Comparison of Coefficients of Induction. 97 



the apparatus would fix a limit far before any such speed was 

 reached. It does not, however, appear to be beyond the mark 

 to say, that the sensitiveness could be in this method made 

 fifty times as large as that of the ordinary mode of experi- 

 menting. 



The conclusion expressed in equation (3) has been experi- 

 mentally verified by comparing <£ and 6 with the following 

 arrangement: — A single Leclanche cell sends a current through 

 the primary of a certain pair of coils. The secondary is con- 

 nected up to the galvanometer; and the induced current at 

 breaking is used to deflect the needle. 



To obtain a constant succession of transient secondary cur- 

 rents in the same direction, the apparatus shown in the sketch 

 is used: — A is a cylinder of hard dry wood 5J inches in dia- 

 meter, which rotates on its axis he in the direction of tho 

 arrow. On the surface of A are fixed the brass strips B, C, 

 five in each row : the pieces B are in metallic connexion with 

 each other and also with b ; similarly the pieces C are con- 



nected to the other end of the axis c. F, G are springs keep- 

 ing an electrical contact with b c respectively whilst the axis 

 rotates. D, E are springs pressing firmly on the cylinder, and 

 making contact with the brass strips when they pass under- 

 neath. F, D are inserted in the primaiy circuit; E, Gin the 

 secondary. Thus the secondary circuit is only closed between 

 each break and make of the primary ; and only the induced 

 current at break flows in it. 



It is found that when the cylinder is rotated, a steady de- 

 flection of the needle is obtained, which increases with the 

 speed of rotation, as equation (2) leads us to expect. 



