692 BELL SYSTEM TECHNICAL JOURNAL 



from 1924 to 1929, during the periods from September 15 to May 15, 

 there appears ^* "to be a decided drop in signal intensity (measured at 

 10 :00 A.M. and 3 :00 p.m.) coincident with heavy rainfall, rising temper- 

 ture, and falling barometer, at the receiving station at the time of 

 general storms, and an increased signal intensity as the storm center 

 passes and is followed by clear and colder weather." A conclusion was 

 presented ^^ that "either radio waves are subject to large absorption in 

 the lower atmosphere, due to a degree of ionization not known to exist, 

 or surface storms are influenced by conditions existing at great 

 heights." 



Disturbed weather conditions may be part of the reason for disturbed 

 transatlantic long-wave (60 kc.) radio reception, but the authors favor 

 the hypothesis that local weather storms are more a symptom of 

 disturbed conditions in higher regions. The same could be said of 

 earth currents.'^ However, since reflection probably occurs at the 

 surface of the earth one or more times on the transatlantic paths, as 

 will be discussed more fully in a paper now in preparation, signals using 

 these paths probably encounter surface weather conditions and earth 

 current influences at least three times. 



Correlation with Meteor Showers 



An attempt was made to find some correlation between meteor 

 showers and long-wave radio transmission. The sketchy nature of the 

 available meteor data does not provide a satisfactory basis for such a 

 study. Comparison of monthly average meteor hour rates ^^ with 

 monthly average radio transmission showed little or no promise of 

 significant relations. Although meteors probably do exert an influence 

 on radio transmission i^, 21, 22 j^y causing random variations of the 

 received signal, due to changes in ionization of the transmission path, 

 the evidence for major changes in field strength at times of meteoric 

 activity apparently is lacking in our data. 



During the four-month period centering on the Perseid shower,* 

 Pickard ^^ observed a considerable depression in the day reception of 

 long-wave signals from Nauen (23 kc.) and a considerable increase in 

 medium-wave signals at night from WBBM (1330 kc), with maximum 

 effects close to the day of meteor shower maximum. No such effect 

 was detected, however, in the data on the long-wave transatlantic radio 

 telephone circuits for the period from 1923 to 1933. In fact, as shown 

 in Fig. 4, if the Perseid shower could be credited with having con- 



* According to Olivier in Bulletin No. 8, the American Meteor Society (quoted 

 also in reference 23), the Perseid shower is by far the longest and largest annual 

 meteor shower and extends about 17 days on each side of August 11. 



