THE ELECTRIC AND LUMINIFEROUS MEDIUM. 
259 
The point, however, to be noticed is that this expression for the toted traction on an 
interface, due to both molar and molecular forcive, which makes the pull on an 
interface lying normal to the lines of polarization twice as great as the push on one 
tangential to those lines, holds only for the case in v/hich the material is polarized in 
the same direction throughout the whole extent of its volume. We may estimate by 
Itself the action of the surrouncling portion only, extending to any distance we please , 
but the action of the remaining outside part of the merlinm will still involve that of 
an inner surface density of uncompensated poles which will remain of undiminlshed 
order of magnitude. This procedure by separate molecules is thus not suitable tor 
discrimination between the forcive due to the medium as a whole, which is trans¬ 
mitted, and the molecular forcive which is compensated locally. 
46. The justification of the theory here applied, which balances on the spot the 
molecular part of the forcive due to the electric polarization, by an Intrinsic cohesive 
stress in the material which is independent of the material elastic constants and 
strains at the place, may be further enforced by consideration of the Ideally simple 
case of a gas. If a system of bodily forces act on it from a distance, they can always 
be balanced by a simple increase of pressure when they are derived from a potential 
function; while if they were not so derived the medium could not be in equilibrium. 
The dual phenomenon of equilibrium of the element of volume maintained by a 
balance between two forcives, an extraneous and an internal one, is really a balance 
between a forcive on the element of volume acting from a distance by the mediation 
of the mther, and another forcive arising according to the explanations of the theory 
of gases from the impacts of the molecules surrounding the element and, in the case 
of dense media, in part also from cohesive molecular actions. It is in this case a 
balance between a static bodily forcive and a steady kinetic molecular one; if the 
force transmitted through the rether from a distance increases, and equilibrium is to 
be maintained, the molecular configuration must be adjusted so that the impacts and 
the local molecular attractions shall continue to preserve the balance. When to the 
forces actino’ from a distance are added coordinated electric attractions between the 
molecules of the polarized medium, a further adjustment of molecular configuration 
must ensue. Now when the gas is electrically polarized, the attractions between 
neighbouring molecules give a forcive, not isotropic like a fluid pressure, but depend¬ 
ing' on the direction of polarization ; its action w^ill thus alter the originally fortuitous 
arrangement of the velocities of the molecules ot the gas so as to impart to then 
distribution a slightly axial character, and when this has resulted in a new steady 
state the pressure due to the impacts will be different according to the manner in 
which the element of interface that is pressed is related to the line of polarization. 
This molecular addition of an intrinsic local stress, which has not the character of the 
ordinary fluid pressure, will just balance the action of the local electric attractions 
when the state of the system has again become steady ; and being thus itself com¬ 
pletely compensated locally, there will remain nothing of the molecular part of the 
2 L 2 
