Terminal velocities for W-R stars can be obtained from ultraviolet P--Cygni absorption profiles (Prinja et al. 1990) or from infrared HeI profiles (Howarth & Schmutz 1992). Recent studies have shown that there exists a good correlation between terminal velocity and stellar temperature (see Table 1) although no significant differences are found between Galactic and LMC stars (Crowther & Smith 1996b).
Figure 1: Comparison of wind performance numbers (see text) for Galactic (open) and LMC (filled-in) WNL stars and LBVs versus surface hydrogen abundance (by mass) adapted from Crowther & Smith (1996b). The single scattering limit is shown (dotted-line), while a clear distinction in wind density is apparent between hydrogen-rich and hydrogen-poor WNL stars. Lower wind velocities and mass-loss rates for LMC stars, anticipated from radiation driven wind theory are not observed
Observationally, mass-loss rates of W-R stars can be determined from radio emission fluxes, assuming spherical symmetry and homogeneity, provided stellar wind velocities, chemical abundances and the outer wind ionization balance are known. However, the difficulty in obtaining radio observations of most W-R stars, coupled with uncertain chemistries and ionization meant that progress was slow prior to the Standard Model. With few exceptions, log ( /yr )-5 to -4, with no obvious relation to spectral type, binarity or metallicity. The theoretically expected relation between mass-loss and luminosity has been tentatively confirmed (Hamann et al. 1995), at least for those without hydrogen.
Winds of O--stars have mass-loss rates that are reasonably explained by radiatively driven wind theory (Puls et al. 1996). The wind performance number /(L/c) indicates the efficiency at which radiation momentum is imparted to the wind, and is equal to unity in the `single scattering limit' (i.e. scattering of a photon in a single line). In contrast, W-R stars often have performance numbers substantially greater than unity as shown in Fig. 1 for Galactic and LMC WNL stars, causing severe difficulties for radiatively driven wind theory, unless multiple scattering is achieved (see Springmann 1994). Hillier (1996) discusses various options which could solve this problem, while other driving mechanisms have also been proposed (e.g. pulsational instabilities, Glatzel et al. 1993, Langer et al. 1994). Although current radiation driven wind theory predicts lower wind velocities and mass-loss rates at lower metallicity, these are not observed for LMC stars, while there is also a significant correlation between wind performance number and helium enrichment (Fig. 1).