1896-] Professor Ewing on Hysteresis, 227 



WEEKLY EVENING MEETING, 



Friday, May 22, 1896. 



George Matthey, Esq. F.R.S. Vice-President, in the Chair. 



Professor J. A. Ewing, M.A. F.R.S. Professor of Mechanism and 

 Applied Mechanics in the University of Cambridge. 



Hysteresis, 



(Abstract.) 



The lecturer explained that the word hysteresis was not a term in 

 neuro-pathology. It had nothing to do with hysterics. The name 

 might be unfamiliar, but the thing it described was exceedingly 

 common. It was scarcely too much to say that hysteresis was to be 

 found everywhere, except, perhaps, in the dictionary. 



The word was derived from the verb vcrrepeo), which signified to 

 lag behind. It was introduced about fourteen years ago to name a 

 characteristic which had been prominent in several researches into 

 the physical qualities of certain materials, especially of iron. The 

 name was invented at a time when the phenomenon of hysteresis had 

 no more than a purely scientific interest ; but in the rapid advance 

 of industrial electricity hysteresis had become a matter of much 

 commercial importance, and the word was now in common use by 

 electrical engineers. Certain materials, when causes acted on them 

 tending to change their physical state, had a tendency to persist in 

 their previous state. This tendency to persist was what constituted 

 hysteresis. 



It was in connection with the magnetic properties of iron and 

 steel that the most conspicuous and practically the most important 

 manifestations of hysteresis were found. An experiment was shown 

 to illustrate hysteresis in the changes of magnetic condition brought 

 about by the application and removal of stress. An iron wire, 

 magnetised by a constant current in a surrounding coil, was hung up 

 and loaded with weights. The weights were alternately removed 

 and reapplied, and the magnetic state of the wire was shown by 

 means of a mirror magnetometer. It was seen that when the weights 

 were repeatedly put on and ofi", the magnetism changed from one to 

 another of two values ; but when half the weight only was left on 

 during unloading, the magnetism assumed a value much nearer to 

 the loaded than to the unloaded state ; whereas when half the weight 

 was put on after unloading, the magnetism took a value nearer the 

 unloaded than the loaded state. In other words, the magnetic efiects 

 of the loading lagged behind the changes in the loading itself. This 



q2 



