W. G. Cady — Electric Arc between Metallic Electrodes. 211 



to maintain the discharge.* Accordingly the voltage of the 

 60-cycle mains was stepped up to about 880 volts and used in 

 place of the usual direct-current supply. With an effective 

 current of about 0*06 amp., oscillations of fair intensity were 

 produced, with which most of the experiments described 

 below could be performed. 



This use of the alternating current to produce high-fre- 

 quency oscillations suggests that the effect described by Dud- 

 clellf in his recent paper on " short spark phenomena" may be 

 due to pulsations of the kind described in the present article. 



§18. Properties of the oscillations. — The following simple 

 experiments illustrate in a striking manner the properties of 

 these oscillations. Unfortunately all attempts at increasing 

 the output of energy have been fruitless, for the maximum 

 potential difference is only that of a glow discharge, and the 

 discharge degenerates into a steady arc if the current is 

 increased much above 0*2 amp. A close-coupled resonance 

 transformer reacts upon the discharge, reducing the oscilla- 

 tions. Still, for purposes of demonstration the oscillations 

 may find a certain field of usefulness, for they exhibit the 

 chief properties of high-frequency alternating currents with 

 extremely simple apparatus. The supply e.m.f. may be alter- 

 nating, but it is better direct, and the latter is here assumed. 



(a) Connect in series with the tube a direct-current ammeter 

 and an incandescent lamp of about 0*5 amp. capacity. When 

 the electrodes are touching the ammeter reads, say, 0*2 amp. 

 and the lamp glows dimly. On separating the electrodes the 

 ammeter indicates a diminished current, but if oscillations are 

 present the lamp bums brightly, showing that the effective 

 alternating current is greater than the direct current supplied 

 by the generator. From the increase in brightness of the lamp 

 a rough estimate can be made of the efficiency of conversion 

 of direct into alternating current. A miniature lamp, or one 

 of several hundred ohms resistance, may be used equally well. 

 In nearly all of the work an 8-volt 2-candle power lamp was 

 kept in series with the discharge tube, and it was found to 

 serve as an excellent indicator of the presence of oscillations. 



(b) Connect a few turns of heavy wire in series with the 

 tube, and in parallel with this coil place a high-resistance 

 lamp. The latter glows brightly. It was by this means that 

 the first estimate of the frequency of the oscillations was made, 

 in terms of the self-inductance of the coil, and of the resist- 

 ance and volts (estimated from the light) of the lamp. 



(c) Connect in series with the tube a miniature lamp, then 

 et resistance of several hundred ohms, and finally a second 



*Cf. Peukert, Elektrot. Zeitschr., xxiii, p. 562, 1908. 

 fDuddell, Proc. Phys. Soc. London, xxi, Part III, p. 275, 1909. 



