6i6 



NA TURE 



[November 4, 1922 



The Origin of Magnetism. 



AX/HEN the proposal was first made to hold in 

 * * Section A of the British Association at Hull 

 this year a discussion on " The Origin of Magnetism," 

 it was met with the criticism from eminent quarters 

 that the time was not yet ripe for the consideration 

 of this subject. Those who attended the meeting 

 will probably agree that this view was justified, for 

 it can scarcely be said that the position was advanced 

 appreciably, or that any real, or even plausible, 

 answer' was given to the main question involved. 

 Perhaps this was in some measure due to the regret- 

 table absence of Prof. Langevin, who had promised 

 to make the opening remarks, and had expressed his 

 intention of using the opportunity for a critical survey 

 of the whole subject. But a recurrence of the ill- 

 health from which he has intermittently suffered for 

 a long time deprived the Section of Prof. Langevin 's 

 presence and his eagerly anticipated contribution to 

 the discussion. As it was, the discussion lacked co- 

 ordination ; the remarks of the various speakers bore 

 little relation to one another. There was the exposi- 

 tion by Prof. Weiss of his theory of the molecular 

 field and the existence of magnetons ; then Sir J. A. 

 Ewing's description of his new molecular magnet 

 models ; then the remarks of Dr. A. E. Oxley on the 

 changes of susceptibility imparted to platinum and 

 palladium by the occlusion of hydrogen ; and, finally, 

 an account by Mr. L. F. Bates of the measurements 

 of the Richardson effect recently carried out by Dr. 

 Chattock and himself, — all contributions of consider- 

 able individual interest, but not closely related to 

 one another nor providing an answer to the essential 

 question of the origin of magnetism. 



In spite of the comparative failure of the discussion 

 in its wider aspects, one felt that the time had not 

 been wasted, principally because it afforded an oppor- 

 tunity for Prof. Weiss to give a most interesting 

 account of his work in connexion with ferromagnetism 

 and paramagnetism, which is not too well known in 

 this country. Prof. Weiss at very short notice under- 

 took to open the discussion in place of Prof. Langevin, 

 and a fairly complete account of his remarks will 

 eventually appear in the Report of the Association. 

 An outline of this exposition may be profitable here. 



Starting from the analogy of the difference between 

 the laws of fluid compressibility for low and high 

 densities, Prof. Weiss snowed how Langevin's kinetic 

 theory of paramagnetic substances may be modified 

 so as to include strong magnetism — or ferromagnetism 

 — by the assumption of the existence of a molecular 

 field analogous to van der Waal's internal pressure in 

 fluids. A whole array of experimental facts was 

 brought forward in support of this theory of the 

 molecular field. It provides an explanation of the 

 variation of magnetic saturation with temperature ; 

 it accounts precisely for the transformation of ferro- 

 magnetism to paramagnetism at the temperature of 

 the Curie point, and for the observed law of this 

 paramagnetism. The theory also points to a dis- 

 continuity of specific heat at the Curie point, and the 

 magnitude of the discontinuity, calculated from mag- 

 netic data, agrees with calorimetric measurements. 

 Still more interesting is the recently discovered mag- 

 neto-caloric phenomenon, which consists of a re- 

 versible temperature variation accompanying mag- 

 netisation. This differs from the ordinary hysteresis 

 effect, which is irreversible, and always involves 

 heating. In the reversible effect, magnetisation pro- 

 duces a rise of temperature and demagnetisation a 

 fall. At the Curie point the change is by no means 

 negligible, reaching, as it does, a value of about i° 



NO. 2766, VOL. I IO] 



H„ = 



in fields readily attainable. The extent of temperature 

 variation calculated by means of the molecular field 

 theory agrees with that observed. 



When one comes to calculate from various ex- 

 perimental data the numerical value of the mole- 

 cular field, it proves to be of the order of magnitude 

 10' gauss, which is far in excess of the magnetic 

 field which might in the most favourable circum- 

 stances be produced by the magnetic moments of the 

 molecules of a ferromagnetic body, namely, io 4 gauss. 

 This remarkable result indicates that the so-called 

 molecular field has not itself a magnetic origin. In 

 this connexion Prof. Weiss's own (translated) words 

 are worth quoting : — 



" It is therefore impossible for the mutual actions 

 represented by the molecular field to be of a magnetic 

 nature. It is just a notation for forces of a non- 

 magnetic character, with a symbol borrowed from 

 magnetism. I prefer, in place of the primitive de- 

 finition given earlier, the equivalent definition 

 5U 

 51' 



where U is the intrinsic energy per unit volume, and 

 I the intensity of magnetisation. This definition is 

 advantageous in that it does not prejudge the nature 

 of the forces. ... It does not appear to be impossible 

 that the forces may be electrostatic ; that, however, 

 is at present a pure supposition." 



In the second part of his address Prof. Weiss 

 directed attention to another important aspect of the 

 combined kinetic theory of Langevin and his own 

 theory of the molecular field. The possession of these 

 theories permits the calculation of the values of the 

 molecular or atomic magnetic moments which have 

 been the underlying assumption in all theories of 

 magnetism. A great number of atomic moments 

 have thus been evaluated from many experimental 

 sources, such as the measurement of the magnetisa- 

 tion of ferromagnetic substances and their alloys both 

 in the neighbourhood of absolute zero and above the 

 Curie point, the investigation of the paramagnetism 

 of solutions of salts, and the like. The general law 

 which emerges is that " all atomic moments are in- 

 tegral multiples of the same elementary moment, to 

 which the name magneton has been given." For 

 example, six different and independent observers have 

 found for nickel, over a temperature interval of about 

 400°, 8-03, 7-99, 8-04, 8-05, 8-03 and 7-98 magnetons re- 

 spectively, numbers which, it will be seen, are in the 

 immediate neighbourhood of the integer 8. It is, 

 besides, a general property of atoms to possess dif- 

 ferent integral numbers of magnetons according to 

 various conditions, such as their state of chemical 

 combination, or their temperature, whether in the 

 ion, or in the undissociated molecule. Prof. Weiss 

 affirms that the magneton is a real entity, and he 

 pointed to the fact that the Rutherford-Bohr atom, 

 together with Planck's quantum theory, actually does 

 indicate the existence of a universal elementary 

 magnetic moment, which, however, proves upon cal- 

 culation to be almost exactly five times as great as 

 the magneton. 



Prof. Weiss's general conclusions may be summed 

 up by quoting him again : — 



" 1. One of the essential conditions for the pro- 

 duction of strong magnetism — or ferromagnetism— 

 is the existence, between molecules possessing mag- 

 netic moments, of important mutual actions which 

 are numerically expressed by the molecular field, and 

 are certainly of a non-magnetic nature. 



