6 Concluding remarks

In the present paper we studied a sample of reddened bulge globular clusters (and candidates) as well as some projected on the disk, by means of integrated near-infrared spectroscopy. The results, in conjunction with other available information from different observational techniques, especially recent CMDs, shed new light on their nature and properties. In particular we try to classify them into families.

Armandroff (1989) indicated a sample of 19 bona-fide metal-rich globular clusters, most of them towards the central regions of the Galaxy and not much reddened. To that list, the following 9 bulge clusters have been conclusively added: Terzan2, NGC6380, Tonantzintla2, Palomar6, Djorgovski1, Terzan5, Terzan6, ESO456-SC38 and Terzan12 (with coordinates $\alpha_{1950}=18^{\rm h}~09^{\rm m}~14.0^{\rm s}$ and $\delta_{1950}=-22\mbox{$^{\circ}$}~45\mbox{$^{\prime}$}~18\mbox{$^{\prime\prime}$}$). In addition, Palomar8, already in Armandroff's list has its properties confirmed in the present study. Liller1 and Terzan1 are certainly members of the metal-rich bulge family, possibly among the most metal-rich ones; however, owing to crowding, contamination and reddening effects, deeper observations are necessary.

NGC6256, HP1 and NGC6717 belong to an intermediate metallicity family in the bulge with $\mbox{$[Z/Z_{\odot}]$}=-1.0$ to -1.5 with prominent blue-extended HBs in post-core collapse clusters, similar to NGC6522 (Barbuy et al. 1994) and NGC6540 (Bica et al. 1994). Terzan9 probably belongs to this family, but deep CMDs are necessary to check. Terzan10 is probably of intermediate metallicity but is not in a post-core collapse phase, and also requires deep CMDs for more conclusive results.

The integrated light of Terzan4 is heavily contaminated by field stars; its nature appears to be that of a metal-poor globular cluster as recently unveiled by a deep CMD (Ortolani et al. 1997).

Although previous studies of the difficult globular cluster UKS1 pointed to an intermediate metallicity, the present result indicates a metal-rich nature, unless the integrated light is contaminated by metal-rich field stars with reddening comparable to that of the cluster itself. The object requires deep CMDs to establish its properties.

From the present spectroscopic results, the object TBJ3 = TJ13 is reinstated as a globular cluster candidate and a deep CMD is required to establish its nature. It is expected that the dynamical evolution of globular clusters produces some post-core collapse fossils, after evaporation of their halos. TBJ3 is the best candidate to be an example of such end products of cluster evolution.

For the clusters BH176, Lyngå7 and Palomar10, which are projected on the disk, the present spectroscopic results show that they have a nearly-solar metallicity consistent with recent CMD studies which point them as metal-rich globular clusters. Such kind of object might be a link to the oldest solar-metallicity open clusters, like NGC6791 ($l=+70.0\mbox{$^{\circ}$}$, $b=+11.0\mbox{$^{\circ}$}$) with age $\approx8$Gyr (Demarque et al. 1992). Finally, ESO93SC-08 and UKS2 are intermediate age open clusters, and the integrated spectroscopic results point to a subsolar metallicity.

We point out that quite often CMDs and integrated studies of reddened clusters in crowded fields show evidence of stellar population mixtures. The first interpretation in most cases would be contamination by field stars. Nevertheless, alternative interpretations like the possibility of mergers between globular clusters (Catelan 1997), and capture of field stars by globular clusters in dense bulge fields (Bica et al. 1997), might be explored by means of deep CMDs, velocities and abundance determinations of individual giants as membership criteria.

Acknowledgements

We thank the staff and personnel at CASLEO for hospitality and assistance during the observations. We acknowledge the Antorcha and VITAE foundations, the Argentinian institutions CONICOR and CONICET, and the Brazilian ones CNPq and FINEP for support.