and on the Ball-ended Magnets of Robison. 29 



The magnetometer is suspended by a fibre of unspun silk 10 to 

 15 cm. in length from a small stand furnished with a simple torsion 

 head. 



| 3. By placing the magnet in a small coil and noting how 

 the deflexion of a mirror magnetometer varied with the current 

 flowing in the coil, I found that up to a field of 4-10 gauss the 

 change of magnetic moment due to induced magnetisation was 

 proportional to the applied magnetic force, and that it amounted 

 to I'll per cent, per gauss. I also found that up to 4*10 gauss 

 the magnetic moment was not permanently changed by the appli- 

 cation of the magnetic force. Thus the change of magnetic 

 moment due to the earth's horizontal force (18 gauss) is "20 per 

 cent. These results agree roughly with those obtained by Mr T. 

 Gray (A. Gray, Absolute Measurements in El. and Mag., vol. II. 

 part I. p. 89). 



Thus, in comparing a field of about '9 (= 5 x '18) gauss with the 

 earth's field, the error due to neglecting the induced magnetisation 

 only amounts to 4 x *20 or *80 per cent. 



If the magnetic moment of the magnet when placed in a field 

 of zero strength be M , then when it hangs in a field H its moment 

 is M (1 + \H) where A, = '0111 in the experiment just described. 

 Hence if T E be the time of vibration under the earth's force H E , we 

 find that the magnetic moment in a field which gives the time T 

 is approximately M (l + \H E T E 2 /T 2 ). 



The aluminium pointer allows the time of vibration to be 

 accurately observed even though the arc is quite small. Thus for 

 30 vibrations in the earth's field, the initial arc being 10° on 

 either side of the zero, I found 189-6, 189"7, 190-0, 189-7 sees., 

 the greatest divergence from the mean (189*8) being only about 

 yoth per cent. 



§ 4. In accurate work a small correction must be made for the 

 torsion of the suspending fibre. The magnetic north and south 

 line is first marked out on a drawing-board by aid of a long, 

 pivoted magnet, and the torsion head from which the bob is hung 

 is so adjusted that the magnet and the aluminium pointer are 

 parallel to the iV — S line, when the system is at rest. The magnet 

 is next turned through one complete revolution and is then set 

 free, and is brought to rest. The magnet will no longer be 

 parallel to the N — S line but will make an angle of (f> radians 

 with it ; this angle can be measured by a small protractor. If //, 

 be the torsional couple per radian of twist, M the magnetic 

 moment of the magnet, and H E the earth's horizontal force, then 



^ = MH E £^^=fMH E , szy. 



