TERRESTRIAL MAGNETISM. 743 



It is extremely difficult to prepare a needle which before magnetisa- 

 tion will rest in neutral equilibrium, and the measurement of the angle 

 of inclination requires very delicately constructed, expensive instru- 

 ments. For merely observing the dipping of the north pole, a 

 knitting-needle should be suspended by a thread tied round its 

 middle and adjusted until it swings horizontally. The knot should 

 be fixed with a little bee's- wax, to prevent it from shifting while 

 the needle is magnetised. After magnetisation it will assume a con- 

 siderably inclined position, although the angle of inclination is of 

 course sensibly smaller than that at which a needle would incline 

 which was suspended exactly at its centre of gravity. 



If the earth is a magnet it should be capable of 

 acting inductively upon soft iron. A bar of soft iron 

 becomes, indeed, feebly magnetic if it is held in the same 

 position which a freely suspended magnetic needle 

 assumes under the influence of the magnetism of the 

 earth, that is, if held in the approximate north- south 

 direction of such a needle, and at the same time so as 

 to be inclined to a horizontal plane at an angle of about 

 68. The end of the bar which dips becomes a north 

 pole, the upper end becomes a south pole, as can be 

 proved by bringing the poles of a small magnet near the 

 end of the soft iron bar. The lower end repels the 

 north pole of the magnet, the upper end repels the 

 south pole. 



The longer and thicker the bar of soft iron, the more apparent 

 becomes its magnetisation by the inductive action of the earth. 

 The experiment is best made with a bar 0*5 to l m long, and 2 or 

 3 cm thick ; if such a bar cannot be obtained, a small iron rod of the 

 size of a somewhat large pencil may be used. The bar must be 

 heated and then allowed to cool very slowly, so as to be as soft as 

 possible. For testing the magnetisation of the bar and the distribu- 

 tion of its magnetism, a very small feebly magnetic needle must be 

 used ; otherwise the test-magnet itself acts inductively on the bar, 

 and attraction wouJd take place in any case. As test magnet a needle 

 bent like that in fig. 375, but without cap and pivot, tied to a piece 

 of untwisted thread, should be used ; a magnet which swings on a 

 pivot cannot be well used for the purpose, because in the position in 



