260 BELL SYSTEM TECHNICAL JOURNAL 



introduction of a single side-band system and toward the develop- 

 ment of single side-band equipment for use at these frequencies. 



Another method which might be employed to reduce this type of 

 distortion is to pick up the signal on a number of antennas spaced 

 more than about 10 wavelengths apart, since it is found that at points 

 this far distant from each other, while the general average signal 

 values are the same, the instantaneous values of the signals are ap- 

 parently random within the fading limits. By an automatic arrange- 

 ment for selecting the best signal from, let us say, three antennas 

 arranged in this manner, voice distortion can be diminished. 



During periods of magnetic storms, however, the signals are so 

 very much reduced in intensity at times that they cannot be heard 

 above the noise level. There appears to be nothing in the present art 

 which will fully cope with this situation. Of course, some of the time 

 which is now lost during these periods may be expected to be regained 

 by further transmission improvements. As was indicated earlier in 

 the paper, it is an interesting but rather discomforting fact that these 

 particularly severe conditions are due to some peculiarity in the con- 

 dition of ionization as indicated by the magnetic and earth current 

 disturbances referred to above and by the fact that aurora displays 

 are likely to be pronounced at these times. Furthermore, it appears 

 that the transmission is most adversely affected during these times 

 along paths which pass near the aurora zones surrounding the mag- 

 netic poles. This is indicated by the marked effect which these storms 

 had on the North Atlantic circuits while showing only a slight effect 

 on the South American circuit. 



The advantages to be gained by the use of directive antenna sys- 

 tems were touched on a little earlier in this paper. So far, most of the 

 gain has been obtained by sharpening the transmission in the hori- 

 zontal plane. This can be done advantageously only up to a certain 

 point, corresponding to an antenna spread of from six to ten wave- 

 lengths—at any rate for transatlantic signals — and representing a 

 gain when a reflector is used of about 15 db. A further gain of 3 to 5 

 db can be obtained by sharpening in the vertical plane; and while 

 a still greater gain can at times be obtained in this manner, the pro- 

 cedure has so far appeared not worth its possibilities of trouble. This 

 is due to the fact that with varying conditions in the upper atmosphere, 

 the waves as they reach the receiving station apparently approach 

 from different vertical angles and care must be taken not to build an 

 antenna with such a sharp vertical characteristic that the received 

 waves will fall on the antenna at such an angle that its calculated 

 gain cannot be realized. We have, in fact, constructed several an- 



