and Characters of the Metals. 179 
iron, but equally defined and distinct. If heated and then 
cooled, it remained unmagnetic long after it had fallen below 
a heat visible in the dark: and, in fact, almond oil can bear 
and communicate that temperature which can render nickel 
indifferent to a magnet. By a few experiments with the ther- 
mometer it appeared that the demagnetizing temperature for 
nickel is about 630° or 640°. A slight change about this 
point would either give or take away the full magnetic power 
of the metal. 
Thus the experiments, as far as they go, justify the opinion 
advanced at the commencement of this paper, that all metals 
have similar magnetic relations, but that there is a certain 
temperature for each beneath which it is magnetic in the man- 
ner of iron or nickel, and above which it cannot exhibit this 
property. This magnetic capability, like volatility or fusibi- 
lity, must depend upon some peculiar relation or condition of 
the particles of the body; and the striking difference between 
the necessary temperatures for iron and nickel appears to me 
to render it far more philosophical to allow that magnetic 
capability is a general property of ail metals, a certain tem- 
perature being the essential condition for the development of 
this state, than to suppose that iron and nickel possess a 
physical property which is denied to all the other substances 
of the class. . 
An opinion has been entertained with regard to iron, that 
the heat which takes away its magnetic property acts somehow 
within it and amongst its electrical currents (upon which the 
magnetism is considered as depending) as flame and heat of 
a similar intensity act upon conductors charged with ordinary 
electricity. The difference of temperature necessary for iron 
and nickel is against this opinion, and the view I take of the 
whole is still more strongly opposed to it. 
The close relation of electric and magnetic phenomena led 
me to think it probable, that the sudden change of condition 
with respect to the magnetism of iron and nickel at certain 
temperatures, might also affect, in some degree, their con- 
ducting power for electricity in its ordinary form; but I could 
not, in such trials as I made, discover this to be the case with 
iron. At the same time, although sufficiently exact to indicate 
a great change in conduction, they were not delicate enough 
to render evident any small change; which yet, if it occurred, 
might be of great importance in iilustrating the peculiarity of 
magnetic action under these circumstances, and might even 
elucidate its general nature. 
Before concluding this short paper, I may describe a few 
results of magnetic action, which, though not directly con< 
U2 
