Fine-Tuning the Wilson-Bappu Effect

We have used the re-reduced Hipparcos parallaxes and extensive new observations from various sources to investigate the details of the Ca II emission line width vs. M-nu relation first noted in Wilson & Vainu Bappu (1957). Only stars whose parallaxes exceed their probably error by at least a factor of 5 were used. The full range of M-nu now extends from M-nu =+ 14 to -5. Small but well defined deviations from a linear relationship are clearly present. The data have also been separately plotted for each luminosity class from V to Ib. Main sequence stars show significant curvature from M-nu = 14 to 8. Stars of class III show a very nice linear curve. The supergiants, almost all of which are of class Ib, show the greatest scatter as originally noted by Wallerstein et al. (1999). For those stars the Wilson-Bappu Effect is not a useful method for deriving M-nu. We discus the shape of the dispersion in log W-0 at a given M-nu and show that it is asymmetric. Individual measurements of log W-0 are more likely to be above the mean rather than below it. This is probably due to intermittent activity that enhances the width of the Ca II emission line.