NO. II STRUCTURE OF THE ATOM PARSON 9 



mentioned, but leaves the superstructure of his theory almost intact, 

 is a strong argument in favor of the magneton. That this substi- 

 tution can be made will be made clear by a short quotation from the . 

 conclusion of his paper : 



.... and we can form a simple and exact picture of all the facts of 

 magnetism and of diamagnetism by imagining the individual currents pro- 

 duced by the electrons to be indeformable but movable circuits of no resistance 

 and very great self-induction, to which all the ordinary laws of induction are 

 applicable. 



The substitution I have suggested has further advantages : it 

 makes a great advance upon Langevin's theory, owing to the fact 

 that, whereas the reaction of one of Langevin's orbits to its environ- 

 ment must vary with the phase of the motion of its electron, each 

 magneton has the properties of an ordinary current circuit at every 

 instant, and it is no longer necessary to think of the orbits statis- 

 tically either in respect to their number or in respect to time. The 

 importance of this difference is easily shown. I will first give another 

 quotation from Langevin. 



After showing that a single one of his orbits can have a moment as 

 great as that of the oxygen or iron atom, he says (loc. tit., p. 122) : 



Since the individual currents due to the other electrons present in the 

 molecule neutralize one another just as in a purely diamagnetic body, it 

 follows that, in magnetic molecules, one or more electrons are sharply 

 separated from the rest and are alone responsible for the magnetic properties, 

 while all the electrons co-operate to produce diamagnetism. 



These are perhaps the very same electrons, situated in the outer part of the 

 system forming the molecule, that play a part in chemical actions, where we 

 know that electrons equal in number to the valence come into action. That 

 would account for the profound influence of the state of molecular association, 

 physical or chemical, upon paramagnetism, and its virtual lack of effect upon 

 diamagnetism. 



It is remarkable how completely the present theory, by means of 

 the magnetic forces between magnetons, realizes in a quite definite 

 manner the state of affairs here hinted at by Langevin. 1 It should be 

 observed, however, that he does not specify that the chemical forces 

 due to his electrons are magnetic in nature. This is very probably 



langevin's deduction that the magnetism of the oxygen or iron atom 

 must be due to a few sharply distinct orbits is perhaps not altogether valid 

 on his theory: a rotation of the whole of an otherwise diamagnetic system 

 of electrons, whether moving in individual orbits or in chaotic motion, could 

 give the same result. It may also be pointed out that if the orbits containing 

 the few valence electrons were distinct, as Langevin suggested, the radiation 

 from them could not possibly be reduced to almost zero by compensations, on 

 account of their small number. 



