Quantitative studies of hot luminous stars in galaxies are important for a number of reasons. First, and probably foremost, is the information they provide on the effect of the environment on such fundamental properties as the mass-loss rate and stellar evolution. In the standard picture (e.g. Maeder & Meynet 1987) mass-loss significantly affects the evolution of massive O stars as they proceed through an Of phase, then an unstable Luminous Blue Variable (LBV) phase, before becoming late WN (WNL) stars. The current theory of driving mass-loss through radiation pressure predicts mass-loss to scale as the square-root of the metallicity (Kudritzki et al. 1989). The evolution of massive stars should therefore be strongly dependent on the local environment (Maeder 1991; Maeder & Meynet 1994). Thus studies of massive stars at different metallicities should enable a direct comparison of observations with evolutionary and radiation driven wind theory. The question of exactly how mass-loss scales with metallicity is of enormous importance. For example, it affects the early chemical enrichment of galaxies, and the evolution of starbursts containing many thousands of O and Wolf-Rayet (WR) stars (e.g. Leitherer & Heckman 1995).
In this series of papers we have studied the physical and chemical nature of predominantly Galactic Wolf-Rayet stars by combining detailed model atmosphere calculations with high quality, multi-wavelength observations. In particular, we have investigated WNL stars (Crowther et al. 1995a, b, c, hereafter Papers I-III), weak-lined, early WN (WNE) stars (Crowther et al. 1995d) and intermediate WN/C stars (Crowther et al. 1995e). Other related studies have been conducted on LBVs (L.J. Smith et al. 1994), Of stars (Crowther & Bohannan 1997), an M33 WR star (L.J. Smith et al. 1995), and strong-lined WNE stars in the infrared (Crowther & Smith 1996). WC and WNE stars have not been widely investigated to date since we anticipate that their parameters will be the most susceptible to the effects of line blanketing, given their dense winds and high excitation.
In this, the final paper of the current series, we present a study of WNL stars in the Large Magellanic Cloud (LMC). These stars have the distinct advantage over their Galactic counterparts in that they lie at a known distance and suffer only moderate interstellar extinction. It is thus much easier to determine reliable absolute fluxes and mass-loss rates. A direct comparison with Galactic WNL stars (Papers I-III) will allow us to investigate any differences in fundamental properties and evolution, as a result of the lower metallicity of the LMC, by a factor of 2-3 (Spite & Spite 1991). The first quantitative comparison of the properties of LMC and Galactic WR stars was performed by L.J. Smith & Willis (1983) who found no differences except that the LMC WN stars had stronger He II emission lines. More recently, Koesterke et al. (1991) suggested that, while the LMC WN stars showed no clear differences from their Galactic counterparts in mass-loss rates or wind velocities, they do have lower stellar luminosities. Nevertheless, it is well known that our Galaxy and the LMC exhibit quite different WN:WC ratios; in the solar neighbourhood this ratio is 1:1 and in the LMC it is 5:1. It has been suggested that this difference may be directly attributed to metallicity effects (e.g. L.F. Smith 1991). It is also interesting to note that the frequency distribution of WN stars among the subtypes is different between the two galaxies; in the Galaxy there is a broad peak between WN5-6 whereas the LMC shows a sharp peak at WN4 (L.F. Smith et al. 1996).
Our previous Galactic studies (Papers I-III; Crowther & Bohannan 1997) have led to the discovery of different evolutionary routes for the formation of WNL stars. In particular, we find that very massive early Of stars ( ) evolve to WN6-7 stars with an Ofpe or WNL+abs (hereafter WNLha following L.F. Smith et al. 1996) intermediate stage. For less massive progenitors, evolution instead proceeds through an intermediate LBV stage to the WN8 spectral type, with WN9-11 stars representing either dormant or post-LBVs. This evolutionary scenario naturally explains the observed dichotomy among WNL stars (Moffat 1989). However, this scenario is based principally on studies of Galactic objects, and it is therefore of interest to examine if it is viable in lower metallicity environments such as the LMC.
In Sects. 2-3 we present and discuss new ultraviolet and optical observations of a substantial sample of LMC WNL stars. The fundamental parameters of our sample are derived and discussed in Sect. 4. Finally in Sect. 5, we compare the results of this analysis with our Galactic sample and discuss the implications of our results for radiatively-driven winds and evolutionary theories of massive stars.
© European Southern Observatory (ESO) 1997
Online publication: June 30, 1998