Electromotive Forces of the Electric Arc. 



515 



the alternator spindle due to the driving belt. The speed of the 

 alternator was 35,400 revolutions per minute, and the highest fre- 

 quency 120,000 periods per second. To give an idea of how very high 

 this frequency is, it is mentioned that if a frequency of 100 periods per 

 second be represented by 10 inches, a very ordinary scale in plotting 

 curves, then the squared paper that would be required to plot the curve 

 between impedance and frequency for the solid arc, which extends over 

 the range from 250 to 120,000 periods per second, would be 1,000 feet, 

 or about l/5th mile long. 



It was found that the spindle alone of the alternator without the 

 inductor could be driven at 60,000 revolutions per minute, or 1,000 

 revolutions per second. 



A table of high frequency alternators shows that this alternator gives 

 a frequency seven or eight times as high as the highest value previously 

 attained. 



The Thermo-galvanometer. 



The principle of this new instrument consists in causing the current 

 to be measured to flow through a very fine wire, the heat radiated by 

 the wire being measured by a modified Boys' radio-micrometer. The 

 instrument is practically non-inductive, and may be used equally well 

 for direct or alternating currents. The actual instrument used has a 

 resistance of about 18 ohms, and gives a deflection of 500 scale divi- 

 sions at a scale distance of 2000 divisions (1 scale division = l/40th 

 inch) for a current of about 9 x 10 -4 ampere. 



Telephone and microphone currents can be easily measured with this 

 instrument. 



Results Obtained by Varying the Frequency. 



This, the fundamental investigation of this communication, consists 

 in varying the frequency of the superposed alternating testing current 

 to see whether, at a sufficiently high frequency, the conditions of the 

 arc remain constant. The criterion that the conditions of the arc 

 remain unchanged has been shown to be that the power factor, as 

 measured with the superposed alternating current, is unity. Under 

 these circumstances the true resistance will be equal to the impedance. 

 It is experimentally found by sufficiently increasing the frequency, 

 that the power factor approximates asymptotically to + 1, and that 

 for the highest frequencies used, it is + 1 to within the limits of 

 experimental error, therefore at these frequencies the variations of the 

 P.D. and current obey Ohm's law, and the impedance of the arc is 

 equal to its true resistance. 



With solid carbons the power factor at 250 periods per second is 

 - 0*91, on increasing the frequency it decreases numerically until it 

 vanishes at 1950 periods per second, with further increase of frequency 



