414 SECTIONAL COMMUNICATIONS. 



to deduce from experiments on ferromagnetic and paramagnetic sub- 

 stances the exact numerical values of these atomic moments. 

 According to the type of experiment utilised, tliese moments 

 classify themselves among a certain number of groups. In 1909 

 M. Kammerlingh Onnes and I measured the magnetisation to satura- 

 tion for iron and nickel in the neighbourhood of absolute zero. From 

 these first measurements there immediately appears a result of which 

 we shall later on see the generality. 



All atomic moments are integral viultiples of the savie- elementary 

 moment, to which I have given the name ' magneton.' There are eleven 

 magnetons in the iron atom and three in the nickel atom at very low 

 temperatures. Other atomic moments emerge from the study of alloys 

 made of ferromagnetic metals, also at low temperatures. These 

 measurements have given indirectly nine magnetons for cobalt, and 

 have led to the discovery of the alloy FejCo, which is interesting because 

 its molecule possesses thirty-six magnetons — a number greater than the 

 sum of those of the constituent metals. This alloy, which' at ordinary 

 temperature has an intensity of magnetisation some 10 per cent, in 

 excess of that of iron, has been brought into practical use in the con- 

 struction of the pole-pieces of electromagnets. 



Tlie study of ferrom_agnetic substances at temperatures higher than 

 the Curie Point has been a new source of magnetic moments. The 

 magnetic moment of nickel for a temperature interval of about 400° has 

 been determined six times by means of independent series of observa- 

 tions made by different observers. They have found 8.08, 7.99, S.04, 

 8.05, 8.03, and 7.98 magnetons respectively — i.e., numbers in the 

 immediate neighbourhood of the integer 8. It is worthy of notice 

 that above the Curie Point the atomic moment of nickel is different 

 from its value at low temperatures. This possibilty of the same atom 

 assuming different magnetic moments is a general property. 



Investigations with solutions have furnished very many atomic 

 moments. The important researches of Prof. Cabrera and his pupils 

 should be mentioned in particular. It turns out that in the case of 

 the dissolved salts of nickel, for example, there is found the same 

 atomic moment, whatever the concentration may be. This moment 

 proves with great precision to be equal to sixteen magnetons for nickel 

 in the chloride, the sulphate, and the nitrate. The two series of experi- 

 ments, one in Madrid and the other by Mile. Bruins in my laboratory, 

 were absolutely concordant. In other cases the atomic moment calcu- 

 lated by means of the paramagnetism of the solution assumes a definite 

 value only at extreme concentrations, whether very weak or very strong. 

 In the interval the atomic moment apparently varies. This arises from 

 the fact that the moment of the metal in the ion, or in the hydrolised 

 molecule, is not the same as in the undissociated molecule. Thus Fe 

 possesses 27 magnetons in very dilute solutions of PeClg, and tends 

 towards 29 magnetons in very concentrated solutions of this salt. 



The study of the magnetisation of paramagnetic salts in the solid 

 state has brought to light magnetic moments charact-eristic of the 

 type of combination; this also lends support to the theory of the 

 magneton. 



