312 MAGNETISM 



Values frequently chosen are d = 30 cm. and d' 40 cm. 



Corrections for temperature and for magnetism induced bv tin- 

 earth in the deflecting magnet must be made. 



Dr. Chree has shown* that the approximation represented by 

 the factor 1 + P/f* is not justifiable with the magnets usually 

 employed and that another term, Q/r 4 , should be taken into account. 

 The reader is referred to Dr. Chree's papers for a discussion of this 

 point. 



The value for H in this country is in the neighbourhood of 0'18, 

 and different instruments may give values di fieri ng by as much 

 as two or three units in the fourth place, say by as much as 1 in 

 1000 of the whole. 



In the regions near the magnetic pole-, i.e. the two points 

 where the force is vertical, the hori/ontal intensity is only a small 

 fraction of the total intensity and too small to be nun- 

 accurately. Special instruments which we need not describe f have 

 been devised to measure both dip and total intensity in these i 



Recording instruments. H.M.I,., the instruments \\hich 

 give the absolute values of the magnetic elements, it is usual to 

 equip magnetic observatories with self-recording instruments which 

 register continuously the small changes which are alwav> occurring 

 in the direction and intensity of the force. For this purpose t 

 magnets are used, one to record changes in declination, another 

 those in horizontal intensity, and a third those in vertical intensity. 

 A satisfactory self-recording ; dip instrument ha> not been de\i-. d. 

 Each magnet is suspended in a dark box, and is furnished \\ith a 

 mirror which reflects a small beam of light from a lamp on to a 

 volvingdrum covered with sensitixd paper.J Each magnet is sur- 

 rounded with a thick copper ring to damp it- vibrations. The 

 declination magnet is suspended so that it is five to set in the 

 direction of the hori/ontal component. The hori/ontal force 

 magnet has a bifilar suspension, and the torsion head is tiirn<d 

 round so that the magnet is at right angles to the a .tic- 



meridian. Let Fig. 232 represent a plan. If II changes in direction 

 only and not in magnitude, the change being through the small 

 angle <S, we shall have H sin 6 along SN with no turning effect. 



and H cos 8 H f 1 -^-J perpendicular to NS : therefore the moment 



on the magnet is decreased by MH ' which is negligible. A \ 



small change in direction is, therefore, without effect. If H 

 changes in magnitude, say from H to H + ^H, the moment on the 



* Proc. Roy. Soc,, Ixv. (1899), p. 375; Phil Man. [t;], vol. viii. (Aug. 1904), 

 p. 113. 



t See Walker's Terrestrial and Cosmical Magnetism, p. 209, for early forms. 

 Recent forms are described in Rational Antarctic Expedition Magnetic Observation*, 

 Royal Society, 1909. 



J For details see references in footnote on p. 301. 



