244 Mr. Christie on the Permanence of 



Finding that not only no change took place in this time, 

 but that in much longer intervals scarcely any appreciable 

 change could be observed, I was led to make observations 

 with the view of determining whether, during a very long 

 interval of time, the magnetism unequally distributed in a 

 steel bar would return to a state of symmetrical distribution. 



For this purpose I made use of four bars of steel, which 

 had been somewhat softened, in order that they might be 

 reduced to the same thickness by filing, and which I had not 

 afterwards hardened, as I considered that with such bars 

 changes would be most likely to occur. These, for the sake 

 of distinction, I marked I., II., III., IV. Each of them is 

 0'15 inch in breadth, and 0*1 inch in thickness ; I. and II. 

 are 8'91 inches, and III. and IV. 5'94 inches in length. 

 I. and II. were placed by the side of each other, and strongly 

 and carefully magnetised by double touch, by means of two 

 twelve-inch bar magnets ; and the same was done to III. and 

 IV. As usual, the south pole of each was indicated by a 

 mark on that end of the bar ; and in order to avoid any am- 

 biguity which might arise from a change of position in the 

 bars, when I determined their magnetic centres and poles, I, 

 in all cases, made use of the terms, marked and unmarked 

 ends, instead of south and north poles. The magnetism of 

 I. and III. was disturbed by passing the marked end of a 

 twelve-inch bar magnet from their centres to their marked 

 ends twice ; and that of II. and IV. was disturbed by a similar 

 operation with the unmarked end of the bar magnet, from 

 their centres to their unmarked ends. 



In order to determine the position of the magnetic centre in 

 each bar, I placed it on a rectangular wooden scale, parallel 

 to, and equally distant from, the sides, the scale being fixed so 

 that the axis of the bar was in the magnetic meridian, and its 

 marked end towards the north. A compass, with a small 

 trial needle, an inch in length, was fixed on another rectangular 

 piece of wood furnished with a vernier ; so that the scale being 

 graduated across to tenths of an inch, and the side of the 

 rectangle having the vernier being applied to it, the position of 

 the point of the magnetised bar opposite to the centre of the 

 trial needle could be determined to the hundredth of an inch. 

 The side of the rectangle carrying the trial needle and that 



