484 BELL SYSTEM TECHNICAL JOURNAL 



Data published by Bion and David * to show the inadequacy of 

 Sommerfeld's solution for the propagation over sea water in the wave- 

 length range 150 to 700 meters, have been plotted in Fig. 5. While 

 only eight points are shown they represent data taken at regular 

 intervals on a ship whose distance from the transmitter was con- 

 tinuously increased up to 1050 km. so that their precision is far superior 

 to that possible with single measurements. The points have been 

 plotted against the parameter y = d/yI\K^ using 4/3 for the value of 

 K. The points lie close to the theoretical curve, substantiating the 

 theoretical curve and indicating that atmospheric refraction was 

 sufficient to increase the effective radius of the earth by the factor 4/3 

 for radio propagation (in this frequency range) over the Mediterranean 

 Sea in January and February, 1932. 



Effect of Imperfect Conductivity 



Due to the complications introduced into the problem of the propa- 

 gation of electromagnetic energy around the surface of the earth by 

 the effect of imperfect conductivity, no rigorous solution has been 

 made to date. The approximate solution due to T. L. Eckersley,^"- ^^ 

 however, has been used ^" to calculate the field strength of the ground 

 wave at distances beyond those for which the solution for transmission 

 over an imperfectly conducting plane applies. The results of this 

 solution will be compared with the rigorous solutions of special cases, 

 leaving a discussion of some of the approximations made and the un- 

 certainties introduced thereby for the appendix. 



Theoretical curves obtained by various methods for propagation 

 over the surface of the earth are presented in Fig. 6 for comparison. 

 Curve A is for perfectly conducting spherical earth based on Watson's 

 solution. Curve B is based on Eckersley's solution for a spherical 

 earth whose conductivity is small enough so that its magnitude is 

 unimportant but large enough so that it is essentially a conductor 

 rather than a dielectric. ^^ Curves C and D result from using the coef- 

 ficients given by Eckersley corresponding to the values of a^^'^X^^^ 

 indicated on the curves. Curves E, F, G and // are for imperfectly 



^ J. Bion and P. David, "Sur L'Affaiblissenienl des Ondes Moyennes et Inter- 

 mediaires se Propageant de Jour sur Mer," Comptes Rendus 194, 1723-1724, May 

 17, 1932. 



'" T. L. Kckersley, "Radio Transmission Problems Treated by Phase Integral 

 Method," Proc. Roy. Soc. {London) A136, 499 527, June 1, 1932. 



" T. L. lickersley, "Direct Ray Broadcast Transmission," Proc. LR.E. 20, 

 1555-1579, October, 1932. 



'-See for example, "Report of Committee on Radio Propagation Data," Proc. 

 LR.E. 21, 1419 1438, October, 1933. 



'^P"or detailed explanation of this curve see the appendix. 



1 



