3 86 Mr. Barlow on the effects produced in the rates of 
with certain parts of a chronometer, and to appoint a day to 
come to Woolwich, and make such experiments as might 
suggest themselves to him or to me in the interval. He there- 
fore prepared for the purpose a new compensation balance, 
in which, of course, the usual care was taken to prevent the 
excitement of any local magnetism ; he also brought with 
him a brass balance, with two springs of different tempers, 
which might, either of them, be affixed to the balance in the 
usual way ; he had likewise, beside the proper frame for sus- 
pending these parts, constructed a brass stand, whereby the 
whole might be nicely adjusted to horizontality. 
Our first experiment was on the new compensation balance; 
but although it was brought almost in contact with the iron 
ball, and at that place where, by means of our experiments, 
page 370, the intensity was known to be the greatest, no 
action whatever could be discovered. We afterwards re- 
peated the same experiments in several other places, but 
without producing the least apparent effect. The weights 
of the balance were now removed, in order to render it more 
light and sensible ; but no species of action could be discerned. 
We now detached the balance entirely from the ball, and 
presented to it the north end of a bar magnet ; and then, 
giving the balance a very slight motion, it stopped after a 
short time, and arranged itself, so that the cross steel bar 
was directly in a line with the magnet; and immediately, 
upon being disturbed from its position, it returned to it again. 
The balance being now turned half round, so that the other 
end of the bar was directed towards the magnet, the same 
effect was produced. We now turned the magnet end for 
end, but found the attraction still the same between either 
