22 



MAGNETISM. 



(88.) Magnetic curves of a different 

 kind are constituted by the balanced 

 actions of two poles of the same de- 

 nomination placed near to each other. 

 When, for instance, a second north pole 

 N' (fig. 38) is substituted, instead of the 

 south pole S, both poles will act in a 

 similar manner, and in directions not 

 very different. 



Fig. 38, 



lar. For this purpose they must be 

 taken in a different order of arrange- 

 ment, and followed in the lines of the other 

 diagonals of the lozenge-shaped intervals 

 between the intersecting radii ; that is, 

 of the diagonals which cross those con- 

 stituting the curves in the former case, 

 as is shown in the lower half of fig. 34. 

 These divergent curves, as they have 

 been called in contradistinction to the 

 former or convergent ones, are deli- 

 neated in. fig. 39 ; and may, in like man- 



Fig. 39. 



In order to render the conditions of 

 this case as simple as those of the last, 

 \ve must suppose that the action of the 

 south poles belonging to the two north 

 poles N, N', whose action we are ex- 

 amining, is, from their remoteness, too 

 feeble to influence the results. In the 

 former case, where the actions of the 

 two poles were of a contrary kind, the 

 resultant of their joint action, or the line 

 C T (fig. 31) passed in a direction in- 

 termediate between N C prolonged and 

 C S, and therefore cut the axis N X 

 at some point in the prolongation of 

 N S. But in the present case, the two 

 magnetic poles being of the same kind, 

 their action is similar, and their resultant 

 is a force of which the direction is inter- 

 mediate to the lines C N and C S ; and 

 this line produced must cut the axis 

 somewhere between N and N'. In con- 

 sequence of this change of position, 

 which produces a change in the sign of 

 the cosine of the angle GST, which is 

 now C N'T, the relation of the cosines of 

 the polar angles is as follows ; namely, 

 that the sum (and not, as before, the 

 difference) of the cosines of the angles 

 which lines, drawn from any point in 

 the curve to the two poles, make with the 

 axis, taken on the same side, is constant. 

 This applies to the case in which the 

 angle formed by C N' with the pro- 

 duced axis is acute, and its cosine posi- 

 tive. When it is obtuse (or CN'N 

 acute), the cosine becoming negative, it 

 is their difference which is constant. 



(89.) The intersections of the radii, 

 drawn according to the method above 

 described, 82, will also point out the 

 course of those curves which belong to 

 the case where the acting poles are simi- 



ner, be exhibited by the arrangements 

 of iron-filings round two similar poles. 



(90.) When the actions of the four 

 poles of two magnets are taken into ac- 

 count, the magnetic curves expressive of 

 the direction of a needle influenced by 

 them, become, of course, much more 

 complicated. 



CHAPTER III. 



Terrestrial Magnetism. 



$ 1 . Variation of the Compass. 



(91.) IT has been already stated ($ 6), 

 that if a magnetic bar be poised on its 

 centre so as to move freely in a horizon- 

 tal plane, and if no ferruginous body be 

 sufficiently near to affect it sensibly, it 

 will assume, when left at liberty, a di- 

 rection nearly north and south. When 

 disturbed from this situation, it returns, 

 after several oscillations, to the same 

 position. On this property is founded 

 the mariner's compass, which is of such 

 essential use in navigation. In moving 

 horizontally towards the position which 

 it thus tends to assume, the needle of the 

 compass is said to traverse. 



(92.) It is found that in this country, 

 as well as throughout Europe, the north 

 pole of the compass deviates a certain 

 number of degrees to the westward of 

 the exact northern direction. This de- 

 viation from the true geographical me- 

 ridian haa been called the magnetic 



