CONTEMPORARY ADVANCES IN PHYSICS 331 



By increasing the magnetizing field very slowly it is possible to 

 space the peaks in the oscillographic curve, or the clicks in the receiver, 

 so widely that the bigger can be counted. Listening to the separated 

 clicks, van der Pol estimated that the process of magnetizing a cubic 

 centimetre of iron or of an iron-nickel alloy involves several thousand 

 of the jumps. It is also possible to measure the area under each of 

 the larger peaks in a curve obtained with a good oscillograph, and 

 calculate from it the magnetic moment of a magnet, the sudden 

 creation of which within the substance would have resulted in just 

 such a peak. One observed by E. P. T. Tyndall could have come 

 about through the sudden creation of a magnet of moment .0027. 

 The word "creation" must not be taken too literally; it might imply, 

 for instance, that two adjacent magnets were at first pointed contrari- 

 wise to one another, and one of them was suddenly wheeled around 

 by the field, so that they ceased to neutralize each other. Data such 

 as that just cited from Tyndall would then indicate the sizes of the 

 magnets preexisting in the substance; data such as those of van der 

 Pol, their number. Both sets of data show that one cannot identify 

 these magnets with individual atoms; they are too large (the moment 

 .0027 is as great as that of a piece of saturated iron 0.12 mm. on a 

 side) and too few. Neither can they be identified with individual 

 crystals; a piece composed of a single crystal makes as much noise 

 in the receiver, while being magnetized, as a fine-grained sample. 

 The data suggest that ferromagnetic metals are built up out of mag- 

 netic units larger than atoms and smaller than crystals— a suggestion 

 which to the theorists is often extremely acceptable. It is also a wel- 

 come fact, that the peaks and the crackling are associated with the 

 steeply-sloping segments of the magnetization-curve, while the initial 

 and final nearly-horizontal arcs of the curve are smooth and silent.* 



Magnetization of Single Crystals 

 Ferromagnetic crystals large enough to be studied are only just 

 ceasing to be a rarity. Only two sorts occur in Nature: those of 

 magnetite (a modification of one of the oxides of iron, Fe304) and 

 those of pyrrhotine (a sulphide of iron, FeySs). To procure single 

 crystals of a metal or an alloy, it used to be necessary to wait on the 

 hazards of the foundry, out of which there might arise at long intervals 

 a single large uniformly-crystallized lump. This condition prevails 



* Attention must be drawn to the possibility that the peaks in the curve, or the 

 clicks in the sound, are due to fortuitous coincidences of events individually too 

 insignificant to be perceived. Should this turn out to be the case, the Barkhausen 

 effect would resemble the Schroteffect of thermionic emission, and the interpretation 

 of the data would be changed. 



'>9 



