STUDIES IN RADIO BROADCAST TRANSMISSION 



185 



east may be inserted with their true radial relation. The two most 

 evident in Fig. 1 are shown to be west of the line extending from 

 transmitter through the center of the obstruction located in Fig. 26. 

 The presence of Long Island Sound east of the geometrical center 

 of the shadow has made an extensive survey of this section imprac- 



ti 



/ 



«<$ 



x^- 



s'i-' 



j Jo 20 30 « 50 60 TO 60 90 100 110 120 130 140 150 IbO 



V^avc Lenflhts from Transmitter 



Fig. 31 — Plot showing intensity of definition of wave interference pattern 



tical. However, a single section taken across the Sound at about 

 90 w^ave-lengths from the station shows quite unquestionably the 

 presence of a low channel about as indicated to the right of the ob- 

 struction designated in Fig. 29. 



We have, therefore, a deep shadow with a more or less orderly 

 array of maxima and minima within its limits. These maxima 

 and minima grow more distinct at a distance from the transmitter, 

 contrary to what we might expect for ordinary shadows. Further- 

 more, we find that they move as the frequency is changed. These 

 facts lead to the belief that the phenomena in question are due to 

 wave interference such as has already been described in connection 

 with night-time fading, but characterized by very much smaller path 

 differences. This daytime interference condition is fixed while we 

 have seen that the nocturnal patterns appear to wander continually. 

 To explain this more in detail let us return to the shadow and con- 

 sider the phenomena which might accompany it in a little more 

 detail. 



The study of light has made available much information concern- 

 ing the subject of wave interference. It is known, for instance, that 

 the edges of shadows are not sharply discontinuous changes from 

 light to darkness, but that a series of dark and light bands, called 



