Abstract
It is shown that, unlike the Oppenheimer-Brinkman-Kramers approximation a first-order capture theory satisfying Coulomb boundary conditions (B1B) yields remarkable agreement with experimental data, both for symmetric (p-H) and asymmetric (p-Ar) collisions in the intermediate velocity regime. This result suggests that a correct treatment of first-order capture ought to precede a detailed treatment of higher-order corrections.