240 W. G. Cadi/ — Electric Arc between Metallic Electrodes. 



§ 16. As long as the circuit on at least one side of the tube 

 was free from too large self-inductance or resistance for a few 

 meters from the discharge, good oscillations could be obtained. 

 During most of the experiments large copper wires extended 

 from the tube about one meter apart for a distance of about 

 15 m with as few bends as possible. At the farther ends of 



Fig. 1. 



these wires were first, coils of large self-inductance, then the 

 regulating resistances, ammeter shunt, and electromotive force, 

 the latter consisting of a small 650 volt generator, in series 

 with which were connected at times about 350 volts additional. 

 The currents used v ranged from 0-0025 to 0'25 amp., the 

 strongest oscillations being produced at about 0*1 amp. 



A voltmeter connected across the discharge indicated, dur- 

 ing oscillations, about 215 volts when- the current was 0*1 amp. 

 This value, however, caunot be depended upon, for the pres- 

 ence of the voltmeter as a shunt across the arc greatly dimin- 

 ished the intensity of the oscillations (cf. § 31). The true 

 value of the mean discharge potential was obtained from a 

 knowledge of the current, e.m.f., and external resistance, and 

 it proved to be about 280 volts under conditions similar to 

 those described above. This is intermediate between the volt- 

 age drop usually found for an arc and that for a glow. 



When the electrodes were touched together and then 

 separated slightly, the discharge formed a minute spot of light 

 between their nearest portions. It was of a bluish color, not 

 very bright, and differed in appearance from either an arc or 

 a glow. The presence of oscillations penetrating through the 

 circuit as far as the self-inductance coils was shown by the 

 heating of a bolometric detector in the neighborhood, or by 

 the brightening of a pilot lamp in the discharge circuit (see 

 §18). 



§17. Alternating current as source of energy for oscilla- 

 tions. — Considering the shortness of the discharge-gap, it was 

 thought that a moderately high alternating e.m.f. might suffice 



