Let us here briefly note that surface tension is, in itself, a comparatively small force, and easily measurable: for instance that of water is equivalent to but a few grains per linear inch, or a few grammes per metre. But this small tension, when it exists in a _curved_ surface of very great curvature, gives rise to a very great pressure directed towards the centre of curvature. We can easily calculate this pressure, and so satisfy ourselves that, when the radius of curvature is of molecular dimensions, the {207} pressure is of the magnitude of thousands of atmospheres,—a conclusion which is supported by other physical considerations. The contraction of a liquid surface and other phenomena of surface tension involve the doing of work, and the power to do work is what we call energy. It is obvious, in such a simple case as we have just considered, that the whole energy of the system is diffused throughout its molecules; but of this whole stock of energy it is only that part which comes into play at or very near to the surface which normally manifests itself in work, and hence we may speak (though the term is open to some objections) of a specific _surface energy_. The consideration of surface energy, and of the manner in which its amount is increased and multiplied by the multiplication of surfaces due to the subdivision of the organism into cells, is of the highest importance to the physiologist; and even the morphologist cannot wholly pass it by, if he desires to study the form of the cell in its relation to the phenomena of surface tension or “capillarity.” The case has been set forth with the utmost possible lucidity by Tait and by Clerk Maxwell, on whose teaching the following paragraphs are based: they having based their teaching upon that of Gauss,—who rested on Laplace. Let _E_ be the whole potential energy of a mass _M_ of liquid; let _e__{0} be the energy per unit mass of the interior liquid (we may call it the _internal energy_); and let _e_ be the energy per unit mass for a layer of the skin, of surface _S_, of thickness _t_, and density ρ (_e_ being what we call the _surface energy_). It is obvious that the total energy consists of the internal _plus_ the surface energy, and that the former is distributed through the whole mass, minus its surface layers. That is to say, in mathematical language, _E_ = (_M_ − _S_ ⋅ Σ _t_ ρ) _e__{0} + _S_ ⋅ Σ _t_ ρ _e_. But this is equivalent to writing: = _M_ _e__{0} + _S_ ⋅ Σ _t_ ρ(_e_ − _e__{0}); and this is as much as to say that the total energy of the system may be taken to consist of two portions, one uniform throughout the whole mass, and another, which is proportional on the one hand to the amount of surface, and on the other hand is proportional to the difference between _e_ and _e__{0}, that is to say to the difference between the unit values of the internal and the surface energy. {208}