FROM THE MAGNETISM OF THEIR BALANCES. 357 



tism, a change of rate took place, whenever a change was made 

 in the relative position of the watch and the magnet. Had 

 any of the balances been free from magnetism, its rate, in all 

 horizontal positions, I expected, would have been uniform, 

 whether under the action of terrestrial, or of moderate artificial 

 magnetism: For although iron acquires polarity, by mere 

 juxta-position with a magnetised body, and the balances of 

 watches placed near magnets become magnetic ; yet, were the 

 magnetism of the balance only transient, the poles would 

 change with every change of position in the watch, and its 

 rate would be unaffected by any slight magnetic influence; 

 but were any of the balances already permanently magne- 

 tic, there would be a repulsive action from the magnet in 

 some positions, and an attractive in others, which affecting the 

 vibrations of the balance in different ways, might be expected 

 to produce a change in the rate of the watch. These effects, 

 it was presumed, would be the same whether the watch were 

 tried in different positions, under the influence only of terres- 

 trial magnetism, or under the action (directed in a similar 

 way) of artificial magnetism ; the only difference being in the 

 quantity. ' - "■ " 



One of the watches under experiment was remarkable for 

 the beauty of the movement, and the uniformity of its rate. 

 I shall confine my remarks, therefore, to the results obtained 

 with this. 



A bar magnet, 12 inches long, was laid in the magnetic 

 meridian, and the watch placed in the same line, four inches 

 distant from the magnet, with its XII o'clock mark towards the 



-^"•''^ north. 



at the centre. The -other balances had generally two pdes <aily'; but, in some, 

 the poles of the rim were not exactly coincident with the pole? of the rays. 



