650 BELL SYSTEM TECHNICAL JOURNAL 



the grounding electrodes and the conductivities, and the thicknesses of 

 the several strata. By means of measurements of the mutual resist- 

 ance for direct currents of circuits with suitably located ground elec- 

 trodes, the conductivities and thicknesses of the strata can then be 

 determined. 



Practically, the experimental work presents many problems, among 

 them being the elimination of the effects of stray earth currents and 

 evaluation of the effects local irregularities in earth conductivity. 

 A preliminary trial of the method was made in connection with the 

 tests at Massillon, and while local irregularities were found to be 

 quite marked, yet the average earth conductivity in the region cov- 

 ered by the tests was about 1.5 X lO^^^abmhos per cm. cube, which 

 is not greatly different from that indicated by the coupling tests. A 

 quite extended series of tests at Cross Keys, using an improved tech- 

 nique, yielded results in excellent agreement with the hypothesis that 

 at this location the earth is stratified, having an upper layer about 

 150 ft. thick with a conductivity about 3.4 X 10~^^ and a conductivity 

 in the lower stratum about 2.6 X 10~^^. 



Conclusions 



In conclusion, it is well to recall the end towards which the work 

 described in this paper has been directed. It was desired, first, to 

 obtain a sufficiently detailed experimental study of the mutual imped- 

 ances of ground-return circuits to enable the formation of an adequate 

 picture of the physical phenomena involved; also to test out the 

 theoretical formulas available. Second, the aim was to investigate prac- 

 tical means for enabling the calculation of the ground-return mutual 

 impedances of power and telephone lines. 



With regard to the first item, it was found that an analysis in terms 

 of an "equivalent ground-plane" was inadequate to represent com- 

 pletely the observed phenomena. However, when information is 

 available as to the proper value of ground-plane depth, this method 

 can be used to advantage in many cases where approximate results only 

 are desired. 



A theory based on the assumption of a d.-c. distribution in the earth 

 gave a somewhat better explanation, particularly in connection with 

 the mutual impedances of circuits in which the points of ground con- 

 nection were in close proximity, but left much to be desired in the way 

 of quantitative agreement with the experimental results. The results 

 of a theory which considers the effect of eddy currents in the earth are 

 shown to be in fair qualitative agreement with some of the test values, 



