384 



BELL SYSTEM TECHNICAL JOURNAL 



As Is shown in the appendix, the small percentage of water vapor, 

 present normally in the atmosphere, has a very marked effect on the 

 radius of curvature of the radio ray. While the bending, there cal- 

 culated, does not quantitatively explain our September 27, 1933 results 

 at 1000-foot altitude, it is qualitatively in the right direction. The 

 weather bureau data given us were. 



and were taken on top of a New York City building. The humidity 

 and its gradient, at the ocean surface may well have been greater. 

 It is not impossible, either, that water vapor absorption bands occur 

 in the ultra-short-wave region. High and irregular refraction effects 

 would then occur. 



It is evident that a slight change in ray curvature under grazing 

 incidence transmission conditions will give a marked increase in range. 

 The fading of weak signals under these conditions, which has been 

 observed in this country by Bell Laboratories engineers at Deal Beach, 

 New Jersey, and by Radio Corporation of America observers, and in 

 Europe by Senatore Marconi and International Telephone and Tele- 

 graph Company engineers, may possibly be explained in this manner. 



This work is being continued. 



APPENDIX 



1. From the accepted theory ^ the dielectric constant of a gas is 

 given by 



€ - 1 



47r7V /. , , Ayi 

 he- -j- 



e -f 2 3 V ' 3RT 



^ See Debye, "Polar Molecules," Chemical Catalogue Company. 



