xlii Introduction to the Makerstoun Observations, 1844. 



vibration was scarcely exhibited ; and it may be added, that in eight months since the 

 needle has been adjusted with its axis in the magnetic meridian, the time of vibra- 

 tion appears to remain nearly constant and equally independent of temperature and 

 arc of vibration. 



56. It appears certain from these results, obtained from an instrument of the 

 best character, treated with the greatest care, that the time of vibration in a vertical 

 plane cannot be depended on as an element in the reduction of the observations of 

 the balance magnetometer. 



57. The question still remains, to what extent these anomalies in the time of 

 vibration affect the observations for the varying vertical component of the magnetic 

 force ? The following fact appears to render it certain that they have no effect 

 whatever. In determining, by the method of comparisons (see pages 1. and li.), the 

 effect of a change of 1° Fahr. upon the position of the needle, it has been found 

 that this effect is nearly constant, while the time of vibration in the vertical plane 

 has varied from upwards of 11 seconds in 1843, to less than 6 seconds in 1846 

 (No. 77, 3d) ; the differences of the results for the temperature coefficient being in 

 all probability due to considerable changes of vertical force in the periods selected 

 for the determinations, and certainly having no relation whatever to the varying 

 time of vibration. Since the temperature coefficient in micrometer divisions (</') 

 has remained constant, it follows that the coefficient of reduction (k) must also be 

 constant. This conclusion renders it the more desirable that the value of the angu- 

 lar motion of the needle in parts of the vertical component should be determined by 

 another method which does not involve the time of vibration ; the statical method 

 already described for the bifilar magnetometer, has been employed for this purpose 

 with some modification. 



58. January 6, 1848. Wooden beams having been placed horizontally at right 

 angles to the magnetic meridian, and a line having been drawn upon them, which 

 was a projection of the prolongations of the balance needle, a small deflecting bar 

 (3'65 inch long, and having a temperature coefficient = 0*000285), was placed verti- 

 cally at different distances on the beam, and the angles of deflection of the needle 

 were observed ; the centre of the deflecting bar was in the prolongation of the axis 

 of the balance needle when horizontal, and the distances were measured from the 

 knife edges of the axle, which, however, was found not to be the centre of magnetism 

 of the needle. The resulting deflections and values of lc for each distance will be 

 found Table 6. 



January 10, 1848, the balance needle was taken out of its box, and attached to 

 the brass detorsion bar of the declinometer ; the brass bar was mounted with a glass 

 scale and lens, and was suspended in the declinometer box ; the balance needle was 

 then deflected by the bar used for the previous deflections, which was placed to the 

 north and south of the suspended needle, so that in both series of deflections the 

 prolongation of the balance needle in its normal position passed through the centre 

 of the deflecting bar. The deflections for three distances are given, Table 7. 



