Prof. Challis on a Theory of Magnetic Force. 95 
dynamic action on any given atom. We have not the data for 
inquiring what that action may be in any particular substance, 
because for that purpose we require to know its composition and 
the magnitudes and arrangement of its atoms. I shall there- 
fore assume, asa result of the experiments above referred to, 
that divergent and convergent magnetic streams act dynamically 
in different manners and in different degrees on a great variety 
of substances which they permeate. 
20. If the substance be iron, the moving force of the mag- 
netic streams on its individual atoms, apart from any action 
resulting from interference with secondary streams, tends along 
the lines of motion towards the poles of the magnet. ‘The fol- 
lowing remarkable experiment by Faraday (2368) establishes 
this law. Cylindrical glass tubes containing ferruginous solu- 
tions of different degrees of strength were suspended with the 
axes vertical in a ferruginous solution of given strength between 
the poles of the magnet. If the solution in the tube was stronger 
than that of the surrounding fluid, the tube was attracted to- 
wards either pole; and if weaker, it was repelled. Both these 
effects may be explained on the assumption that the magnet 
attracts the solutions in proportion to the quantity of iron they 
contain. The tube was drawn towards the poles in the first 
case, because the contained fluid being more ferruginous, was 
more attracted than the surrounding fluid. In the other case, 
the hydrostatic pressure of the surrounding fluid, due to the 
assumed magnetic attraction towards the poles, is greater on 
the side of the tube nearest either pole, than on the opposite 
side, and this difference of pressure, not being wholly counter- 
acted by the attraction on the contained weaker solution, causes 
a repulsion of the tube. The cases are exactly analogous to 
those of a body specifically heavier than water smking in it, 
and a body specifically lighter rising in it, by the action of 
gravity. Now it is true that the magnetic forces here assumed 
have the same directions as those which the variation of the 
pressure of the ether would of itself produce. But if the 
action were either wholly or chiefly due to this cause, no reason 
appears why the observed motions should depend, not upon the 
difference of the specific gravities of the fluids within and without 
the tube, but upon one being more or less ferruginous than the 
other. ‘The experiment, therefore, allows us to infer theoreti- 
eally, that etherial streams, flowing through ron, and affected 
by the number, size, and state of aggregation of its atoms, 
exert on each atom an accelerative force, which is in the direc- 
tion of the current where it is convergent, and in the direction 
contrary to that of the current where it is divergent. The 
atoms being by hypothesis spherical, it may be inferred from 
