[en] It is suggested that the optical emission lines in spectra of X-ray binaries can originate from the surface of the nondegenerate member. This is an alternative to the picture in which the emission lines come from the vicinity of the compact member of the system. The emission lines are due to a temperature inversion in the photosphere which is induced by X-ray heating. The inverted phase of the emission velocity curves and the observed ''eclipse'' of the lines are explained. Individual systems are discussed. In some cases I have to assume a small soft X-ray flux. I also point out a few ways to test this suggestion observationally