486 



BELL SYSTEM TECHNICAL JOURNAL 



not valid for values of a^'^X''"'' less than 10"'^. While no glarinij incon- 

 sistencies are evident from Fig. 6 for values of the parameter somewhat 

 greater than this it is the writer's opinion that implicit faith should not 

 be placed in the results without experimental verification due to the 

 nature of the approximations made in obtaining the solution. ^^ Com- 

 parison of curves D and A shows that the Eckersley modification of 

 the Watson solution for values of the parameter of the order of 10"*' 

 is small which is consistent with the results for plane earth, curve G. 

 It is in this region that Eckersley presents experimental data to sub- 

 stantiate his solution. 



001 



01 



10 100 



TTd/X 



1000 



10000 100000 



c2a/f 



Pig_ 7 — Revised Eckersley curves for imperfectly conducting spherical earth. 

 (The conductivity in e.m.u. is represented by a and in e.s.u. by c-a, c = i 

 X 10>o.) 



Recently Eckersley-" has made the plausible suggestion that his 

 curves should be shifted vertically until they are tangent or nearly 

 tangent to the Sommerfeld curve. This results in the curves of Fig. 7. 



19 Since the writing of this paper an article by Jean Marique entitled " Note sur 

 Ouelques Measures due Rayonnement des Stations de Navires," L'Onde Eleclrique 

 13, 149-156, March, 1934 has come to the attention of the author. Experimental 

 data are presented from which he concludes that Eckersley 's results do not apply lor 

 distances of the order 400 to 500 km. at a wave-length of 600 meters. 



2»T. L. Eckersley, "Stuih- of curves of propagation of waves," Document 

 A. G. (1934), No. 11, comm. 11; International Scientific Radio Union, \'lh 

 Assembly. 



