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



miniature lamp. The first lamp glows brightly, the second 

 little or not at all, showing that the oscillations have been 

 absorbed by the high resistance. 



(d) Substitute a coil of self-inductance, but of low resist- 

 ance, for the high resistance in (c). The result is the same 

 as before, owing to reflection and absorption of the waves by 

 the self-inductance. These experiments explain the fact that 

 the remote parts of the circuit are entirely devoid of oscilla- 

 tions. Thus, if there are connected in series with the dis- 

 charge a small lamp, a few meters of straight wire, a second 

 lamp, and a large self-inductance, the lamp nearer to the tube 

 will brighten up under the oscillations, while the other remains 

 dark, showing that a node of current is located at the self- 

 inductance, as if at the end of an open oscillator. 



I have repeated these experiments, using a hot-wire 

 ammeter instead of lamps, and have thus explored various 

 parts of the circuit for oscillations. An example of such 

 observations is 



shown in Table 



I. 



Each set of readings 



Table 



I. 





H i 





t\ ==.*"— to 



063 -243 





•J 80 



•093 "265 





•172 



•124 • -284 





•J60 



corresponds to a different external resistance. i is the current, 

 in amperes, given by the generator, i that indicated by the 

 hot wire close to the discharge, and i x the effective alternating 

 current, which in one case is nearly three times as large as the 

 direct current supplied. 



(e) If a coil of moderately large self-inductance is con- 

 nected close to the tube on each side, the discharge becomes 

 irregular and noisy, and no high-frequency oscillations appear. 

 There seems to be a critical value of the self-inductance at 

 which the oscillations suddenly cease. The cessation of 

 oscillations apparently shows that the glow-arc pulsations 

 cannot take place above a certain frequency. 



(f) Hold one terminal of an incandescent lamp of high 

 resistance in the hand, and touch the other terminal to the 

 discharge circuit near the tube. The lamp lights up, the light 

 being brighter when the lamp is on that side of the tube nearer 

 to a self-inductance, as it is here that the antinode of the 

 e.m.f. wave occurs. 



(g) By means of a frequency meter with a range of from 

 0*2 XlO 6 to 5X10 6 periods per second, the fundamental vibra- 

 tion and its upper harmonics can be detected. The best 

 method is to allow a coil of a few turns in the discharge 



