2i2 BELL SYSTEM TECHNICAL JOURNAL 



originalK- with constant amplitude and frequency can be recei\ed as 

 one of variable amplitude only if certain characteristics of the medium 

 are variable with the time. So far as the atmosphere is concerned, 

 these characteristics may be the distribution of electrons and heavier 

 ions and the intensity and direction of the earth's magnetic field. 

 If these are functions of the time, the velocities, bending, absorption 

 and rotation of the plane of polarization will all be variable, 

 the amplitude of variation depending upon the variations of A^, 



-r-, //, — J-, as well as the frequencv of the wave, the effects being in 

 as as 



many cases magnified greatly in the neighborhood of the critical fre- 

 quency. These effects are obviously sufficiently numerous to account 

 for fading of almost any character and suggest a number of experi- 

 ments to determine the most effective causes. The question of rota- 

 tion of the plane of polarization, fading and distortion is now being 

 examined experimentally. 



From the formulas it is clear that the velocity, curvature and absorp- 

 tion of an electromagnetic wave as well as the rotation of its plane of 

 polarization can all be affected by a time variation in the intensit\' 

 and direction of the earth's field. An examination of the probable 

 time and space variations of each, however, lead us to the conclusion 

 that these are not of primary importance in determining large ampli- 

 tude fading except, perhaps, during magnetic storms. One result of 

 the last two years of consistent testing between New York and London 

 at about GO, 000 cycles has shown that sexere magnetic storms are 

 always accompanied by corresponding \ariations in the strength of 

 received signals. Thus, although the earth's magnetic field can 

 well exercise a large influence upon the course and attenuation of 

 radio waves, it does not seem likely that its time variation is ordi- 

 narily a large contributing cause to fading. 



This leaves as the probable principal cause of time variations the 

 number and distribution of ions in the earth's atmosphere. It is 

 impossible in this paper, which is devoted primarily to a development 

 of a theory of transmission involving the earth's magnetic field, to 

 consider adequately all the possibilities resulting from changes in 

 ionic distributions, but some general remarks may be made. Imagine 

 a wave traveling from the source to the receiver. At a short distance 

 from the source the wave front will be more or less regular but as it 

 progresses, due to the irregularities in ionic distribution, the wave 

 front will develop crinkles which become exaggerated as the wa\e 

 goes on. These crinkles in the wa\e front will be due to irregularities 

 in the medium and can be obtained by a Huyghen's construction at 



