264 Prof. W. Stroud. [June 19, 



vantages of this form of deflector have not even yet been enumerated. 

 The circular magnet may be made, or, rather, requires to be made, 

 very weak indeed ; this arises from the fact that the action of the two 

 poles on the needle is a summational instead of being a differential 

 one, as in the usual method of performing the deflection experiment. 

 There can be little doubt, too, that it is an advantage rather than the 

 reverse to use only weak magnets in determining the value of H. 

 Lastly, any variation in magnetic moment arising from, changes in 

 temperature or other causes does not affect the determination, and, 

 what is a matter of some importance in accurate determinations, 

 there is no correction corresponding to that required in Gauss's 

 method for the varying inductive action of the earth in the different 

 positions which the deflector assumes with reference to the magnetic 

 meridian. 



Description of Instrument. 



The instrument consists of a rectangular wooden box, ABCD, 

 fig. 1, mounted on levelling screws, and provided with a plate-glass 

 window in one of the large vertical sides, which can be opened for ob- 

 taining access to the interior. Attached to the upper side of this 

 box is a second, EF, also provided with a door for adjusting the 

 bifilars G,G in position. The little needle N is suspended at the 

 centre of the large box by a silk fibre some 10 cm. long attached to 

 a brass arm, K, which is screwed into the side of the box opposite to 

 the plate-glass door. Fastened to the needle at right angles to its 

 magnetic axis is a plane mirror (P), 1 cm. in diameter. The needle 

 and attached mirror are prevented from turning completely round by 

 a foi'ked piece of wood, F, which also enables the experimenter to 

 observe when the needle is at the centre of the box. 



Soldered to the large circular magnet MMM are two hooks of 

 brass of an indented V" sna P e ' H,H, figs. 1 and 2. These are for 

 suspending the magnet from the brass bar L, which forms the lower 

 end of the bifilar arrangement (figs. 1 and 2). The form of this 

 bar is a knife-edge of brass with a V- no ^ cn > Q> near one en( 5j so 

 made with the object of enabling the circular magnet to be un- 

 hooked and reversed in position. This eliminates any error arising 

 from the circumstance that the plane of the circular magnet may 

 not be placed accurately magnetic east and west when the bifilars 

 are vertical. A long aluminium wire, W, riveted at each end to the 

 bar, descends in the form of a wide loop, and carries a plane 

 mirror, R, to enable the deflection of the circular magnet to be read 

 off. This mirror is placed just below the mirror P, previously 

 mentioned, so that only one telescope is needed in reading the de- 

 flections of both needle and magnet. Soldered to the bar is a strip 



