MAGNETISM. 



of computation are much more compli- 

 cated than those we have given; and 

 we must therefore refer our readers for 

 the details to Biot's Traite de Physique *. 



(335.) The determination of the in- 

 tensity of terrestrial magnetism may, in 

 general, be made with much greater 

 accuracy, by observing the oscillations 

 of a needle moving horizontally, than 

 in any other way : because the mode of 

 suspension we can employ for obtaining 

 a horizontal motion is much more deli- 

 cate, and much less impeded by friction, 

 than any other motion or a fixed axis 

 can be. The greater duration also of 

 the period through which the oscilla- 

 tions continue enables us to ascertain 

 with greater exactness the average time 

 of the vibration. Hence a silken sus- 

 pension is much to be preferred for 

 delicate experiments with horizontal 

 needles, to that of balancing them on 

 a point by an agate cup. 



(336.) The following description of 

 the apparatus used by Captain Sabine, 

 in the voyages to the arctic regions in 

 the year 1822 and 1823, may furnish 

 valuable practical information to those 

 who may hereafter conduct experiments 

 of the same nature t. A mahogany 

 box was provided, made, for convenience, 

 of an octagonal shape, with a top of 

 stout glass ; its height was fifteen 

 inches, t and its diameter sufficient to 

 allow a horizontal bar of seven inches 

 in length to vibrate freely when sus- 

 pended by a silk line passing through a 

 brass button, inserted in a perforation 

 in the middle of the glass top. A metal 

 circle, fixed in the bottom of the box, 

 of rather more than seven inches dia- 

 meter, measured the arc of vibra- 

 tion. The bar was carried in a light 

 stirrup, into which it was slid until cor- 

 rectly balanced. The silk thread, from 

 which the stirrup was suspended, was 

 fifteen inches long, and consisted of a 

 sufficient number of silk fibres to sus- 

 tain the weight. In order to remove 

 all influence from the tendency in the 

 silk to untwist, a brass bar, equal in 

 weight to the magnetic bar, was first 

 introduced into the stirrup in place of 

 the latter, and the silk thread allowed to 

 untwist itself, and then adjusted by 

 turning the button in such a manner as 

 that the brass substitute should settle, 

 when at rest, in the magnetic direction. 

 This being now removed, the magnetic 



* Tom. iii. p. 33. > 



j- Account of experiments to determine the figure 

 Of the earth, p. 477. 



bar was replaced in the stirrup, and its 

 horizontally ascertained by its accord- 

 ance with the circle, the degree to which 

 it settled being registered as the zero. 

 It was then drawn about forty degrees 

 out of the meridian, and retained by a 

 copper wire passing through the glass 

 top, and capable of being moved in azi- 

 muth from its outside, and of being 

 raised so as to release the needle at 

 pleasure, in order to commence its os- 

 cillations. These were not noticed until 

 the arc had diminished to thirty degrees, 

 when the registry of them commenced, 

 and was repeated at the close of every 

 tenth vibration, until the arc had still 

 further diminished to ten degrees, when 

 the experiment was concluded. The 

 box was usually placed on the ground, 

 in a sheltered situation, far from build- 

 ings, or other sources of local inter- 

 ference ; the only adjustment required, 

 besides that of the silk thread, was to 

 render the graduated circle horizontal, 

 which was accomplished by a pocket 

 spirit level, and wooden wedges placed 

 beneath the box. 



(377.) Six bars were used in this ap- 

 paratus, differing from each other con- 

 siderably, both in rapidity of vibration, 

 and in the duration of the interval of 

 oscillation between thirty and ten de- 

 grees. They were seven inches long, a 

 quarter of an inch broad, 0.15 inches 

 thick, and strongly magnetized. When 

 not in use, they were kept in pairs, in 

 the usual manner, as described in J 218, 

 Jig. 60, being combined, with their op- 

 posite poles united, in separate boxes ; 

 and each bar was placed by itself in the 

 direction of the meridian for two or 

 three hours before its time of vibration 

 was ascertained. The times were re- 

 gistered to fractional parts of a second 

 by the beats of a chronometer, having 

 a rate inappreciable in the interval. 



(338.) It should be observed, how- 

 ever, that comparative experiments on 

 magnetic intensities, made by the oscil- 

 lations of needles balanced horizontally, 

 are liable to a source of error when 

 they are made in places in which the 

 dip is considerably different. For one 

 of the poles having a tendency to in- 

 cline below the horizon, the axis of 

 suspension must, in order to cqmpen- 

 sate for this tendency, pass through a 

 point on that side of the centre of gra- 

 vity, so as to give an equal prepon- 

 derance to the other side. Hence arises 

 a difference between the two arms of 

 the lever, and consequently a difference 



