A modification of the standard theory of stimulated Raman scattering (SRS) first proposed by Sparks (1974, 1975) is analyzed and shown to incorporate a possibly important physical effect; however, its original formulation is incorrect. The analysis is based on an exact numerical integration of the coupled equations of the modified theory, the results of which are compared with both the conventional theory of SRS and with one set of experimental data. A reformulation of the modified theory is suggested that leads to a gain which is in somewhat better agreement with the data than is the conventional theory.