1902.] on Musical and Talking Electric Arcs. 65 



arcs fed from these mains to give out the characteristic noises, even 

 though considerable distances intervene between the points of con- 

 nection to the mains. 



In some experiments on the resistance of the arc, in the course of 

 which a small alternating current of high frequency was added to a 

 steady direct current, I found that if even as small an alternating 

 current as yoVo ampere was passed through an arc taking 10 

 amperes direct current, a distinctly audible note was produced by the 

 arc at frequencies ranging from a few hundreds up to 8000 periods per 

 second ; and further experiments showed that if the added alternating 

 current was only J^- to -^^ ampere, the note did not become inaudible 

 until frequencies as high as 30,000 periods per second were 

 reached.* 



This great sensibility of the arc to small and rapid changes in 

 the current through it at once suggests that the arc might be used as 

 a telephone receiver. This was first tried experimentally by Herr 

 Simon,-]- who caused the current from an ordinary microphone to flow 

 round one coil of a mutual induction, or transformer, the second coil 

 of which was placed in series with an arc. The variations in current 

 flowing through the microphone due to speaking into it, induce cor- 

 responding currents in this latter coil and through the arc, causing 

 the arc to give out sounds similar to those made near the microphone. 

 By this means the arc may be caused to sing, whistle, or even talk 

 distinctly. 



There arc a large number of other ways of superposing the micro- 

 phone current or the direct current through the arc, the microphone 

 may be shunted by an inductive or non-inductive resistance and 

 placed directly in series with the arc ; or the microphone may be 

 connected in series with a battery and highly inductive coil, the 

 terminals of the latter coil being connected direct to the arc terminals 

 through a condenser, or the microphone and battery may be connected 

 to the primary of a transformer the secondary of which is connected 

 in series with a condenser as a shunt to the arc. I have usually 

 found the last-mentioned arrangement give the best results. J 



Any good microphone will work, though the resulting speech from 

 the arc varies greatly according to the microphone and the battery 

 power used. A " Hunningscone " or " Delville " with two accu- 

 mulators will, if carefully adjusted, give loud and distinct speech. 



As the sound-waves are produced by the variations in the volume 

 of the vapour column, it is natural that the loudness will depend on 

 its size* In order to obtain loudness it is therefore necessary to use 

 a long flaming arc and a fairly large current ; such arcs can be best 

 produced by using cored carbons or by introducing impurities, such 



* Journal of the Institution of Electrical Engineers, 1901, vol. 30, p. 238. 

 t Annalen der Physik und der Chemie, 1898, vol. 64, p. 233. 

 X Journal of the Institution of Electrical Engineers, 1901, vol. 30, p. 239 ; 

 ElektrotechniBche Zeitschrift, 1901, pp. 197 and 510. 



Vol. XVII. (No. 96.) r 



