AUDIO FREQUENCY ATMOSPHERICS 515 



the telephone. So far as it can he expressed in letters the tone sounded 

 about like peou}'' From the physical viewpoint, it was an oscillation 

 of approximately constant amplitude, but of very rapidly changing 

 frequency . . . beginning with the highest audible tones, passing 

 through the entire scale and becoming inaudible with the lowest 

 tones. . . . The entire process lasted almost a full second." 



Barkhausen presents two possible explanations for these sounds. 

 The first assumes the presence of a reflecting layer in the upper 

 atmosphere. An electromagnetic impulse originating at the earth's 

 surface arrives at a distant receiver first over the direct path and then 

 from reflections in the order 1, 2, 3, to n. Such a series of reflections 

 would result in a wave train of rapidly diminishing frequency becoming 

 asymptotic to a value dependent upon the height of the reflector. 



The second of Barkhausen's theories depends upon ionic refraction 

 in the Heaviside layer, resulting in the breaking up of an impulse into 

 its component frequencies and a delay in the transmission of the lower 

 frequencies with respect to the higher. It gives a rate of frequency 

 progression which varies with distance and with the refractive index of 

 the medium. 



Eckersley ^^ in a paper on "Musical Atmospheric Disturbances" 

 discusses apparently the same type of atmospherics. As an experi- 

 mental background he notes frequent observations of audio-frequency 

 disturbances received over large radio antennas. He states: "These 

 (tones) have a very peculiar character : the pitch of the note invariably 

 starts above audibility, often with a click, and then rapidly decreases, 

 finally ending up with a low note of more or less constant frequency 

 which may be of the order of 300 to 1000 a second. 



"The duration . . . varies very considerably; at times it may be 

 a very small fraction of a second, and at others it may be even 1/5 of 

 a second." He observes that they are infrequent in morning, in- 

 creasing throughout the day and reaching a maximum during the night. 

 He develops a theory based on ionic refraction to account for these 

 disturbances. 



It appears that in these latter observations both swishes and 

 tweeks were heard, but were not recognized as distinct phenomena. 

 Such an error might be attributed to the irregularities of response 

 which are common in the ordinary telephone receiver. 



Barkhausen's first theory fails to explain swishes because of their 

 upward as well as downward progression, long duration and frequency 

 range. The theory, as previously pointed out, is adaptable to the 



^^ Peou slowly pronounced in a whisper excellently portrays a descending swish 

 accompanied by the rushing sound. 



18 T. L. Eckersley, Phil. Mag., 49, p. 1250, 1925. 



