CORRECTION FOR MAGNETISATION INDUCED BY THE EARTH. 607 



By using a current with the bifilar and the tangent galvano- 

 meter (858), we may arrange the apparatus so as to make the two 

 observations at the same time. 



Suppose, for example, that the deflecting bar M is placed above 

 the declinometer, at a distance R, and is supported by a bifilar 

 suspension. When this bar is perpendicular to the meridian, the 

 deflection a produced on the declinometer is 



M 



and the couple is determined by the torsion of the bifilar. This 

 arrangement enables us even to regulate with facility the position 

 of the bifilar, for there should be no deflection when the deflecting 

 bar is exactly in the meridian. 



This would also be the case if the bar were placed at the same 

 height as the declinometer, either in the meridian or in a plane 

 perpendicular to the meridian ; and this latter case would corre- 

 spond to almost double deflections. 



1179. CORRECTION FOR THE MAGNETISATION INDUCED BY THE 

 EARTH. When the couple MH is estimated by oscillations, as is 

 usually done, it is necessary to determine experimentally the co- 

 efficient <, which corresponds to induced magnetisation. 



In strictness, this coefficient should be deduced from an oscil- 

 lation experiment; and we might work in the following manner. 

 The bar being supported by a bifilar suspension, and in equilibrium 

 in the magnetic meridian, the oscillations of the system give the 

 couple 



The bar is next reversed end-wise ; the induced magnetisation is 

 then opposed to the magnetic moment, and, if we assume that it has 

 sensibly the same effect, the fresh oscillations give the couple 



C-HM(i-<H) = A 2 . 

 It follows that 



2 HM = A 1 -A 2 , 

 2C 

 and consequently 



