October 6, 1892] 



NATURE 



553 



nelisni and magnetizing force, and pointed out that in special 

 cases the three stages became extraordinarily distinct. Curves 

 of I he same kind were used to show what happened after a 

 magnetizing force had been applied, if it were withdrawn or 

 varied in any way. The magnetism in all cases tended to lag 

 behind when the magnetizing force was varied, and hence these 

 curves in any cyclic process became loops enclosing a certain 

 area. It had been proved that this area served to measure the 

 energy expended in carrying the substance through a cyclic 

 magnetizing process, the reason why energy had to be spent 

 being the tendency which the magnetism always had to lag behind 

 the force that was operating to change it. To this tendency he 

 had given the name "hysteresis," a term which was already of 

 formidable significance in the ears of practical electricians. For 

 the existence of hysteresis was the chief reason why the trans- 

 fdrmers which were used in alternate current systems of electrical 

 distribution absorbed wastefully a considerable amount of power. 

 The iron core of a transformer was being carried through a j 

 cycle of magnetization from a positive to a negative value and 



Weber postulated an arbitrary directing force, which tended to 

 hold them in their original direction. The lecturer proceeded 

 to show by means of experiments conducted on the projecting 

 table of the lantern, and shown on a large scale on the screen, 

 that no arbitrary directing force was necessary. The mutual 

 actions of the molecular magnets on one another supplied all 

 the control that was required. It accounted completely for the 

 three stages of the magnetizing process and for all the pheno- 

 mena of hysteresis. It accounted also for the effects which 

 were found to be produced by mechanical vibration and mechani- 

 cal strain. Experiments were made exhibiting the breaking up 

 of molecular groups, bound together by their mutual forces, 

 under the influence of a gradually increased external directing 

 force. In these experiments models were used, consisting of a 

 number of small magnets, pivoted like compass needles on fixed 

 centres, and arranged on the horizontal table of a large pro- 

 jecting lantern. A pair of coils placed one on either side of 

 the group supplied deflecting force, and as the current in these 

 was gradually increased the three stages of the magnetizing 



.—Pro. Ewing's Mognetic Curve Tracer. General viev,- of apparatus. 



Kack again some 80 or 100 or 120 times a second, and every one 

 of these periodic reversals of magnetism implied a waste of 

 energy, which went on even when no useful current was being 

 drawn off. It was a question of considerable practical interest, 

 whether the amount of work wasted on the iron of a transformer 

 was the same per cycle at high speeds of reversal such as were 

 usual in practice, as it was in the slow speed laboratory experi- 

 ments, by the help of which these cyclic curves had been drawn. 

 The lecturer proceeded to give some account of molecular 

 theories which had been framed to account for the character- 

 istics of the magnetizing process. It was suggested, originally 

 by Weber, that the molecules of iron are always little magnets, 

 and that when the iron, as a whole, is not magnetized, it is 

 because as many of the molecular magnets are facing one way 

 as another. According to this view the process of magnetization 

 consists in turning the molecular magnets round, so that they 

 face, more or less, one way. When a very strong magnetising 

 force i? applied, the molecules are forced all to face one way ; 

 the piece is then saturated. To explain why they did not at once 

 turn round completely when any magnetizing force was applied, 



NO. II 97, VOL. 46] 



' process and the phenomena of hystersis exhibited themselves in 

 the manner in which rearrangement of the elementary magnets 

 composing the group took place. In some of the models the 

 magnets turned under water, so that their vibrations were rapidly 

 damped out. Slides were also shown which gave some of the 

 results of observations recently made in the lecturer's laboratory 

 i by Miss Klaassen, of Newnham College, which demonstrated 

 1 an extraordinarily close agreement between the phenomena 

 i noticed in the magnetization of actual iron and those presented 

 ! by a model consisting of groups of little pivoted magnets. 

 I Even the less conspicuous features of the actual process were 

 reproduced in the model with a fidelity which went far to con- 

 firm this molecular theory of magnetism. It was shown, for 

 instance, that the model reproduces a phenomenon familiar in 

 real iron, namely, the tendency which magnetic changes exhibit 

 to be imperfectly cyclic, under cyclic changes of magnetic force, 

 until these are repeated several times, and also that in the 

 model, just as in real iron, this tendency disappears if a process 

 i of demagnetizing by reversals of gradually diminishing magnetic 

 I force has been previously gone through. 



