518 2f. I. Pupin — Electrical Oscillations of 



due to resonance. Thinking that perhaps the development of 

 Foucanlt currents in the rather coarse iron wire or that the 

 mechanical vibrations of the wire may in some way or another 

 modify the period of the circuit and the self-induction of the 

 coil, I endeavored to devise some simple experiments which 

 would test the hypothesis just mentioned. 



Experiment 4- — About 19 microfarads were plugged in the 

 condenser and then the iron wire was gradually introduced into 

 the inertia coil and the rise of potential was observed closely. 

 The potential went up continually until a maximum of about 

 100 volts was reached. From this point on the electrometer 

 needle remained stationary although I kept on adding more and 

 more iron wire, one wire at a time. Suddenly the addition of 

 another wire caused the electrometer needle to drop down con- 

 siderably below the midway point between zero and the maxi- 

 mum point. I tried then to bring it back by removing the 

 iron wires slowly one at a time from the inertia coil and found 

 that I had to remove a considerable number of wires (about 

 one fourth of the total number which was about 300 wires) 

 before the electrometer index began to move up rapidly toward 

 the maximum point previously obtained. The wires were now 

 gradually replaced into the coil and the same phenomenon ob- 

 served as before. The experiment was repeated over and over 

 with invariably the same result. 



Experiment 5. — After the collapse of resonance (as described 

 in the last experiment) by the addition of the last fatal wire, 

 as it were the last straw that broke the camel's back, resonance, 

 indicated by the large rise of potential and the singing of the 

 iron wire, was established again by gradually lifting the whole 

 bundle until the electrometer indicated the highest point. 

 Then the bundle was lowered again very gently until it was 

 entirely in the coil again. The electrometer needle would then 

 remain stationary, apparently for an indefinite time. But the 

 mere approach of an iron wire toward the bundle would upset 

 the resonance suddenly and cause the electrometer needle to 

 drop way down. That this approach of the single iron wire 

 did not disturb resonance by the increase of the coefficient of 

 self-induction was proved by the circumstance that if this iron 

 wire was introduced into the coil and then the whole bundle 

 raised until resonance was established and then slowly lowered 

 again the maximum potential previously obtained was reached 

 again and maintained apparently indefinitely although the iron 

 wire bundle contained now that wire also by whose mere 

 approach resonance was upset before. In fact it is possible to 

 increase in this way, the number of iron wires in the bundle 

 by four or five wires without changing perceptibly anything 

 in the resonant effect, whereas the mere approach of a single 



