Dr. E. Bouty on the Magnetization of Steel by Currents. 195 



had been sufficiently high in my experiments. The value of 

 F corresponding to the minimum of r is much lower in the 

 case of soft tempering ; hence the facility with which its ex- 

 istence is recognized in the latter case. 



IV. Experiments on Cylindrical Bars. 



The preceding experiments were directed so as to permit the 

 investigation of the magnetizing function and the direct de- 

 termination of the situation of the poles ; the object of the 

 present has been to supply the characteristic constants of the 

 distribution of the magnetism in very various circumstances. 

 It will be seen that both problems depend, in the last analysis, 

 on the determination of the same physical elements. 



1. I have already established* that the magnetic moments y 

 of thin cylindrical needles, tempered hard and magnetized to 

 saturation, are represented by the formula 



(9 J 3 2 — e ^2\ p, 



• c -!^r^) with/3= «' • • (1) 



2 eTa — e % ; i ^ £ 



y 



" e p *+e 



first given by Biot, and connected by Green with the theory 

 of coercive force. I have also shown f that the same formula 

 applies to needles derived by breaking from the centre of 

 needles not saturated. At the same period some incomplete 

 experiments J had caused me to suppose that the formula was 

 not applicable to bigger needles ; and this induced me to 

 resume, in the case of the cylindrical bars of 6-10 millims. 

 diameter §, a series of analogous researches, the results of 

 which I now publish. 



The measurement of the magnetic moments was effected by 

 the method applied in the second and third sections of the 

 present memoir. The necessity of sometimes operating on 

 bars of considerable length has demanded some special arrange- 

 ments, which I will briefly indicate. The coil employed to 

 produce the magnetization consists of a zinc tube 1*2 metre 

 long and 5 centims. in diameter, carrying a kilometre of insu- 

 lated copper wire wound in seven superposed layers. The 

 winding is arranged so that either the whole or only a portion 

 of the coil can, at will, be employed. A horizontal railway, 

 in wood, movable about a vertical axis which passes through 

 the prolongation of the wire of the galvanometer-needle || , is 



* Annales de VEcole Normale Superieure, t. iii. p. 34. 



t Ibid. p. 43. % Ibid. p. 44. 



§ French steel, wire-drawn, in rods of two metres length, from MM. 

 Peugeot and Jackson. These steels are extremely homogeneous. 



|| The needle is rendered nearly astatic with the help of a compensating 

 bar : see p. 132. 



02 



