Dr. E. Bouty on the Magnetization of Steel by Currents. 199 



to vary, the greater the value of B the more quickly does the 

 distribution-curve approach the axis of the abscissae, for one 

 and the same value of the extreme ordinate. When B in- 

 creases, the magnetic distribution shortens ; it lengthens when 

 B diminishes. Therefore the temporary magnetization is 

 shorter than the permanent*. 



With M. Jamin, we may call A the coefficient of capacity, 



=d the coefficient of conductivity, of the steel under investiga- 

 tion. The measure of the distance from the poles to the neigh- 

 bouring extremity of a long bar furnishes the absolute value 

 of its coefficient of conductivity under the conditions of the 

 magnetization. The temporary conductivity is invariable, 

 whatever may be the intensity of the magnetizing force ; the 

 permanent conductivity is so only in the case of steel not tem- 

 pered!. 



2. After investigating the distribution in a bar innocent of 

 prior magnetization, I sought to apply the same method to 

 bars already magnetized, submitted a second time to the 

 magnetizing action. I selected the simplest case — that of 

 bars not tempered, primitively saturated. 



Action of a direct current. — A second application of a direct 

 magnetizing force does not alter the permanent moment of a 

 saturated bar; but, however feeble this force may be, while 

 submitted to its action the bar acquires a temporary moment 

 superior to its permanent moment. 



Moreover it is found to be impossible to represent these 

 temporary moments by a single formula of the form of equa- 

 tion (1) ; on the other hand, one succeeds very well on taking- 

 the two-term formula 



A / 2 e 2 — e 2 \ A / 2 ^2 _ e 2 \ 



x e * +e * x e* +e 2 ' 



* No constant force, therefore, can produce upon a bar a distribution of 

 the same form as the permanent distribution. This result is in opposition 

 to 1he theory of the coercive force. 



t The number of functions or constants necessary for the complete 

 magnetical definition of a given steel is, as we see, pretty considerable. 

 The two magnetism-functions A 1 and A 2 (or at least the characteristic 

 constants of the rapid magnetization) must be known, and the two conduc- 

 tivity-functions B x and B 2 . Again, we thus define the state of the steel 

 as it is -, and the entire examination has to be gone through afresh for each 

 degree of tempering. 



