482 BELL SYSTEM TECHNICAL JOURNAL 



a fairly close copy of the other, shifted a certain distance along the 

 horizontal axis. From the magnitude of the shift it is possible to 

 compute the strength H of the earth's magnetic field; or let me rather 

 say, it is possible to compute a numerical value which must agree 

 with H, or else the theory will be vitiated. When the computation 

 is made, the value turns out to be just a few per cent less then the 

 field strength at ground-level. The action of the field through the 

 electrons on the waves is exercised only in the ionosphere, which is 

 hundreds of kilometers up in the sky; and it is quite reasonable to 

 believe that these few per cent are actually the falling-ofif in the field 

 strength from the ground up to that level. Such is the present belief, 

 and many of those who work in terrestrial magnetism are happy over 

 the prospect of measuring thus the field in regions where there seems 

 to be no greater hope of anyone ever actually going, than of going to 

 the moon. 



Actually Fig. 4 shows data obtained in England, which is far from 

 the equator; I point this out in order to mention that even when the 

 waves are traveling obliquely to the earth's magnetic field, there is 

 a separation of the signals into pairs of echoes, and these are still 

 amenable to theory. In this general case of oblique transmission, 

 the waves are polarized elliptically — a feature difficult to visualize 

 without a certain amount of specialized knowledge, but lending itself 

 to some very neat and pretty experimental tests.^ In the special case 

 of transmission parallel to the magnetic field, waves initially plane- 

 polarized should remain of this character but their plane of polarization 

 should rotate as they proceed. There are indeed so many curious and 

 interesting details of the influence of the field through the electrons 

 on the waves, that a writer must be ruthless in ignoring them if he 

 is to observe decent limits of space. I will mention only in closing 

 that the "gyro-frequency" eH jlTrmc— -which, in our latitudes is around 

 1.3-10^ mc. — plays the part of a resonance-frequency. Waves too 

 close to this frequency are liable to great, not to say distressing, 

 anomalies in transmission. Theoretical statements about waves in 

 general are likely to assume two difi^erent forms, one appropriate to 

 those of frequency higher and the other to those of frequency lower 

 than the gyro-frequency; it is the former which appears in this paper. 



In Fig. 14 there appears something which, if the war had begun 

 before its discovery, would perhaps have been called a "blackout." 



* It may perhaps be regarded as obvious that the ellipse of polarization should 

 be described in opposite senses in the northern and southern hemispheres, since in 

 one hemisphere the magnetic lines of force are coming out of the ground, in the 

 other they are diving in. This inference was tested by a special experiment in 

 Australia, and the result was taken as establishing the "magneto-ionic theory," 

 as this general theory is often called. 



