Based on observations obtained with the MPG/ESO 2.2-m telescope at La Silla Observatory.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

Tables 1-3 are only available in electronic form at http://www.edpsciences.org

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...

Full Table 2 is only available in electronic form at the CDS via anonymous ftp to cdsarc.u-strasbg.fr (130.79.128.5) or via http://cdsweb.u-strasbg.fr/cgi-bin/qcat?J/A+A/443/851

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... uncertainty

The reddening of the program stars can be quite large and varies from star to star, while the Landolt (1992) standards used have a very limited reddening range. See Selman (2004) for details.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... magnitudes

Here the fluxes are in ADUs, and are not extra-atmospheric. To use the above equations to reduce extra-atmospheric WFI magnitudes we need to correct by a ubv extinction per airmass of approximately 0.414, 0.179 and 0.119 respectively.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... to

Grebel et al. (1996), studied this effect and found a flattening of the IMF by 0.1-0.3 units in the exponent, depending on the fraction of Be stars and the age of the stellar population.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... gap

See Massey (2002) for a discussion on contamination by MW disk stars on different places of the HR diagram.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... universality

As stated in the introduction, we do not intend the previous statement to apply to Population III stars. It was shown by Bromm & Loeb (2003) that below a critical metallicity of $Z_{\rm crit}\sim10^{-3.5}~Z_\odot$ one can expect a top-heavy IMF (see also Abel et al. 2002). The Thomson optical depth $\tau_ {\rm e} = 0.17\pm 0.04$ derived from the polarization observations by the Wilkinson Microwave Anisotropy Probe (WMAP) appears to require a top-heavy IMF for Pop III stars (Kogut et al. 2003; Cen 2003), but see the dissenting views in Ciardi et al. (2003) and Tumlinson et al. (2004).

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.