BeppickEr—On the Influence of Magnetism on the Rate of a Chronometer. 47° 
needle; he thinks the most probable explanation would be the assumption that 
the magnetism was chiefly lying in the spiral spring instead of the balance. 
Finally, Airy condenses the above method of compensating the effect of terrestrial 
magnetism into six rules, of which the first four describe the way of adjusting 
the height of the chronometer above the compass. No. 5 points out that in 
different magnetic latitudes it will be necessary to repeat the experiment from 
time to time, and to alter the adjustment accordingly. And, lastly, No. 6 runs 
literally : ‘‘ The amount of discordance depending on magnetic action is a matter 
of no consequence whatever, and needs not to be ascertained beforehand. The 
same arrangement of the compass which corrects a small discordance will correct a 
large one, and will do no harm to the chronometer if it have no discordance 
whatever.” 
XIV. 
From the experiments and speculations, reproduced and discussed in the 
preceding pages, we perceive that experiments with single chronometers are not 
likely to lead to general results. But still it appears that such experiments would 
be exceedingly valuable and interesting, since they would always throw more light 
upon the peculiarities and complications which the disturbance of rate produced by 
magnetism may assume.in different instruments. It is unnecessary to say that 
these experiments would increase in importance with the number of chronometers 
made use of. It is therefore highly desirable that the institutes for testing chro- 
nometers would take up these magnetic investigations. 
The chronometers experimented upon should all be instruments whose compen- 
sation for the effects of temperature is known to be efficient within the usual limits, 
or has been sufficiently determined, so as to render the elimination of the errors 
arising from this source possible. Besides this, the precaution should always be 
used to keep the temperature—if necessary, by artificial means—as equal as pos- 
sible throughout the duration of the experiments. As a further general rule, it is 
to be deduced from the preceding pages, that the most comparable and important 
results would be obtained if the different kinds of experiments described, viz., 
observations of chronometers in different azimuths, under the influence of soft 
iron, and of magnets; were all made, one after the other, upon one and the 
same chronometer. Thus a relation between the effects of terrestrial and arti- 
ficial magnetism and of soft iron (or induced magnetism) might probably be 
arrived at. It is likely that a chronometer which is influenced by terrestrial mag- 
netism will also (and only then) be affected by soft iron, and that, on the other 
hand, a magnet will disturb each chronometer, no matter whether it shows reac- 
tion to the earth and soft iron, or not. 
