MOLECULAR HYPOTHESIS 201 



some of the groups become unstable, and for those groups we pass 

 to the second stage, represented in Fig. 148. As the temperature 

 rises, more and more groups pass to this stage, and between 700 

 and 800 all the groups have become unstable and have then come 

 more or less into line with the magnetising force. This is repre- 

 sented by the great rise in permeability. The sudden disappearance 

 of magnetic quality at the critical temperature, about 785, may 

 perhaps be explained by supposing that after that temperature 

 the vibration has become so violent as to change into molecular 

 rotation. 



With strong fields the iron is already in the second or third 

 stage. Vibration due to heating has two opposite and neutralising 

 effects. On the one hand it eases the breaking up of groups and 

 alignment with the force and so tends to increase the magnetisa- 

 tion, while on the other it tends to reduce the magnetisation by 

 the movements out of the line of the force in the way illustrated 

 in Fig. 156. We may suppose that these opposite effects 

 keep the curve of permeability nearly level, till at a temperature 

 near 800 the diminution due to vibration overpowers the increase 

 due to facility of group breaking and the permeability decreases. 



The change in length of a bar on magnetisation is yet to be 

 explained on the molecular hypothesis. It would obviously require 

 some further assumptions as to molecular grouping or as to 

 molecular dimensions in different directions. 



Attempts to explain the constitution of the 

 molecular magnets. \Ve have already pointed out that the 

 molecular hypothesis only carries down the magnetisation from 

 the visible mass of the bar as a whole to its molecular constituents, 

 and does not account for their magnetisation. Attempts have 

 been made to explain the magnetisation of the molecules in terms 

 of electricity. The earliest was Ampere's hypothesis. He started 

 from the- experimental fact that a current flowing in a circular 

 wire forms a magnet of which the axis is perpendicular to the 

 plane of the wire. He supposed that there was a channel round a 

 magnetic molecule through which an electric current could flow, 

 and he further supposed that this channel was perfectly conducting, 

 so that the energy of the current was not dissipated, and therefore, 

 if once started, the current should persist unless an equal and 

 opposite current were in any way superposed. But a continuous 

 closed current, such as that flowing in a circular wire, is essentially 

 a locus into which electric lines of force are continually converging 

 and are there dissipated. As these lines sweep through the 

 surrounding space they are accompanied by magnetic action, and 

 >o in a sense produce the magnetic field. We know of no such 

 current as that which Ampere supposed. 



Hut if we could imagine a bundle of lines of magnetic force in 

 the form of those of a small magnet, bound round as it were with 

 a perfectly conducting channel, then no lines of force could be 



