320 Captain Wilson's experiments in China and St Helena, 



trials ; for although, when within the influence of the ball, the 

 needle would be materially affected, it would, when left to it- 

 self, take its natural direction in the true magnetic meridian. 

 I therefore placed the needle on the east or west point of the 

 table, and lowered the shell, so that its centre was level with 

 the pivot of the needle, in which position the shell, according 

 to Mr Barlow's experiments, ought not to have any effect. In 

 this position I found, that, by rubbing the needle in different 

 parts with the magnet, I could cause a greater, less, or con- 

 trary deviation at pleasure, that is, if the south end of the 

 needle was attracted by applying the north pole of the magnet 

 to the south point, it would cause it to take its true direction ; 

 or, if it failed to do so completely, by applying the south pole 

 of the magnet between the north end of the needle and the 

 centre, it would be accomplished, and vice versa. 



I am not aware that Mr Barlow experienced the same devia- 

 tions in his needle, probably not, as the less the angle that the 

 plane of no attraction makes with the horizon, the less the 

 needle would be inclined to deviate. Having discovered the 

 mode of rectifying the needle, if I may so call it, I proceeded 

 with the experiments, confining myself to the 75th, 80th, and 

 85th degrees of the circle, from the north and south towards 

 the east and west, as they were, from the difficulty of getting 

 the shell in the exact centre of the circle, the most convenient, 

 frequently placing the needle on the east or west points to cor- 

 rect it, and found the angle that the plane of no attraction 

 makes with the horizon to be, as in the annexed table, the dip, 

 as found by my dipping needle, being 31° 20' north. 



Mean, 54°55 

 Radius of the circle, 10.55 inches. Diameter of the shell, 

 12.68 inches. 



II 



