8 BappickerR—On the Influence of Magnetism on the Rate of a Chronometer. * 
IE 
The observations which prove the magnetic state of different steel parts in the 
work of a chronometer are of considerable interest, as they show which parts are 
chiefly liable to become magnetic, and how the magnetic poles are situated, and 
because they contain some valuable suggestions as to how to avoid or to remove 
this cause of disturbance of a chronometer’s rate. The first statement was given 
by S. Varley (No. 1), who was led, by the inferior performance of first-rate 
watches, to test their balances upon magnetic polarity. He took the steel balance 
out of a watch, removed the balance-spring, and put the balance into a poising 
tool, in order to observe the effect a magnet would have upon it. But before he 
could begin the experiments, and before he approached a magnet, he found that 
the balance in one position of the tool seemed to be out of poise (though in that 
particular examined before by a very careful workman), and, in another position 
of the tool, in poise. So, when placed vertically, the balance was in poise as often 
as its plane was east and west (and, of course, its axis north and south), and 
whenever it was placed north and south—“ the only position in which the magnetic 
influence could make itself most apparent”—it was out of poise. Then the axis 
was placed in a vertical position, and the balance was found to have sufficient 
polarity to overcome the friction of the pivots, and to turn readily its north pole 
towards the north, ‘‘in every respect like an equally-suspended magnetic needle.” 
And when now a magnet was approached, the north pole of the balance continued 
at rest before the magnet’s south pole, and receded immediately when the north 
pole was presented. 
Varley now—after having made some observations of rate-changes with the 
above balance—resolved always to try a steel balance before application. He laid 
them upon a slice of cork sufficient to make them float on water, but out of many 
dozens he could not select one that had not acquired polarity. Some of them had 
it in a weak degree, and not more than one or two out of the whole quantity had 
it so strong as the one with which he had made the above experiments. 
Barlow (No. 3) communicates some experiments upon detached parts of a 
chronometer in the proximity of a mass of iron, and gives some ideas how always 
to distinguish if the balance or the approached iron is magnetic. I do not consider 
it necessary to repeat these rules, and of his experiments only the following facts 
are to be reproduced. A balance with its spring was brought near a piece of iron, 
and a slight repulsion in one and attraction in another position were observed, 
indicating a slight degree of magnetism in the balance or its spring. Mr. 
Frodsham—in whose workshop the observations were made—‘‘ had no doubt that 
such an action as we then noticed was amply sufficient to change the rate of the 
chronometer, of which the balance formed a part, when brought within the sphere 
