We computed the line profiles and wavelength positions of all the important emission lines due to carbon for a variety of steady state siphon flow loop models. We have used the models calculated by Noci et al. and have taken their results for the nonequilibrium ionization densities. We found for the lines from the lower ionization states (C II-C IV) a preponderance of blueshifts, contrary to the observations. The lines from the higher ionization states can show either a net red- or blueshift, depending on the position of the loop on the solar disk. We expect similar results for oxygen. We conclude that the observed redshifts cannot be explained by our models.