312 BELL SYSTEM TECHNICAL JOURNAL 



the fact that an iron wire, which undergoes the experience of being 

 violently pulled and then relaxed, displays very different I-vs.-H 

 curves before and after this adventure. In certain cases we do 

 observe some sort of an attendant change, as for instance when the 

 iron wire has been stretched so forcefully that it is permanently 

 lengthened, or cold-rolled so vigorously that the X-ray diffraction- 

 pattern due to its little crystals is affected. In other cases we observe 

 no concurrent change whatever, and are forced to assume that there 

 has been an internal alteration of the metal, for which there is no 

 evidence beyond the testimony of the changed magnetization-curve. 

 In the same way, we are prone to assume that when "the burnt child 

 dreads the fire," something is altered within his brain-cells, for which 

 there is no evidence except his change of conduct. As a rule, one 

 would not speak of the brain-cells; one would say that the child has a 

 memory of the painful burn. The ferromagnetic substance also 

 changes its conduct after each experience, as though it remembered. 

 No one, I presume, supposes that it actually has a consciousness which 

 remembers; but the actual responsible alteration, whatever it may 

 be, is often as far beyond detection as the alteration in the brain-cells. 

 The resulting change in conduct, the result of this "memory" of the 

 metal, is what is known as hysteresis. 



All this makes the designing of a model for a ferromagnetic substance 

 a very difficult and perplexing problem indeed, as we shall discover 

 in due time. For the moment we are concerned only with knowing 

 how much of the biography of a piece of metal must be recorded, 

 in order to give background and value to a determination of its I-vs.-H 

 curve. A curve inscribed "This is the I-vs.-H curve for iron'' would 

 not be worth much, no matter how carefully it had been determined 

 nor how nearly pure the iron had been. At this point the physicist 

 must betake himself to the foundry and the rolling-mill, and confer 

 with the metallurgist, and learn the usage of a number of uncouth 

 words such as swaging and sintering and cold-working and quenching, 

 and grasp the distinction between cast-iron and wrought-iron and 

 pig-iron and soft steel and hard steel, and observe a number of processes 

 which were discovered so long ago that originally they were practiced 

 without the least assistance from the guiding hand of pure science. 

 The curve for his sample of metal must be labelled with the processes 

 which the sample underwent before and after it came into his hands. 

 Even yet it is not completely settled how many of the details of these 

 processes should be recorded, nor how far back the history of the 

 sample should be traced. One piece of knowledge, however, dispenses 

 us from the risks of this uncertainty; it is known that a long-continued 



