Si II and Fe II lines in mean spectra at phases 0.15–0.25 (top row) and 0.65–0.75 (bottom row). The profiles of the Si II and Fe II lines clearly differ. Whereas there is only modest substructure in the Si II lines, there are two distinct components in the Fe II lines. The weaker component is likely stellar in origin while the deeper component has characteristics, such as a narrow width, that are suggestive of a circumsystem shell. The Fe II components are cleanly split near phase 0.7, and similar behavior was noted among many of the lines at blue wavelengths examined by D. M. Popper (1977); the exception was Mg II 4481.