356 ON THE ERUORS OF CHRONOMETERS, ARISING 



and the balance replaced in the watch, Mr Varley laid the 

 watch with the dial upward, and the north pole of the balance, 

 as determined by the previous experiment, towards the north ; 

 — in this situation it gained 5' 35" in twenty-four hours. He 

 then directed the north pole of the balance towards the south, 

 every thing else being as before, and it now lost & 48" in twen- 

 ty-four hours. Mr Varley afterwards took away the steel- 

 balance, and substituted one made of gold ; then having 

 brought the watch to time, he carefully observed its rate, and 

 found it as uniform as any watch of like construction. He 

 subsequently examined many dozens of balances, out of which 

 he could not select one that had not polarity. 



The instance observed by Mr Varley is no doubt an ex- 

 treme one ; but yet the influence of magnetism in the balance 

 most probably affects the rate, in some degree, of almost every 

 watch and chronometer. I have attempted to ascertain how 

 far this influence might be considered as general, by experi- 

 ments both on detached balances, and on the rates of watches 

 under magnetic influence. Similar experiments have been 

 made by others ; but there were some circumstances neglect- 

 ed in all, which I was anxious to examine, particularly the 

 position of the watch or chronometer when its rate was deter- 

 mined. These experiments, owing to the arrival of the season 

 for undertaking my usual voyage to the Polar Seas, I have 

 been unable to complete ; but this much was accomplished : 

 In seven detached chronometer balances, very sensible mag- 

 netic properties, both attractive and repulsive, were found in 

 all * ; and in three watches subjected to the action of magne- 

 tism. 



• 111 some of these balances, the magnetism was strong. One, with three arms, 



had a vigorous south pole at the extremity of each ray, and a common north pole 



I at 



