of Magnetization in a Magnetized Iron Cylinder. 65 



the bar had quite run down. The experiment was tried, and 

 always gave the expected result. 



If, then, the exact time taken for pole changing could be 

 measured, when the galvanometer just showed no sign of an 

 induced current, this period of time would be that of the fall 

 of magnetization of the bar. In practice such a change is 

 just as well performed by having an arrangement, shown in 

 the diagram, where an inductor-coil is suddenly connected 

 to the galvanometer instead of the primary circuit, and the 

 time is very gradually increased until no deflexion is produced. 



A B is a long primary coil of known dimensions ; within 

 this the iron cylinder to be tested is placed ; it is in circuit 

 with a battery C, and a contact-piece E. I is an inductor- 

 coil of known induction area, in circuit with the galvano- 

 meter G, the inductor H, and the contact-piece F. 



The inductor H forms part of a system, consisting of a long 

 helix MK, a battery N, an ampere-balance L, and a key 0; 

 by means of this system the value of the throw or deflexion of 

 the galvanometer is determined in C.G.S. units, by the 

 method of Sir W. Thomson. 



The contact-pieces E and F differ from each other in their 

 action. The circuit at E is broken when E is hit by a pro- 

 jection, D, attached to the carriage of the chronograph moving 

 in the direction of the arrow, whereas the induction circuit 

 is closed when D hits F. E is fixed, and F can be placed at 

 any required distance from it. The distance E F determines 

 the time which elapses between breaking the primary circuit 

 and making the induction circuit. The primary circuit is 

 always broken before the inductor-coil is connected to the 



Phil. Mag. S. 5. Vol. 31. No. 188. Jan. 1891. F 



