70 



THE BELL SYSTEM TECHNICAL JOURNAL, JANUARY 1952 



the free space levels. Also during the winter montlis, when the humidit}^ 

 content of the atmosphere is low, signal variations are usually very small. 

 However, on clear summer nights with little or no wind, non-uniform 

 distributions of temperature and humidity can create steep dielectric 

 constant gradients in the lower atmosphere, thus causing anomalous 

 propagation and fading. 



When fading occurred on our experimental transmission paths, an 

 alarm circuit connected into the continuously recording equipment was 

 arranged to operate when the signal level dropped below a predetermined 

 value. This enabled observers to be present during severe fading periods 



SOUTHARD 



HILL 



240 FT 



CRAWFORD 



HILL 



385 FT 



10 12 14 



DISTANCE IN MILES 



¥ig. 2 — Profiles of the transmission paths. 



and to seek, by means of the special experiments, to determine the causes 

 of the fading. 



Although it has not been possible to pro\'ide satisfactory explanations 

 for all of the observed fading phenomena, much of the fading (occasions 

 when the signals are depressed to levels 15 to 20 decibels or more below 

 the normal daytime value) can be explained qualitatively in terms of 

 simple ray pictures. Fig. 3 is intended to illustrate some of the observed 

 fading phenomena. The case of multiple path transmission, the most 

 common cause of fading on either transmission path, is shown in Fig. 3(a). 

 Two, three and sometimes more signal components are found to arrive 

 at various angles in the vertical plane, usualh^ above the line of sight. 

 Wave interference among these components produces fading, the severity 

 of which depends upon the relative amplitudes and delays of the com- 

 ponents. At these times, different frec}uencies fade differently and the 

 signals received on two verticallj^ spaced antennas also fade different 1}'. 

 The use of either frequency or space diversity would be effective in this 

 type of fading. 



A relati\'ely rare tj'pe of fading, observed onl}' on the Murray Hill 

 path, is believed to be caused by the mechanism illustrated in Fig. 3(b). 



