AUDIO FREQUENCY ATMOSPHERICS 505 



and weak tweeks observed before sunset or after sunrise probably 

 originate at considerable distance respectively to the east or west 

 within regions not exposed to sunlight. 



The multiple reflection theory of tweeks, as explained above, con- 

 cerns a single wave train originating in a disturbance located near one 

 of the reflecting surfaces. It may be shown that an impulse originating 

 anywhere in the intervening space might produce a similar eff'ect, 

 although the initial frequency would be altered by the location in 

 altitude. Furthermore, were the point of origin well separated from 

 both surfaces, two simultaneous wave trains diff^ering somewhat in 

 rate of frequency change would occur. Phasing effects, which might 

 be attributed to this have been found in several oscillograms. 



Based on the multiple reflection theory, the curve C in Fig. 4 was 

 calculated assuming the point of origin to be located near the earth's 

 surface. The altitude of the reflecting layer was taken as 83.5 km. 

 (55 miles) and the distance between source and observer as 1770 km. 

 (1100 miles). While this curve only roughly approximates the form 

 of the tweek curves of Fig. 4, an explanation of the discrepancy may 

 lie in a variation in effective layer height in accordance with the change 

 in angle of incidence of the successive impulses. Such a relation in 

 the case of radio frequencies has been described by Taylor and 

 Hulburt.io 



Comparison of the lower limiting frequencies of individual tweeks 

 with an oscillator calibrated in small steps has shown at times an 

 almost continual drift in frequency. This may be interpreted as a 

 corresponding variation in the effective height of the reflecting layer. 

 In one five-minute period during complete darkness, examination of 

 24 tw^eeks showed the lower limiting frequency to vary irregularly 

 between 1690 and 1720 c.p.s. This indicates a variation in effective 

 layer height between approximately 88.5 and 87 km. The variations 

 of lower limiting tweek frequencies noted at our various observation 

 points have indicated the reflecting layer to vary between 83.5 and 

 93.2 km. during the hours of complete darkness. No marked vari- 

 ations of mean tweek frequency, in respect to either season or latitude, 

 have been observed. 



During experiments carried out in New Jersey and New Hampshire,* 

 a calibrated tone producing apparatus was available whereby fre- 

 quencies of musical atmospherics, as observed by ear, could be closely 

 followed. It was found that in addition to tones, which could be 

 considered as individual tweeks, there appeared at times a slight, 

 almost unbroken resonance quality in the static. This resonance was 



1" A. H. Taylor and E. C. Hulburt, "Propagation of Radio Waves," Pliys. Rev., 

 27, p. 189, February, 1926. 



