ir. G. Cadi/ — Electric Arc between Metallic Electrodes. 243 



circuit to act inductively on a second coil in fairly close coup- 

 ling with it. If the second coil is in series with a small lamp 

 and a variable condenser, then as the capacity is slowly varied 

 the lamp can sometimes be seen to light up three or four times, 

 as successive harmonics are passed. The lamp sometimes glows 

 more brightly than a similar one in the discharge circuit. 



(A) If the coupling in the last experiment is close, then on 

 closing the resonating circuit it is found that the direct current 

 from the generator is reduced, showing that the secondary 

 circuit has reacted on the primary. 



£ 19. Energy and efficiency of the oscillations. — The follow- 

 ing experiment was performed in order to discover how large 

 a percentage of the direct current energy expended in the 

 discharge could be converted into the energy of electrical 

 oscillations. Two -±- candle power 110 volt incandescent lamps 

 of about 700 ohms resistance each were connected in the dis- 

 charge circuit, one on each side of the tube. They absorbed 

 nearly all the energy of the oscillations, and glowed more 

 brightly when the discharge passed than when the electrodes 

 Avere touching. 



The total power absorbed by the lamps is given in the first 

 column of Table II. It was estimated from the brightness of 



Table II. 



Watts in lamps Watts supplied Per cent Generator 



Total d. c. a. c. to arc efficiency current 



13-8 5-4 8-1 18-0 46 '06 amp. 



19-6 7'4 12-2 21*2 57 '072 " 



25-0 9-0 16'0 24-2 66 '082 " 



36-0 20-4 15-6 345 45 '125 " 



the lamps. In the second column are recorded the watts 

 expended in the lamps by the direct-current component, 

 obtained by multiplying the square of the generator current 

 (last col.) by the lamp resistance. The difference between the 

 first and second columns gives the watts of alternating current 

 power consumed by the lamps. The watts supplied by the 

 generator to the discharge (neglecting the external resistance) 

 are given in the fourth column, while the fifth column shows 

 the efficiencies, obtained by dividing col. 3 by col. 4. It 

 will be observed that the highest efficiency recorded is 66 

 per cent. 



g 20. Deter ini notion of the frequency. — The following 

 resonance method was used in determining the frequency of 

 the oscillations. In the main discharge circuit, about 30-50 cm 

 from the tube, was a coil of wire which in most of the experi- 

 ments consisted of six turns, having a diameter of 22 cm . Its 



