chronometers by the proximity of masses of iron. 387 
end of the magnet, and on each end of the steel bar of the 
balance ; and in no case could there be obtained the slightest 
indication of repulsion ; from which we may conclude, that 
the balance itself was free from any polar magnetic quality, 
and that every part of it was alike susceptible of the power of 
the magnet, although it was wholly insensible to the action of 
the iron ball. A chronometer, therefore, with such a balance, 
and with a spring equally free from magnetism, would, I 
conceive, preserve the same rate both on shore and on ship- 
board, although it might be as sensibly affected with a magnet 
as any of those experimented upon by Mr. Fisher. 
We now took the brass balance, and having suspended it 
in its frame, applied to it the end of the bar magnet, in order 
to ascertain whether any magnetic quality could be discovered 
in the brass of which it was composed ; but no action of that 
kind could be rendered sensible. One of the springs being 
now attached to it, it was applied to the ball, whereby a small, 
but sensible, effect was produced by the action of the iron ; 
and by repeating the experiment in various ways, it was ob- 
viously of that kind which indicated magnetism in the spring ; 
and a very similar action was discoverable with the other 
spring. 
Each of these was now applied, at a short distance, to a very 
light and sensible compass needle ; when the polar quality of 
both was rendered manifest in a very peculiar manner, but 
which it is not necessary to detail in reference to the present 
enquiry. 
All we learn from these experiments appears to be, that 
when a balance, or its spring, acquires a magnetic quality, 
the rate of the chronometer, of which it forms a part, will 
mdcccxxi. 3D 
