A simple analytical expression, which closely approximates the Coulomb potential between two uniformly charged spheres, is presented. This expression can be used in the optical potential semiclassical analyses which require that the interaction be analytic on and near the real r-axis. PACS number(s): 24.10. Ht, 25.70.Bc, 03.65.Sq The semiclassical analyses of the optical potential cross sections play an important role in understanding the scattering mechanism of light-and heavy-ions [1]. The direct evaluation of the semiclassical amplitudes is based on the calculation of the integral actions between the turning points of the radial motion. The turning points of the radial motion are solution of the equationwhere k, λ, E, and U (r) are, respectively, the wave number, the angular momentum in units of , the center of mass energy, and the complex optical potential. The integral actions are defined bywhere r i and r j are appropriate solutions of Eq. (1). In the semiclassical approximations both these quantities are treated as analytical functions of λ around the real λ-axis. This is true only if the complex optical potential is an analytical function of r, on and near the real r-axis.In the phenomenological analyses the optical potential between heavy-ions is written in the formwhere V (r) and W (r) are the real and the imaginary part of the nuclear interaction, and V C (r) is the Coulomb interaction potential. As a rule, analytic functions are used for the real and the imaginary part of the interaction. On the contrary, the Coulomb part of the interaction is described using either the Coulomb potential of one point charged particle with a uniformly charged sphere of radius R C (in the least recent works) or the Coulomb potential between two uniformly charged spheres of radii, say, R 1 and R 2 (in the more recent ones). These potentials are not analytic functions around the real r-axis.This fact, rigorously, excludes the possibility of applying the semiclassical methods which assumes the analyticity of the potential. The lack of analyticity must however be considered accidental. It arises from our preference to use simple expressions for the interaction, and we do not really think that it is connected with some physical fact. Owing to this we expect that this, and similar troubles, can be cured with a little effort.To do this we here briefly recall some techniques used in the past to treat the point-sphere case, and we present a similar method for the sphere-sphere case.Both these Coulomb potentials can be written in the form