472 BELL SYSTEM TECIIMCAL JOURXAL 



occurs in daylight would seem to suggest for its explanation the pres- 

 ence of some factor in atldition to sunlight, such as electron emission. 



Field Strength Formulae. The two major phases of the diurnal 

 variation of signal field strength which lend themseK'es to possible 

 predetermination are the daylight values and the established night- 

 time \alues. As to the night-time values our data show, within the 

 limits of experimental error, that the maximum values do not exceed 

 that defined b>' the inverse distance law. This fact seems to support 

 the viewpoint * that the high night-time values are merely the result 

 of a reduction of the absorption experienced during the day. Fig. 11 

 presents the monthly a\erages of the daylight field strengths for the 

 various frequencies on which measurements were taken. The chart 

 at the left is for reception in England and that at the right for recep- 

 tion in the United States. 



The difficulty in predicting b\- transmission formuliK', \alucs to be 

 expected at any one time will be evident and the best that can be 

 expected is to approximate the average. The formulae of Sommer- 

 field, Austin-Cohen and Fuller take the form 



„ ,,, 377/// _«o 



where the coefficient -^7) — represents the simple Ikrizian radiation 



field and the expoiuniial '' X' the attenuation factor. From theo- 

 retical considerations, Stjmmerfeld (U)OV)) gave a=.0019 and x = }/i. 

 In the Austin-Cohen formula a is given as .0015 and x = l^. Fuller 

 gives a = .004.^ and .v = 1.4. The Austin-Cohen formula was tested 

 out experimentally chiefly with data obtained from the Brant Rock 

 station (1911) and from the Arlington station by the U.S.S. Salem 

 in Fel)ruar>' and March, 1913. Fuller derived his .0045 value of a 

 from 25 selected observations from tests between San I'rancisco and 

 Honolulu in 1914. 



An attempt has been made to determine the constants of a formula 

 of the above form which would approximate axerages of some 5,000 

 observed values of field strength over this particular New York to 

 London |)alh and over the freciuency range of 17 kc. to tiO kc. For 

 each transmitting station a series of comparatively local measure- 

 ments were taken to determine the power radiated. By combining 

 these local measurements with the values obtained on the other side 



' Sec also " Kailio Kxtcnsiun of Telephone System to Ships at Sea," Nichols and 

 Espcnschied, I'roc. I. R. E., June, 1923, pages 226-227. 



