ACTION OF MAGNETS ON GASES, ETC. 
57 
as to persuade one at once that air must have a great and perhaps an active part to 
play in the physical and terrestrial arrangement of magnetic forces. 
2433. At one time I looked to air and gases as the bodies which, allowing attenu- 
ation of their substance without addition, would permit of the observation of corre- 
sponding variations in their magnetic properties ; but now all such power by rarefac- 
tion appears to be taken away ; and though it is easy to prepare a liquid medium 
which shall act with other bodies as air does (2422.), still it is not truly in the same 
relation to them ; neither does it allow of dilution, for to add water or any such sub- 
stance is to add to the diamagnetic power of the liquid ; and, if it were possible to 
convert it into vapour and so dilute it by heat, it would pass into the class of gases 
and be magnetically undistinguishable from the rest. 
2434. It is also very remarkable to observe the apparent disappearance of mag- 
netic condition and effect when bodies assume the vaporous or gaseous state, com- 
paring it at the same time with the similar relation to light ; for as yet no gas or 
vapour has been made to show any magnetic influence over the polarized ray, even 
by the use of powers far more than enough to manifest such action freely in liquid 
and solid bodies. 
2435. Whether the negative results obtained by the use of gases and vapours de- 
pend upon the smaller quantity of matter in a given volume, or whether they are 
direct consequences of the altered physical condition of the substance, is a point of 
very great importance to the theory of magnetism. I have imagined, in elucidation 
of the subject, an experiment with one of M. Cagniard de la Tour’s ether tubes, 
but expect to find great difficulty in carrying it into execution, chiefly on account of 
the strength, and therefore the mass of the tube necessary to resist the expansion of 
the imprisoned heated ether. 
2436. The remarkable condition of air and its relation to bodies taken from the 
magnetic and the diamagnetic classes, causes it to point equatorially in the former 
and axially in the latter. Or, if the experiment presents its results under the form 
of attraction and repulsion, the air moves as if repelled in a magnetic medium and 
attracted in a medium from the diamagnetic class. Hence it seems as if the air were 
magnetic when compared with diamagnetic bodies, and of the latter class when com- 
pared to magnetic bodies. 
2437- This result I have considered as explained by the assumption that bismuth 
and its congeners are absolutely repelled by the magnetic poles, and would, if there 
were nothing else concerned in the phenomenon than the magnet and the bismuth, 
be equally repelled. So also with the iron and its similars, the attraction has been 
assumed as a direct result of the mutual action of them and the magnets ; further, 
these actions have been admitted as sufficient to account for the pointing of the air 
both axially and equatorially, as also for its apparent attraction and repulsion ; the 
effect in these cases being considered as due to the travelling of the air to those 
positions which the magnetic or diamagnetic bodies tended to leave. 
MDCCCXLVI. I 
