202 Prof. E. Bouty's Studies on Magnetism. 



merely say that the analogy on which it rests may be carried 

 very far, and that we could find arguments in favour of it even 

 in our present researches : thus we have seen that repetition of 

 the action of a current upon a needle augments the permanent 

 magnetic moment communicated to the latter * ; in like manner, 

 when an imperfectly elastic thread is twisted by means of a con- 

 stant force, the permanent torsion it acquires is increased by a 

 second or a third application of the same force, and tends towards 

 a new limit. 



We do not charge Wiedemann's hypothesis with being abso- 

 lutely false ; we only say that it is incomplete ; and this is why 

 a certain number of phenomena can only with extreme difficulty 

 be accommodated to it. Of this number are some relative to the 

 superposition of a certain permanent magnetism and a tempo- 

 rary magnetism opposite in direction, particularly those pro- 

 duced in the separation of magnetized bundles f. 



Instead of attributing, as Wiedemann does, temporary and 

 permanent magnetism to the same molecules, other physicists J 

 prefer to assume that the condition (whatever it may be) which 

 corresponds to the conservation of a certain degree of permanent 

 magnetism is communicated, in the steeling or the tempering, 

 only to a certain number of molecules, the rest retaining the 

 magnetic properties of soft iron. It is further conceivable that 

 the different molecules may acquire the coercive power in differ- 

 ent degrees. 



Be this as it may, I think that important light would be shed 

 on many facts in connexion with the magnetization of steel by 

 supposing that it is magnetically heterogeneous. Now this is 

 not a gratuitous hypothesis. Chemists will readily admit that 

 the true nature of the chemical species which constitute steel 

 is still very imperfectly known, and that a bar of steel may con- 

 tain normally a mixture of several of those species. When a 

 steel bar is attacked by chlorhydric acid, it is ascertained, accord- 

 ing to A. Holtz §, that the corroded surface is rugous, the aspe- 

 rities being formed of a highly carbonaceous steely substance 

 (Kohleneisen) scarcely or not at all capable of being attacked by 

 the acid ; and the arrangement of this substance in the interior 



* Vide supra, Part II. sect. 2 (pp. 90-98). 



f When an imperfectly elastic thread has undergone a permanent tor- 

 sion, a temporary torsion in the opposite direction can, it is true, be super- 

 posed to the permanent one by the application of a force insufficient to 

 untwist it entirely. But how can a temporary torsion be associated with 

 a permanent one in the opposite direction in a thread submitted only once 

 to torsion and then left to itself ? But this is what happens in magneti- 

 zation (see above, p. 197 et seq.). 



X See Verdet, Conferences de Physique faites a VEcole normale, p. 219. 



§ Pogg. Ann. vol. cli. 1874. 



