Oscillations of the Horizontal Needle. 13 
the magnetic pendulum, those induced in the magnetism of the 
bar itself, from temperature and other causes, may be considered 
the most important. The effects of heat in the magnetic tension 
seem to be extremely uncertain, since it tends more freely to a 
state of neutrality in some magnets than in others. It is therefore 
requisite also to examine experimentally, in the way above stated, 
the influence of change of temperature in each individual bar, so 
as to discover whether it has an invariable magnetic state or not. 
The influence of heat also, evidently decreases more rapidly than 
the magnetic intensity, so that bars of weak intensity seem inva- 
riable under certain low ranges of temperature. I have hence 
found it desirable to work with bars of low tension,{reated in the 
way above mentioned (16). The interesting and profound re- 
searches of Mr Curistiz*, however, would lead us to infer that 
certain variable states may occur in magnetic intensity from heat, 
without a final loss of force, and for which he has endeavoured 
to deduce adequate expressions. Although such authority must 
necessarily be highly appreciated, yet the imperfect state of our 
knowledge in the sciences of electricity and magnetism requires 
a very extensive examination of exceptions, deducible by expe- 
riment from any particular instance. 
24. After a careful inquiry into the effects of temperature on 
certain small magnetic bars employed to determine terrestrial in- 
tensities, I have not been enabled to discover, by the method of 
vibration, variable changes in their magnetic state, such, that the 
tension, being weakened by heat, it could again be restored by 
cold ; nor have I, in the particular instances which came under 
my notice, been enabled ever to increase the magnetic energy of 
a bar by reducing its temperature. On the contrary, the exposure 
of small bars to very low temperature, seemed, as tested by the 
method of vibrations, to be attended by a decreased energy. 
Thus, a small bar of moderate force was observed to make in va- 
* Transactions of the Royal Society for 1825, 
