324° ‘Astronomical and Nautical Collections. 
important to repeat them with soft iron and with pure nickel at dif- 
ferent temperatures, and even with other metals, which have not 
hitherto been found to be subject to magnetical action: and, 
in fact, the experiment of Mr. Gay Lussac, which has established 
the difference of the action of the magnetic fluid in different sub- 
stances, may afford some reason to suspect, that the intensity of 
this action is only very weak at ordinary temperatures, though 
perhaps not absolutely wanting, in cther metals. 
After having explained the hypothesis, or rather the physical 
foundations of the question which forms the subject of this memoir ; 
we must endeavour to ascertain precisely in what manner we can 
represent, from these principles, the disposition of the boreal and 
austral fluids in magnetized bodies. 
Let us first suppose that the substance is a cylindrical needle of 
soft iron, of a very small diameter, and of any finite length; and 
that, in the direction of its axis produced, there are one or more 
centres of magnetic action. In the natural state of the needle the 
two fluids contained in it are united in equal quantities throughout 
its substance ; so that their actions being equal and opposite in all 
distances, they destroy each other completely, and no sign of mag- 
netism is exhibited. The effects of the centres of magnetic action 
on the two fluids will separate them from each other; but each 
boreal or austral particle will be very little removed from its pri- 
mitive situation ; and in this new state the two fluids will succeed 
each other alternately throughout the length of the needle, and this 
length will consist of a series of very small parts, each of which 
will contain, as in the neutral state, the two fluids in equal quan- 
tities. Itis unnecessary to inquire whether the length of these 
parts is equal to that of the constituent molecules of the iron; it is 
sufficient for our calculations that their length should be very mi- 
nute, so that it may.be neglected in comparison with the diameter 
of the needle, and, in general, with the smallest dimensions of the 
magnetized bodies which are to be considered. However small 
this length may be supposed, it is still conceivable that it may be 
different for the different substances which are capable of being 
magnetized, as for iron and for nickel; but it will appear inthe 
