November 7, 1912] 



NAT URL 



293 



attenuaies much less fast than a pure Hertzian wave 

 ana can iravel round the curvature of the earth quite 

 easily. 



On the whole, we may say that the theory, as given 

 by Zenneck anci Sommerfeld, is a valuable attempt 

 to bridge over the very serious gap in our knowledge 

 of the reasons for certain well-ascertained facts in 

 radio-telegraphy. Nevertheless, there are still un- 

 explained difliculties. 



Another unsolved problem in radio-telegraphy is the 

 explanation of the effects of the atmospheric conditions 

 and daylight upon it. 



The suggested explanations are in many respects 

 imperfect. In the earliest days of radio-telegraphy it 

 was found that atmospheric electric discharges pro- 

 duced irregular and false signals, which sometimes 

 greatly interfered with working. These were more 

 objectionable at the time when the receiving instru- 

 ment was a coherer of some kind associated with the 

 Morse printer. Nowadays, when the reception is by 

 telephone, it is usual to have the spark frequency at 

 the sender hiarh enough to give a shrill note in the 

 telephone. The receiving operator can then distin- 

 guish, to a great extent, between the clear musical 

 note of the right signals and the lower squeaks or 

 grunts in the telephone due to atmospheric dis- 

 charges. Nevertheless, at certain times and in certain 

 regions the so-called atmospherics present serious 

 obstacles to radio-telegraphic communication. 



When we turn to the effect of sunlight on the 

 propagation of radio-telegraphic waves, which was 

 discovered and described by Mr. Marconi in 1902, we 

 find that even after ten years we are still, intellectually 

 speaking, very much in the dark as to the reason for 

 this daylight effect. 



The first observation made by him in 1902 was that 

 by night signals could be received over sea from the 

 Poldhu station at a distance of 2099 rniles, whereas 

 by day the same kind and type of signal ceased to 

 be detectable at about 700 miles ; also that at the 

 time when the sun rose over the sending station the 

 signals at 700 miles' distance quite quickly became 

 very weak.'^ 



Of recent years he has noticed that in the morning 

 or evening, when the boundary between light and 

 darkness occurs, about half-way across the Atlantic 

 signals sent across become weak. 



Also he has noticed that in sending with a coupled 

 transmitter radiating waves of two wave lengths, 

 whereas the longer wave length is the one generallv 

 received, there are certain periods at sunrise and 

 sunset when the shorter wave gives the best signals. 



It has also been pointed out, both by Mr. Marconi and 

 G. W. Pickard, that soon after the time of sunrise 

 at the sending station there is a very pronounced 

 decrease in the strength of the signals received a few 

 hundred miles', or at some considerable distance from 

 a power station, but that after sunrise there is a 

 partial recovery of strength. There is also a gradual 

 rise in the strength of the signals soon after "sunset, 

 and a very pronounced maximum value after or about 

 midnight. An interesting curve has been given by 

 Prof. Pierce in his book on "Wireless Telegraphy," 

 p. 135, taken from Pickard's observations showing the 

 general variation of the strength of received signals 

 at a distance of 600 miles from the Marconi station at 

 Glace Bay during the hours of the day and night. 

 It appears that the current in the receiving telephone 

 at midnip-ht was about thirty times greater than by 

 day. Confirmatory observations have been published 

 by the Telefunken Co. 



1- See G. Marconi, Proc. Roy. *^oc Lond., June 12, 1902. A note on 

 the effect of daylight upun the propagation of electromagnetic influences 

 over long distances. 



NO. 2245, VOL. 90] 



Two theories have, so far, been proposed to explain 

 this effect : — 



1. The original suggestion of Mr. Marconi (which 

 has been tentatively adopted by Prof. Zenneck) was 

 that it is due to the effect of light in discharging the 

 sending antenna, so that it does not reach at each 

 oscillation such a high potential by day as in darkness. 



2. The theory that the daylight effect is due to the 

 ionisation of the air by sunlight giving it increased 

 conductivity, and so producing absorption of the 

 electric waves. 



Neither of these theories seems to meet all the facts. 

 If the daylight effect were an action of light on the 

 j sending antenna alone, it should be produced inde- 

 ' pendently of the distance of the receiving station, 

 I whereas it is essentially a cumulative or long-distance 

 1 effect. 



I Again, so far as measurements of the electric con- 

 ductivity of air have been made, they do not give 

 support to the theory that the daylight effect is due 

 to air conductivity produced by ionisation, because 

 measurements of this conductivity show it to be too 

 small to account for the observed wave attenuation. 



Prof. G. W. Pierce has made calculations which 

 show that the air conductivity would have to be 

 100,000 times greater than it actually is to account 

 for even a part of the observed effect at 3000 km. 

 distance. 



Prof. Zenneck also agrees that atmospheric con- 

 ductivity bv ionisation cannot account for the 

 phenomena. 



Let us, in the first place, consider theoretically the 

 propagation of an electromagnetic wave through a 

 medium possessing dielectric constant (K) and con- 

 ductivity (o-). 



It is quite easy to obtain an expression for the 

 absorption coefficient of such a medium. Starting 

 from the Maxwellian equations as above, we have the 

 quantity denoted by fe on a previous page defined by 

 the equation — 



since It is a complex, it can be represented bv 



k = a + j0. 



Hence we have — 



3-2 = _ mK/'-' , / 1 67rV-/x-j)- ^"Ky 

 '^ 2?-' - \ AC* 4^-* ' 



and if 



/K 



4C* 4C-* 



is a small quantity, as it is when the 



conductivity is small and the frequency f /2t is large, 

 w'e have then — 



If we consider a plane wave the amplitude of which 

 varies as eHlix-pt)^ it attenuates in amplitude to i/e 

 of its initial value in travelling a distance — 



If p is the resistivity 

 p = (9+io")(r, and we hav( 



in ohms per c.c, then 



/i 6o7r ■ 



We can therefore determine at once the absorption 

 effect of any conducting dielectric of which the re- 

 sistivity in ohms per c.c. and the dielectric constant 

 are known. Thus, supposing the dielectric constant 

 is 9, and the resistivity 500 megohms per centimetre 

 cube, the value of i/3 would Ije nearly 80 km., or 

 fifty miles. This is, then, the distance in which the 



