296 



NATURE 



[November 7, 1912 



part horizonlal, and that radiation takes place most 

 energetically in the opposite direction to which the 

 free end of the horizontal wire points." Also by the 

 law of exchanges a bent antenna which radiates un- 

 equally absorbs unequally in different azimuths. The 

 question is as to the explanation of the action of this 

 bent antenna. 



In igo6, starting from a suggestion by Sir Joseph 

 Larmor, I gave a theory based on the view that the 

 bent antenna is equivalent to a combination of an 

 open and closed circuit and assumed the earth to be 

 a perfect conductor. The objection has been raised to 

 this theory that it implies that the directivity should 

 fall off with distance. 



Experiments have not yet been made, so far as I am 

 aware, on a sufficiently large scale and at sufficiently 

 great distances to settle this point, and the experi- 

 mental problem is undoubtedly complicated by the 

 effect of the nature of the soil'surface over which the 

 waves travel in different dielectrics or regions. Never- 

 theless, Mr. Marconi's experiments show that the 

 directivity persists for several hundred miles. Recently 

 the problem has been discussed bv H. von Hoerschel- 

 mann in the Jcihrbuch dcr Drahtlosen Telcgraphie 

 (see vol. v., p. 15, 1901). According to his theory, the 

 effect of a bent antenna is entirely due to vertical 

 electric currents which are produced' in the earth just 

 under the horizontal part of the bent antenna. Hence 

 the directivity depends on the conductivity of that 

 region of the earth. 



Starting from the same system of equations as 

 Sommerfeld, he obtains expressions for the vector 

 potential or Hertzian function which show that the 

 function which determines the vertical field in the 

 earth and in the air is dependent on the azimuthal 

 angle, in such manner that there is no directivity if 

 the earth is a perfect conductor, but that if the earth 

 has a finite conductivity under the antenna then an 

 unsymmetrical radiation takes place and the direc- 

 tivitv, once created, persists even though the waves 

 travel out later on over a good conducting surface. 

 The mathematical work in Hoerschelmann's paper 

 needs careful consideration by pure mathematicians 

 to test whether the transformations of the Besselian 

 functions he employs are valid and his analvsis cor- 

 rect. He points out that as a consequence of his 

 theory a bent antenna situated over a sea surface 

 should not have the same degree of directivity as one 

 situated over a poor conducting soil. These conclu- 

 sions could easily be checked experimentally. 

 _ The general result of all the theoretical investiga- 

 tions of Zenneck, Sommerfeld, and Hoerschelmann is 

 to sho\v that the function of the earth in radio-tele- 

 graphy is by no means confined merelv to guiding a 

 space wave, but that it fulfils a most important func- 

 tion in assisting to create surface waves and in per- 

 mitting cartel currents which have directive effect. 

 Recent experiments with antenna laid on the ground 

 or under the ground bv Dr. Kiebitz have directed atten- 

 tion afresh to the matter, although many of Kiebitz's 

 results seem only a repetition of those' obtained by 

 Marconi in t<)o6 with antennae laid on the ground or 

 a little above it." 



Kiebitz used as receiving antennae wires carried on 

 insulators placed in ditches about one metre deep. 

 The ends of the wires were earthed through con- 

 densers. The receiving appliance was at the centre. 



Bv such antennae properly oriented he found he 

 could receive signals from 'all the principal radio- 



'^ See G. \farc"ni on methods whereby the radiation of electric waves 

 m.aybe mainly confined to certain directions, etc., Proc. Roy. See. Lond., 

 vol Ixxvii., p. 413. IQ06. 



'« See Dr. F. Kiebitz's " Recent Experiments on Directive Wireless 

 Teleerarhy with Earthed Antenna." T/ie F.ltcMcian, March 8, 1012, 

 vol. twin., r. B68. 'VI 



^o. 2245, VOL. go] 



telegraphic stations in England, France, and 

 .•\meri(a. 



inere is no need to assume that these received 

 waves are propagated through the deep strata of the 

 earth. The eifects are exactly what might be ex- 

 pected from waves travelling over the surface. 



The chief interest of Kiebitz's experiments lies in 

 the confirmatory proof they give that an elevated 

 antenna is not necessary for reception. On the other 

 hand, for long-distance transmission an elevated 

 aerial wire or one raised above the earth is requisite. 



The chief problem yet to be faced in connection 

 with sending antennae is to find a form of antenna 

 which will radiate a large power, say, 100 to 500 kw., 

 at a relatively low frequency or long wave length 

 consistently with high antenna efficiency. 



There is room for an immense amount of research 

 yet on improved forms of antenna. When we con- 

 sider that the function of a sending antenna is some- 

 thing like that of a gas mantle or gas-fire radiator, 

 viz. to transform into radiation of desired wave length 

 as large a fraction as possible of the supplied energy, 

 and remember what has been done in the correspond- 

 ing luminous problem, it is easy to see that countless 

 questions of great practical value in connection with 

 antennas for radio-telegraphy remain unsolved. 



The ingenious methods of directive telegraphy due to 

 Bellini and Tosi deserve mention, and suggest that, 

 in the case of wireless plant erected on ships means 

 for instantly locating the direction of the arriving 

 waves is a matter of the greatest importance. 

 Although the practical problem is to some extent 

 solved, there is room for further invention in con- 

 nection with it. 



We can scarcely leave this discussion without some 

 mention of the state and prospects of wireless tele- 

 phony. 



The essential condition of success in transmitting 

 speech is the possession of means for creating un- 

 damped oscillations or alternating currents a fre- 

 quency not less than 20,000. When the Poulsen arc 

 generator was first introduced it was hailed as a 

 solution of the problem, but practical experience has 

 shown that whilst experimental feats can be per- 

 formed with it, it has not the simplicity and ease of 

 manipulation required for commercial work. The 

 modification recently introduced by Mr. E. L. Chaffee, 

 consisting of a copper-aluminium arc in damp hvdro- 

 gen, the arc being formed between two closely 

 adjacent plane surfaces, appears to be an improve- 

 ment. The practical solution seems, however, to be 

 in the perfection of some simple, easilv managed form 

 of high-frequency alternator. The ingenious inven- 

 tions of Goldschmidt in utilising the properties of the 

 polyphase motor to increase ifrequencv have been 

 developed by the Lorentz Company of Berlin, and 

 seem likely to result in the production of a practical 

 form of extra-high-frequency alternator suitable for 

 radio-telegraphv and radio-telephony in a practical 

 high-frequencv machine. 



In addition to the generator, inventors have wrestled 

 with the difficulties of makine a microphone trans- 

 mitter which shall be able to carry a large current 

 without heating. To conduct \vireless telephony over 

 anv distance we have to modulate in accordance with 

 the wave form of the speaking voice a verv large 

 antenna current. The problem has to some extent 

 received solution in the liquid microphone of Majo- 

 rana, the relay microphone of Dubilier, and a recently 

 invented heavy-current microphone of Riihmer. 



We seem to be, however, on the track of mechanical 

 means for trenerating undamped oscillations and 

 microphonic means for modulating them, and wire- 

 less telephony is therefore even now a practical matter 



