Abstract
The real part of the optical-model potential for 35 MeV protons on target with mass number 10 is calculated from several versions of the nuclear matter approach. All these versions yield good agreement with the dependence upon A of the volume integral of the empirical optical potential, but only several of these versions reproduce the dependence upon A of its root mean square radius. The versions which predict a sizable increase of the potential radius and diffuseness with increasing A are those which take into account the inaccuracy of the nuclear matter approach in the tail of the potential surface. The agreement with the empirical values of the potential radius and diffuseness is best when one introduces between the radial shapes of the proton and neutron density distributions a difference similar to the one predicted by recent theoretical models. The potential radius and diffuseness are also calculated for 11 MeV protons.