566. MOSSOTTI ON CAPILLARY PHA NOMENA. 
dicular to it, and whose height is equal to the distance at which — 
the molecular forces act, if the liquid be not acted on by any 
external pressure, its molecules are situated at such a distance 
from one another that the sum of the respective repulsions of 
the molecules upon the opposite side of the plane on those of the 
nearer portions of the small prism, is exactly equal to the sum 
of the respective attractions of the same molecules of the fluid 
upon the opposite side of the plane on those of the further por- 
tions of the small prism: hence the small prism has no tendency 
either to press on the plane or to move from it, and the fluid 
throughout is in equilibrium and exerts no pressure. This holds 
for every portion of the fluid situated at a distance from the sur- 
face greater than that to which the molecular forces extend ; but 
if we suppose a plane intersecting the fluid parallel to its surface, 
which we shall now suppose horizontal and infinite, at a depth 
less than the distance of molecular action, and if we suppose a 
small prism perpendicular to the plane upon the side towards the 
external surface, this prism not being of sufficient height, will 
not contain a sufficient number of more distant molecules to 
counteract the repulsive action of those nearest the plane, conse- 
quently there will be an excess of repulsion on the former mo- 
lecules, and they will tend to separate from each other. The 
separation of the molecules will be so much greater as the plane 
is nearer the surface of the liquid, so that on approaching this 
surface we shall find a rapid decrease in density, regulated by 
the law that the repulsive action of the fluid beneath the plane 
on the molecules of the portion of the prism between the plane 
and the surface, should be always counteracted by the attraction 
of the parts reciprocally more distant, so that the pressure re- 
main zero for every plane. 
The depth of the stratum in which this rapid decrease in den- 
sity will take place will be very small, since the molecular action 
extends only to inappreciable distances; but we may suppose 
it divided into several very thin strata, in each of which the 
density may be considered uniform, that is the molecules may 
be considered equidistant one from the other. Now while near 
the surface the equilisrium of the molecules in a vertical direc- 
tion requires that the density of the fluid shall decrease rapidly, 
the equilibrium in a horizontal direction will remain unaltered, 
although the molecules are distributed with a uniform density in 
