40 
on the medium within, not merely from its superficial layer of molecules, 
but from all those lying within a certain distance of the surface in such 
a way that no one molecule of the containing vessel contributes more 
than an infinitesimal part to the effect on a molecule of the medium. Up 
to a certain density and pressure the medium will be contained by such 
a vessel; but as soon as the density is made to exceed this limit, the 
medium will begin freely to permeate the vessel, and escape. The re- 
marks made with reference to the last case apply equally to the present 
one, so long as the medium continues of sufficiently low density to be re- 
strained. The third case of limiting conditions arises where the medium 
is kept in by the wall of a containing vessel which it can neither press 
back nor penetrate. That this may be so, the outward motion of the 
superficial layer of molecules of the medium must meet with an absolute 
obstruction from the repulsion of the inner layer of molecules of the con- 
taining vessel. The forces thus coming into play would be at variance 
with hypothesis (4) ; on which account it is needless here to trace further 
the consequences of their action, several of which are, besides, sufficiently 
obvious. 
This kind of examination might, indeed, be continued almost inde- 
finitely, and extended to the crystalline texture of the medium and other 
branches of the subject; but enough has been done for my present 
end, which is to draw attention to the fact, that no known ponderable 
materval possesses a constitution of the kind which has been supposed, 
and, consequently, that we are forced to admit, as demonstrated, the 
following points with reference to this obscure branch of Physics :— 
1°. That no known ponderable matter consists of a continuous sub- 
stance similar in all its parts. 
2°. That by continuing the subdivision of solzds, as far as is s consistent 
with the parts being similar, similarly placed, and acting alike, either 
the whole, or at least a controlling part, of the action on each such mole- 
cule must be contributed by a finite number of those in its vicinity. 
The validity of this conclusion is not disturbed by attributing to the 
elementary molecules such motions as may constitute sound, heat, light, 
and electricity, and the statement is evidently a fortiori true of the parts 
of the elementary molecules. 
3°. If liquids are separable into elementary molecules of the kind 
described in the last paragraph, similar to one another, similarly placed 
and acting alike, then the statement of that paragraph may be repeated, 
word for word, with reference to them; but if, as seems perhaps more 
likely, there is going on between the elementary molecules of liquids a 
constant interchange of their parts, then the language must be somewhat 
modified, although the conclusion remains substantially the same. 
The case of gases is somewhat different. The examination which 
has been made above shows, no doubt, in the case of gases, as it does for 
either solids or liquids, that their constitution cannot be that of a sta- 
tionary system, each molecule of which acts, but to a trifling extent, on 
each of those around it. But it does not show that it may not be such ~ 
a system in a state of vivid molecular motion. The closer investigation 
