10 BELL SYSTEM TECHNICAL JOURNAL 



to measurements made on telegraph traffic, a portion of the large 

 variation in time may be due to experimental errors. 



The wave-like variations in field intensity observed on the Rocky 

 Point-Houlton 60-kc. path on 1/21/34 and 4/15/34 have the appearance 

 of interference fringes,^- ^^ and if explained on this basis the question 

 at once arises whether or not interference is the principal cause of the 

 sunset cycle. If daylight communication is accomplished either by 

 means of a single reflected ray in combination with a ground wave, or 

 by the resultant of a number of reflected rays of nearly constant com- 

 plex propagation constants, at sunset the occultation of the ionizing 

 rays from the transmission medium by tangency with an occulting 

 layer, might reasonably be expected to produce variations in both the 

 real and imaginary portions of the propagation constant of the medium, 

 thereby causing interference fringes through the variation of the rela- 

 tive phase of different rays. 



As an example, we may assume that on the Rocky Point-Houlton 

 transmission path the ground ray provides the principal agency of 

 daylight communication by means of atmospheric refraction and 

 diffraction. At the approach of sunset, reduced ionization between 

 the earth and the reflecting layer reduces the attenuation in the path 

 of the reflected ray, producing a resultant received field which is a 

 function of the relative intensity and phase of the two waves. As the 

 effect of the sun's active rays becomes still less, the decreased ionization 

 of the layers produces variations in the phase of arrival of the reflected 

 wave, either through changes in the propagation constant, or on 

 account of greater path length occasioned by an increased virtual 

 height of the reflecting layer. 



Now since the resultant received field is the vector sum of the two 

 rays, when one ray is much smaller than the other, variations in their 

 relative phase will produce small amplitude fluctuations in the re- 

 sultant, with maxima and minima equal to the sum and difference of 

 the two components. As the two components approach equality, 

 however, the maxima will approach twice the intensity of one com- 

 ponent ray, while the minima will approach zero, thereby producing a 

 very deep "dip" in the received field. 



The "D" Layer 



It has been suggested by Heising,^" Appleton " and Chapman "^^ 

 that passage through a low-elevation layer of ozone produced by solar 

 ionization, is one of the principal causes for the daylight attenuation 

 of the reflected ray.* Radio transmission measurements at broadcast 

 frequencies, reported by the U. S. Bureau of Standards " and by the 

 Australian Council for Scientific and Industrial Research,^ show a rapid 



