Abstract
The Brillouin scattering efficiencies of exciton polaritons in CdS have been calculated with two different models. The first model, which assumes Pekar's additional boundary condition (ABC) results in qualitative disagreement with experiment. The second model employs the ABC proposed by Birman and Sein, by Agarwal et al., and by Maradudin and Mills and assumes a lossy layer on the sample surface. This model results in quantitative agreement between theory and experiment.