484 Pupin — Resonance Analysis of Alternating Currents. 



wires were fairly well insulated from each other. A layer of fine 

 copper wire surrounding this coil formed part of the resonator 

 circuit. First, the secondary current was passed through the coil 

 before the iron core was inserted. The resonator could detect 

 no harmonic worth mentioning even when the current was 

 increased almost to full load. But as soon as the iron core 

 was introduced the odd harmonics appeared, especially the 

 third harmonic ; its strength increased proportionally to the 

 current. Placing now another similar iron core on the top of 

 the first and adjusting it in such a way that it allowed a small 

 rocking motion the two cores could be set into violent vibra- 

 tion by the inductive attraction between them. This vibration 

 manifested itself by a very loud note corresponding in pitch to 

 the frequency of the alternator. The vibration could be 

 diminished very much by pressing the top core against the 

 lower core and against the table. The vibration produced no 

 appreciable difference in the strength of the harmonic ; if 

 anything it seemed to make it stronger. Mechanical vibration 

 produced by striking the cores produced no appreciable change 

 in the harmonic. These experiments seem to me to render 

 the theory which ascribes the origin of harmonics to the hys- 

 teretic action of iron completely untenable. 



I do not think that the proper time has arrived yet for the 

 formulation of a physical theory which will give a complete 

 account of the peculiar behavior of iron, by means of which 

 it superposes odd harmonics upon the wave of a simple har- 

 monic current. The view which irresistibly suggests itself to 

 my mind is simply this : Upper harmonics will be generated 

 whenever more or less abrupt changes of the magnetic state in 

 any part of the magnetic field through which an alternating 

 current flows occur. A slotted core armature or an armature 

 made up of coils with iron cores distributed over a drum com- 

 mon to all of them will introduce such changes. An alternat- 

 ing current induction motor, especially when it is not of a smooth 

 core armature type, will also cause abrupt changes of magnetism 

 and hence cause strong deviations of the feeding current from 

 the simple harmonic form. But if this view be correct, then 

 every complete cycle of magnetization to which iron is sub- 

 jected when under the inductive action of a simple harmonic 

 current must be accompanied by some abrupt changes in mag- 

 netism, and that, too, whether the mean magnetic intensity of 

 the cycle be large or small. One thing seems certain and that 

 is, that hysteresis, as commonly understood, will not account 

 for these abrupt cyclic changes ; for, if they really exist and 

 are the cause of harmonics, they are certainly not affected by 

 mechanical vibrations by which, as is well known, all hysteretic 

 effects are influenced very much. But whatever the real 



