December 2, 1909] 



NA TURE 



141 



RESEARCHES IN RADIO-TELEGRAPH Y.^ 



^ADIO-TELEGRAPHY, popularly called wireless tele- 

 graphy, has outlived the tentative achievements of its 

 precocious infancy and obtained for itself a settled but 

 important position amongst our means of communication. 



This stage, however, has only been reached after a long 

 stiuggle with experimental difficulties and much labour in 

 analysing the processes involved. As many of these 

 matters are of general scientific interest, it is proposed, 

 during the present hour, briefly to summarise the results 

 of some recent research. 



It is well known that the nature of the earth's soil or 

 surface between the sending and receiving stations has a 





K-1 



1 Water. | W; 

 Spccijic Resista 

 .—Depth 



(Dr. 



"ihms per Metre Cittc 

 of W.ives 1000 feet in 

 Zenneck.) 



great effect upon electric waves passing over it. If the 

 surface is a very good conductor the wave hardly pene- 

 trates into it, but glides over the surface. If it is a poor 

 conductor the wave penetrates into it to a greater extent, 

 and the worse the conductivity the deeper the penetration. 



The materials of which the earth's crust is composed, 

 with some exceptions, owe their electric conductivity chiefly 

 to the presence of water in them. They are called electro- 

 lytic conductors. Substances like marble and slate, when 

 free from iron oxide, are fairly good insulators. Dry 

 sand or hard dry rocks are poor conductors, but wet sand 

 and moist earth are fairly good conductors. Sea water, 

 owing to the salt in it, is a much better conductor than 

 fresh water. The following table gives some figures, 

 which, however, are only approximate, for the specific 

 resistance of various terrestrial materials in ohms per 

 metre cube. It will be seen that dry sand or soils are of 

 very high specific resistance, and damp or wet sand or 

 clay fairly low. 



Table I. — Approximate Conductivity and Dielectric 

 Constant of various Terrestrial Materials. 



Specific resistance in Dielectric constant. 

 Material ohms per metre cube .A.ir=r 



to which electric waves of such frequency as are used in. 

 radio-telegraphy penetrate into the sea or terrestrial strata, 

 of various conductivities. For mathematical reasons, it i& 

 customary to define it by stating the depth in metres or 

 centimetres at which the wave amplitude is reduced to 

 1/ 6 = 0-367 of its amplitude at the surface. I have repre- 

 sented in a diagram some of Zenneck's results calculated, 

 for waves of looo feet in length, and for terrestrial surface 

 materials of various kinds, conductivities, and dielectric 

 constants (see Fig. i). You will see that in the case of 

 sea water an electric wave travelling over it penetrates only 

 to the depth of a metre or two, whereas in the case of. 

 very dry soil it would penetrate much deeper. Owing to 

 the conductivity of the soil, this movement of lines of 

 magnetic force through it sets up currents of, 

 electricity which expend their energy in heat. 

 This energy must come from the original store 

 imparted to the sending antenna, and therefore 

 the wave is robbed of its energy as it travels- 

 over the surface. 



Dr. Zenneck has discussed mathematically, in 

 a very interesting paper, the effect of the con- 

 ductivity and dielectric constant of the terrestrial 

 surface, soil or sea, on the propagation of a plain 

 electric wave over it, assuming the radiation to 

 be from an ordinary vertical antenna, and the 

 electric force therefore normal to the earth, and 

 magnetic force parallel to it. The result is to- 

 show that there are, broadly speaking, three cases 

 to consider. First, supposing the surface material 

 to be a good conductor, then the wave moves 

 over the surface and penetrates a very little way 

 into it. The electric force in the air over the- 

 surface is a purely alternating force vertical to 

 the earth's surface, and the magnetic force is an 

 alternating force parallel to it, and there is very 

 little subterranean electric or magnetic force 

 (Fig. 2, a). This is realised approximately or most 

 nearly in the case of radio-telegraphy over sea water. 

 Secondly, let the earth be assumed to have a very 

 poor conductivity and not a very large dielectric con- 

 stant, then analysis shows that the electric force m 

 the air has two components, one perpendicular to the 

 earth's surface and one parallel to it, and the resultant 

 is an alternating and a rotating force, the direction 

 of its maximum value being inclined to the surface 

 and leaning forward (Fig. 2, b). The wave-front 

 therefore slopes forward. Also there is a subterranean 

 electric force, showing that the wave is penetrating into 

 the soil, and there is therefore dissipation of energy owing 



If our earth's surface had a conductivity equal, say, to 

 that of copper, then the electric radiation from an antenna 

 would glide over the surface without penetration. In the 

 case of the actual earth there is, however, considerable 

 penetration of the wave into the surface, and therefore 

 absorption of energy by it. 



Brylinski, and also Zenneck, have calculated the depth 

 1 From a discrurse delivered at the Royal Institution, on Friday, June 4, 

 by Prof. J. A. Fleming, F.R.S. 



NO. 2092, VOL. 82] 



to the conductivity of the soil as the wave travels over 

 the surface. This case is realised when the wave travels- 

 over land composed of dry soil having a small dielectric 

 constant. Thirdly, let the earth be a very poor conductor, 

 having a small dielectric constant from 2 to 3, and ai 

 specific resistance of about 10,000 ohms per metre cube. 

 For example, very dry earth or sand. Then the investi- 

 gation shows that the electric force in the air has two- 

 components, one parallel to the earth's surface and one 

 perpendicular to it differing in phase, and the resultant 

 is represented by the rotating radius of an ellipse, the 

 maximuin value or major axis of which is inclined forward' 

 in the direction of the wave motion (Fig. 2, c). At the- 



