Comparison of synthetic spectra of a pure carbon atmosphere models (blue line) with a ﹩\mathrm{log}\,( \mathrm{C}\,/ \mathrm{He}\,) =-3.0﹩ model (red line) for ﹩T_{\mathrm{eff}\,}=10,000﹩ K and ﹩\mathrm{log}\,g=8.0﹩. Spectra with extremely strong carbon bands similar to that predicted in the pure carbon model should have been easily noticed in SDSS if they existed.