xxx Introduction to the Makeestoun Observations, 1844. 



35. When the observations for 1843 were nearly through the press, observations 

 of deflections of the bifilar magnet were made at two distances; observations of 

 deflection of a unifilar magnet were also made ; the dimensions of the bifilar and 

 unifilar bars were very different, but, as the results for the two distances computed 

 by equation (2.) were very nearly equal, it was considered that the differences be- 

 tween^?, q, and p„ q x , might be neglected; the abstracts were accordingly formed 

 with the aid of the coefficient so obtained. It was discovered, after the volume was 

 published, that the difference of the distances employed was too small to exhibit the 

 error of the assumption that p=p x , &c. In consequence of this error, the coefficient 

 has been redetermined by different methods, as follow : — 



36. Wooden beams having been placed in the prolongations of the bifilar mag- 

 net, and at right angles to these, lines were drawn upon them, passing through the 

 centre of the magnet, — one in the magnetic meridian, the other at right angles to 

 it ; several distances from the centre of the suspended magnet were marked off on 

 each side with a beam compass ; a similar structure was erected for the declinome- 

 ter. The following observations were then made : — 



1st, A cylindrical magnet, 3'65 inches long, was employed to deflect the bifilar 

 and declinometer magnets ; these two magnets are of the same dimensions, 15 inches 

 long, and were obtained at the same time from the same maker. The short de- 

 flecting bar was placed at different distances to the east, and at the same distances 

 to the west, of the bifilar bar, and the deflections of the bifilar were observed in 

 scale divisions. Observations of deflection of the declinometer magnet were then 

 obtained with the same deflector — the deflector, however, being placed at the same 

 distances, as in the other case, to the north and south of the declinometer magnet : 

 in both cases, the prolongation of the suspended bar, in its normal position, passes 

 through the centre of the deflector. The results are obtained in the 1st portion 

 of Table 4. 



2d, The same deflecting bar was placed to the north and south of the bifilar 

 magnet, and to the east and west of the declinometer magnet, the prolongation of 

 the axis of the deflector in both cases passing through the centre of the suspended 

 bar. The results are given in the 2d portion of Table 4. 



3d, A large deflecting bar (15 inches long) was employed in the same manner 

 as the small bar in the 1st instance. 



ith, The large deflecting bar was employed in the same manner as the small 

 bar in the 2d case. 



In the 3d case, deflections of the bifilar could only be obtained to the E, and, 

 in the 4th case, to the S of the bifilar magnet, owing to the proximity of the bifilar 

 to the walls of the Observatory. It was easy, however, from the observations with 

 the small bar to make the requisite corrections for the difference of deflection on 

 the opposite sides : the correction is small. The results for the 3d and 4th cases are 

 contained in the 3d and 4th portions of Table 4. 



