigators 

 arriving at substantially the same result at the same time. Inde- 

 pendently of Professor M.I. Pupin, he worked out a complete theory 

 of the telephone loading coil. They both applied for patents, with 

 the result that an interference was declared, and Pupin was able to 

 establish a slightly earlier date of conception. The loading coil inter- 

 ference was decided in Pupin's favor and the famous patents issued 

 to him. The fact should be recorded, however, that Campbell's 

 analysis of the problem — actually more detailed than Pupin's — led 

 him to formulate rules for the design of loading coils and their spacing 

 which were, from the very beginning, the only ones employed in 

 this country. 



As the effectiveness of telephone instruments increased and the 

 lengths of circuits grew, noise and crosstalk became an outstanding 

 obstacle to telephone advance. It had been shown that this crosstalk 

 was a complex effect resulting partly from electromagnetic and partly 

 from electrostatic induction. In unpublished memoranda written 

 between 1903 and 1907, Campbell pointed out the importance of 

 Maxwell's capacity coefficients in the calculation of crosstalk and 

 coined the much-used term "direct capacity," now modernized to 

 "direct capacitance." It was also at this time that he designed his 

 well-known "shielded" balance, which in one form is a bridge for 

 measuring direct capacities. He showed in these early memoranda 

 that crosstalk between two circuits depends, to a considerable extent, 

 and particularly in the case of loaded circuits, on a function of the 

 various direct capacities between the wires of two circuits. He termed 

 this function the "direct capacity unbalance." This work led to the 

 invention of the well-known capacity unbalance test set, hundreds of 

 which have been used in countless measurements in the manufacture 

 and installation of toll cables. 



This study of Campbell's marked an important advance since, for 

 the precise but unwieldy theory of crosstalk, it substituted a simple 

 approximation — an approximation which was to remain adequate until 

 the advent of carrier systems with their higher frequencies and shorter 

 wave-lengths. 



As control was gradually extended over the characteristics of 

 telephone circuits, both from the standpoint of their transmission 

 effectiveness and their freedom from crosstalk and noise, the art 

 reached the point at which development emphasis shifted to the 



