rROP.tGATlOX OF ELECTRIC WAVES OVER THE EARTH 2.13 



.my point. If we consider the wave a short distance before it reaches 

 the receiver, we will tind regions in which the wave front is conca\e 

 to the recei\er and regions of opposite curvature. Thus at certain 

 portions of the wa\e front energ>- will be concentrated toward a point 

 farther on and at other parts will be scattered. The location of these 

 convex or concave F«)rtions of the wave in the neighborhood o{ a 

 given receiving point will be very sensitive to changes in ionic dis- 

 tribution along all the paths of the elementary rays contributing 

 to the effect at the receiver. Hence, if we knew the location and 

 movement of all the ions between the transmitter and the receiver, 

 it would be possible, theoretically, to predict the resultant eflfect at 

 the latter point. 



To explain fading it is essential that there be a time variation in 

 this distribution. It is clear that effects of this kind should be more 

 marked at short waves than at long waves since a region of the medium 

 comparable in dimensions to a wave length must suffer some change 

 in order to produce an effect upon the received signal. If, for example, 

 there were space irregularities in the medium comparable to the wave 

 length, a kind of diffraction effect would be produced at the receiver 

 which would be ver>' sensitive to slight changes in grating space. 



A possible cause of irregularity may be found in the passage across 

 the atmosphere of long waves of condensation and rarefaction, each 

 of which results in a change in the density and gradient of the ions, 

 even though the average density remains constant throughout a 

 large %olume. If, as seems plausible, the upper atmosphere is 

 traversed by many such atmospheric waves of great wave length, 

 the resulting effect at a given receiving point would be fluctuations 

 in signal strength due to a more or less rapid change in the configura- 

 tion of the wave front near the receiver. 



For radio waves whose length is of the order of a few hundred 

 meters, fading experimentally observed occurs at a rate of the order 

 of one per minute (of course, it is not implied by this statement that 

 there is any regular periodicity to the fading). The pressure wave 

 referred to would travel in the upper atmosphere with a velocity of 

 the order of 300 meters per second at lower levels or 1,000 meters in 

 the hydrogen atmosphere, so that the wa\e length of these "sound" 

 waves would be of the order of 50 of the radio wave lengths. The 

 irregularities of the medium would thus be of sufficient dimensions 

 with respect to the electromagnetic waves so that one of the char- 

 acteristics referred to above might be developed. In this way we 

 might explain variations in intensity of the wave at the receiver re- 

 curring at intervals of a minute or so. 



