AUDIO FREQUENCY ATMOSPHERICS 513 



like nature obtained by us and others indicate a correlation between 

 swish and magnetic disturbances. The accepted connection between 

 auroral and magnetic field variations might justify a supposition that 

 aurorae and whistling tones may be directly related as indicated by 

 the New Hampshire observations. An assumption that the tones 

 originate at the altitudes usually occupied by auroral displays might 

 lead to an explanation of the apparent absence of marked diurnal 

 variations in the swish tones. The observed correlation between 

 certain atmospheric crashes and the subsequent swishes appears to 

 indicate either dependence of the latter on the former or origin of the 

 two from a common source of energy. The first assumption points to 

 multiple reflection or dispersion phenomena which produce either 

 ascending or descending tones. The time lag between the static im- 

 pulse and the following swish would indicate either a low velocity or 

 the traversing of a great distance. In either case, low attenuation is 

 indicated by the long duration of some tones. It appears possible 

 that the two radiations may result from sequential events occurring 

 in the upper atmosphere by means of which non-musical as well as 

 musical atmospherics are produced. Assuming an emission of energy 

 which persists more or less steadily over a period comparable with 

 the duration of a swish, it is possible to account for the approximately 

 uniform amplitude of a swish without the necessity of assuming a 

 very low damping. 



It is suggested that swishes may be related to the occasionally ob- 

 served phenomenon of swinging and flashing auroral beams. In this 

 case it appears necessary to consider a cyclic process in the behavior 

 of the aurora which w'ould account for the time lag between the radi- 

 ation of an initial static disturbance and the following varying tone. 

 The varying tones might be produced by energy radiated from swinging 

 beams resonating w'ithin the space separating beams or in the space 

 between a beam and a stationary reflecting layer. 



It might be possible for standing waves to occur within a beam, 

 variations in the length or other constants of the path producing the 

 varying tones. 



A correlation between swish and auroral phenomena is indicated in 

 statements by witnesses of auroral displays. Professor Chapman '^ 

 reviews the testimony of many observers who have witnessed auroral 

 displays at extremely low altitudes. Some attest to having stood 

 within the glow and to having heard, directly from the atmosphere, 

 disturbances accompanying the visible phenomena. Some of their 

 sound descriptions follow: 



'- Prof. S. Chapman, " Audibilitvand Lowermost Altitude of Aurora," Xature, 127, 

 p. 341, March 7, 1931. 



