MAGNETIC PERMEABILITY OF IRON, STEEL AND NICKEL 43 



and h that from helix F, the number of lines of force passing through 

 the magnet E, expressed in the unit we have chosen, will he 



(9) 



where ri is the number of coils in the ring G, n the number in the 

 helix F, R the radius of G, 6- 27 the total magnetism of the earth, and 

 7450' the dip. The quantity 2n'(6-27 sin 7450')^E 2 is constant for 

 the coil, and had the value 14* 15. This is the number of square metres 

 of a unit field which, when cut once by a wire from the galvanometer, 

 would produce the same deflection as the coil when turned over. 



The experiments being made by reversing the magnetism of the bars, 

 a rough experiment was made to see whether they had time to change 

 in half a single vibration of the needle; it was found that this varied 

 from sensibly to nearly 1 second, so that there was ample time. It 

 was also proved that the sudden impulse given to the needle by the 

 change of current produced the same deflection as when the change was 

 more gradual, which has also been remarked by Faraday, though he 

 did not use such sudden induced currents. As a test of the method, 

 the horizontal force of the earth's magnetism was determined by means 

 of a vertical coil; it was found to be 1' 634. while the true quantity is 

 1-641. 



It is sometimes assumed that some of the action in a case like the 

 present is due to the direct induction of the helix around the magnet on 

 the coil F. I think that this is not correct; for when the helix is of 

 fine wire closely surrounding the bar or ring, all the lines of force 

 which affect F must pass through the bar, and so no correction should 

 be made. However, the correction is so small that it will hardly affect 



the result. If it were to be made, -^ (equation 5) should be diminished 



CL 



by 47r/lf ; but, for the above reasons, it has not been subtracted. As a 

 test of the whole arrangement, I have obtained the number of lines of 

 force in a very long solenoid: the mean of two solenoids gave me 



Q' = 12-67 M(xR<); 

 while from theory we obtain, by equation (7) (n 1), 



which is within the limits of error in measuring the diameter of the 

 tubes, &c. 



All the rings and bars with which I have experimented have had a 

 circular section. In selecting the iron, care must be used to obtain a 



