Abstract
The dynamic or surface‐plasmon contribution to the surface energy of the electron gas derived by Craig is evaluated taking into account correctly the electron–hole excitations, dispersion of the plasmon and surface‐plasmon modes, and the damping of these modes by electron–hole excitations. The resulting surface energy vs. the density parameter is close to the earlier approximation which was criticized since all these influences were neglected there. The success of the simple approximation is explained using the simplified theory of electron correlations by Overhauser.