... clusters[*]
Based on data collected at the European Southern Observatory, Chile.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
... lines[*]
We only used H$_\alpha$. H$_\beta$ was not observed; the other members of the Balmer series were indeed observed, but they are more contamined by metal lines, and their dependence on gravity, metal abundance and details in the convection much larger than for H$_\alpha$ (see Fuhrmann et al. 1993).
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
... subgiants[*]
The errors for $T_{\rm eff}$'s we quote for cluster stars are smaller than those we give for - individual - field stars, in spite of the lower S/N values. This is because errors in individual $T_{\rm eff}$'s are due to the flat fielding, and are then only marginally dependent on S/N; on the other side, temperatures for cluster stars are obtained averaging over the values obtained for all similar stars: for instance, in the case of NGC 6752 TO-stars, the adopted $T_{\rm eff}$ is the average of the values obtained for nine individual stars. Star-to-star scatter in these $T_{\rm eff}$'s is similar to the error bar we obtain for individual field star.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Copyright ESO 2001