on the Rates of Chronometers. 7 
from detached rates, differing only from equality by the small 
quantity 0".3. 
These effects having been produced by the application of the 
north pole of the magnet to the main-springs and other parts 
of the chronometers, an attempt was next made to ascertain how 
far a similar application of the south pole would be productive 
of corresponding effects. For this purpose the magnet was in- 
verted, and its south pole applied to that part of the chronometer 
(B), before occupied by the north pole, as in Fig. 9.? when the 
rate changed from -f- 1".4 to — 1".0; so that the application of 
the south pole increased still farther the losing rate the time- 
keeper had acquired, in consequence of its position having been 
altered from Fig. 3. to Fig. 4. By turning the chronometer, 
so that it might occupy the position represented in Fig. 10., the 
mean rate of four days gave a result of q- 9". 5, clearly proving, 
as in the application of the north pole, that the main-spring was 
powerfully influenced by the magnet. This mean rate, how- 
ever, was less in quantity than either of the rates obtained by 
appl 3 ung the north pole ; and which circumstance would seem 
to indicate, that it resulted from the feebler influence of the 
south pole. By altering the position of the chronometer to ac- 
cord with Fig. 11., the rate changed from -f- 9". 5 to — 
the declension harmonizing in some degree with the observations 
recorded, when the chronometer occupied the position of Fig. 
When the machine was afterwards brought into the situation of 
Fig. 12., the mean rate for four days amounted to -|- 8". 7 ; and 
when restored to its position in Fig. 9., the rate returned to 
— 0".8, agreeing within two-tenths of a second of its former 
rate in the same position. During the continuance of the chro- 
nometer in this situation, the magnet was again inverted, and 
the north pole applied to the part before in contact with the 
south pole, when the change in the rate was from — 0".8 to 
4-0".4; agreeing within a second of its former rate ; and 
which circumstance, from the chronomete.r having gained by 
changing the pole from south to north, corresponding with the 
remark before made, of its having lost, by turning the pole from 
north to south. After remaining in the last mentioned position 
four days, the chronometer was again turned into the position 
denoted by Fig. 10., when its rate increased from + 0".4 to 
