256 



Each of these results in this Table, is the difference of the 

 rates when the magnets were applied, and a mean of the rates of 

 the chronometers before and after the application of the mag- 

 nets. The rate of the chronometer No. 3, is very different from 

 that given in the former Table ; that of No. 4 is nearly the same, 

 and does not appear to have been affected, as No. 3 was, by the 

 close approach of the magnet. 



Upon the whole, however, it appears that chronometers will 

 be generally accelerated (particularly if their balances have not 

 received polarity by the too near approach of any thing magneti- 

 cal) on ship-board. It appears probable, Hkewise, that the 

 force of the balance springs is affected in the same way; since 

 it is well known that chronometers having gold balance springs, 

 although more difHcuU to adjust, yet keep better rates at sea than 

 the others. 



However this may be, these observations show the necessity of 

 not trusting to the rates of chronometers ascertained during the 



time 



