under the Operation of Feeble Magnetic Forces. 227 



The deflections were read in the manner usual with Thomson's 

 galvanometers, by the motion of a spot of light thrown upon 

 a scale after reflection by the mirror. The division is in 

 millimetres, and with the aid of a lens a displacement of ^ of 

 a division can usually be detected with certainty. 



The direct effect of the magnetizing-spiral upon the sus- 

 pended needle was compensated by a few turns of wire C, 

 7 centim. in diameter, supported upon an adjustable stand D. 

 This adjunct might have been dispensed with ; but what is 

 essential is the larger coil, E, by which the effect of the iron 

 core is compensated. This coil consisted of 74 convolutions, 

 of mean diameter 18 centim., tied closely with string, and 

 mounted upon an independent stand, F. By sliding this 

 stand, and ultimately by use of the screw, G, the action of 

 this coil upon the suspended needle can be adjusted with 

 precision. All the coils are connected in series ; and pro- 

 vided that the magnetic condition of the iron under given 

 force is definite, matters may be so arranged that the imposi- 

 tion of the force produces no movement of the suspended 

 needle, or, more generally, the compensation may be adjusted 

 so as to suit the transition from any one magnetic force to any 

 other. If the susceptibility (k) and permeability //, ( = 4:7rk + 1 ) 

 were constant, as has often been supposed in mathematical 

 writings, the compensation suitable for any one transition 

 would serve also for every other, and the magnetometer-needle 

 would remain undisturbed, whatever changes were permitted 

 in the strength of the magnetizing current*. The question 

 now presenting itself is, How far does this correspond to fact ? 

 or, rather, How far is it true for magnetizing forces which are 

 always very small ? for we know already that, under the ope- 

 ration of moderate forces exceeding (say) 1 or 2 C.Gr.S., not 

 only is /a not constant, but there is no definite relation at all 

 between magnetic induction and magnetizing force, whereby 

 the one can be inferred from the other without a knowledge 

 of the previous history of the iron. 



The magnetizing force of the spiral is of course easily cal- 

 culated. The difference of potential in passing through n 

 convolutions of current C is 4z7rnG. If the n convolutions 

 occupy a length I, the magnetizing force is 



4.0 J, 



or, in the present case, 128 7rO. 



* The idea of compensating the iron is not new. The method was 

 employed by Koosen (Pogg. Ami. Bd. lxxxv. S. 159, 1852) to exhibit the 

 phenomena of "saturation." 



R2 



