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



self-inductance was about 20,000 cm . This primary coil acted 

 inductively on a secondary coil at varying distances from it. 

 The secondary coils used had self-inductances varying from 

 T90 cm to 1,170,000™. 



In series with the secondary coil was an adjustable condenser 

 of zinc plates separated by sheets of glass, the whole immersed 

 in kerosene. The secondary circuit contained also a hot-wire 

 ammeter of about two ohms resistance. As detecting devices 

 a thermo-element, and also a bolometer in a tertiary circuit 

 loosely coupled with the secondary circuit, were tried, but 

 neither method gave as good results as the hot wire in the 

 resonating circuit itself. For the purpose in hand, the damp- 

 ing caused by the resistance of the hot wire was not serious. 



According to the method just described, it will be seen that 

 the e.m.f. induced in the resonating circuit is proportional to 

 the current in the discharge circuit. That is, if M represents 

 the mutual inductance between the two coils, 



e > = M ft- 



Since this expression involves the frequency, it follows that 



* 2 = *»,/, 0) 



where Jc is a constant. 



By using secondary coils of varying self -inductance and each 

 time adjusting the capacity for resonance, the fundamental 

 vibration together with its upper harmonics could be investi- 

 gated for frequency, and to some extent for intensity. It was 

 always found that the hot-wire deflections corresponding to the 

 fundamental vibration were very small. This agrees with 

 the relation expressed in equation (1), since f has its smallest 

 value for the fundamental, and the hot-wire deflections are 

 proportional to the square of e v Remembering also that the 

 wave-form of the oscillations is certainly very complex, it is 

 evident that the deflections corresponding to some of the upper 

 harmonics will be much greater than those for the funda- 

 mental. I found it necessary to use very close coupling in 

 order to detect the fundamental at all. Loose coupling could 

 be used for harmonics from the third to the ninth, while above 

 that the coupling had to be made close again, on account of 

 the diminishing value of i x . 



§ 21. When the capacity of the discharge circuit was 

 increased by touching a piece of metal about 10 cm square to a 

 point near the discharge tube, the frequencies were slightly 

 lowered, as would be expected. 



§ 22. Introducing a coil of a few turns of wire into the 

 discharge circuit lowered the frequency. A coil of larger 



