PEIRCE. — MAGNETIZATION IN IRON. 123 



These and other difficulties He in the way of anyone who attempts 

 to use the Isthmus Method in its original form, and every modifica- 

 tion of it, whatcA'er advantages it has, usually introduces some new 

 problems. Notwithstanding all this, the method is a most useful one 

 and has a much wider application than has usually been given it. 

 No other way that has been proposed of making magnetic measure- 

 ments at very high excitations is nearly so good, and the test pieces 

 now employed are small and convenient to make. 



There is still in many cases some uncertainty in the determinaton of 

 H, and Hadfield and Hopkinson discuss the subject in trying to ac- 

 count for the differences between their results^ and those of Gum- 

 lich, obtained at the Reichsanstalt. MoreoAer, there is sometimes 

 irregularity in the values of / measured by the Isthmus Method * for 

 a single specimen. For these reasons, it has seemed to me worth 

 while to push the use of the solenoid for magnetizing test pieces farther 

 than has yet been done, to make sure that there are specimens of 

 metal in which / is higher than 1700 even in much w^eaker fields than 

 those which the Isthmus Method furnishes. This is especially de- 

 sirable if we wish to be able to determine the constitution of a large 

 mas^ of steel by a quick measurement of the specific magnetism of a 

 small test piece in an electromagnet arrangefl for the purpose, as has 

 been proposed. 



According to the molecular theory of magnetization of Weber and 

 Ewing, the molecules, which lie with their magnetic axes in all direc- 

 tions when the metal is in the neutral state, tend to turn in the direc- 

 tion of any magnetic field to which the iron may be exposed, though 

 they are hindered from doing so by the interaction of the molecules 

 themselves. When, however, the applied field is made strong enough 

 to overcome these intermolecular forces, in large measure, all the axes 

 of the elementary magnets point practically in the same direction. 

 It is evident, therefore, that unless the applied field affects the mo- 

 ments of the elementary magnets of which the metal is made up, the 

 magnetic moment (7) of the metal per unit volume should remain 

 nearly constant after the excitation has gone beyond a certain large 

 value. This maximum magnetization is very different in different 

 metals and we may well consider it as characteristic of a material. 



3 Journal of the Institution of Electrical Engineers, Dec, 1910, p. 253. 



4 Ewing's Magnetic Induction in Iron and Other Metals, Tables XI, XII. 



