W. G. Cady — Electric Arc between Metallic Electrodes. 247 



columns give the frequencies in millions per second. The har- 

 monic relation is in general clear, considering that in many cases 

 the settings were difficult and resonance ill-defined. The value 

 of the fundamental given in the first column is that which best 

 satisfies the observed upper harmonics. Owing to the loose- 

 ness of the coupling, the fundamental could not be directly 

 observed in all cases. As a test of the accuracy, the third col- 

 umn contains the quotients obtained by dividing the observed 

 frequencies by the fundamental. These quantities should be 

 integers representing the number of the harmonic. 



The differences between the wave-forms of the various 

 groups are due to differences in current and changes in the 

 electric circuit. 



Fig. 2. 





[ 





















































i 





















































< 







































































































00 













































































































a 





$f 















































, 





















































i 



». 













































50 

































































1 

































































































































































































r 4 





§ 26. The results of the last group are also represented in 

 fig. 2, in which the resonance curve is drawn. All the obser- 

 vations were made within the range of the adjustable con- 

 denser, the same secondary coil and the same degree of coup- 

 ling being preserved for all. The ordinates are not observed 

 deflections of the hot wire, but are in arbitrary units propor- 

 tional to the intensity of the primary oscillations, derived from 

 the formula 



*, = -J- Vd, 



in which i 1 is the effective value of the harmonic in question, 

 /the frequency, d the hot-wire deflection, and c a constant. 

 The deflections are thus corrected for frequency (§ 20) and for 

 the parabolic form of the hot-wire relation. 



§ 27. Perfect agreement between the different harmonics in 

 each group cannot be expected, as indeed other observers have 

 found that not all the harmonic frequencies in electrical 



