TRANSACTIONS OF THE SECTIONS. 
563 
free; in which case, it will continue to vibrate until it again 
comes to rest in the direction of the parallel threads. If the 
number of vibrations made by the ring alone in this way, be¬ 
tween any given limits on the graduated cord, be divided by 
the number made within the same limits, when allowed to 
vibrate about the bar, then the quotient minus 1, may be taken 
as a comparative numerical value of the force of the bar, and 
will be a constant quantity, whatever may be the limits within 
which the vibrations are taken. 
Of the ordinary causes of change in the magnetism of a bar, 
perhaps an increase of temperature may be considered as one 
of the most important. The effect, however, seems to be dif¬ 
ferent for different bars, and also to vary with their magnetic in¬ 
tensity; so that in certain bars, and under certain temperatures, 
the magnetic state may remain unchanged. The results of very 
many experiments on the influence of heat on magnets, has led 
the author to conclude, that a bar formed of the best double 
steel of commerce, well hammered and tempered, may, after 
being rendered magnetic, obtain an invariable magnetic inten¬ 
sity, for all temperatures under 200° of Fahrenheit; by boiling 
it for a short time in water, at 212°; or otherwise, by touching 
the bar, after being prepared as before, whilst its temperature 
is raised to 250 or 260 degrees. 
A decrease of temperature will not restore any of the power 
which heat has destroyed. Thus a bar, after being magnetized 
to saturation, was tested by the rate of vibration, and subse¬ 
quently exposed to the heat of boiling water, when it was again 
tested in the same wav, so as to ascertain the amount of the 
force it had lost by this process. On being afterwards im¬ 
mersed for an hour in a freezing mixture, which reduced its 
temperature to 0 of Fahrenheit, it was again vibrated as before, 
the previous temperature of the bar being first restored ; but 
no increase of power was apparent; rather, on the contrary, 
the bar appeared to have lost somewhat more of its power, but 
in a very slight degree. It would seem, therefore, that correc¬ 
tions for the effects of heat on the magnetism of a bar, are 
limited to such elevations of temperature only as happen to be 
above that to which it has been already exposed. 
The author observes, that the effects of heat on the mag¬ 
netic distribution of a bar, are very precarious and uncertain, 
since the magnetic disturbance tends more readily to a state of 
neutrality in some bars than in others: it is hence requisite to 
examine the influence of heat experimentally, on each indivi¬ 
dual bar, previously to its being employed for investigating ter¬ 
restrial magnetism. 
2 n 2 
