the Magnetism in Steel Bars. 245 



of the scale being in contact, the former was passed along the 

 side of the latter, until the position of the trial needle was 

 exactly reversed ; that is, until its marked end pointed accu- 

 rately south : the point of the bar then opposite to the centre 

 of the needle, I considered as the zero, or magnetic centre 

 of the bar. The distance of this point from the centre of 

 figure of the bar was determined by passing the needle along 

 each side of the scale, and a mean of the two distances taken ; 

 and this distance. I designated M or U, according as it was 

 towards the marked or unmarked end of the bar. 



The points in the bars towards which the small needle, if 

 uninfluenced by terrestrial magnetism, would be directed, when 

 near to them, and which are nearly those where the magnetic 

 intensity is the greatest, I considered as their poles ; arid in 

 order to determine their positions in each bar, the scale was 

 placed at right angles to the meridian, so that the marked end 

 of the bar was towards the west. The trial needle with its 

 vernier was then passed along the north side of the scale until 

 the position of the needle was exactly reversed ; and then again 

 until it assumed its natural direction : the point of the bar 

 opposite to the centre of the needle, in the first case, I consi- 

 dered to be its north or unmarked pole ; and that opposite to 

 it in the second, its south or marked pole. The distances of 

 the poles from the centre of figure of the bar were similarly 

 determined by passing the trial needle along the south side of 

 the scale, the needle being, in this case, in its natural direction 

 when opposite to the unmarked pole, and having its direc- 

 tion reversed when opposite to the marked pole. I took a 

 mean of the distances thus determined for each pole, as the 

 distance of that pole from the centre of figure, although I 

 considered that the position of the needle, when its direction 

 was reversed, was most accurately determined, since in this 

 case the slightest change in its position caused a very sensible 

 change in its direction. I should here remark, that in these 

 bars there were no indications of other poles besides those 

 whose positions I determined, magnetism of the same character 

 predominating from the zero, or near that point, to one end of 

 the bar, and the contrary magnetism predominating from the 

 same point to the other extremity. 



The observations are contained in the following Table : 



