1 Introduction

The determination of globular cluster parameters such as reddening and chemical abundances is fundamental to understand their spatial and metallicity distributions, which in turn can be used to study the formation conditions and evolution of the Galaxy (for a recent discussion see van den Bergh 1996). The first step is to determine whether the star cluster is indeed a globular cluster, since the nature of several objects is still unclear. Several compilations provide coordinates, cross-references and available basic data on globular clusters and candidates, e.g. Webbink (1985); Djorgovski & Meylan (1993) and Harris (1996).

Efforts to obtain such parameters have included colour-magnitude diagrams (CMD) and integrated photometry and spectroscopy, see e.g. the compilation of CMDs by Peterson (1986), the integrated photometry by Zinn (1980) and Bica & Pastoriza (1983), the photographic spectroscopy by Zinn & West (1984, hereafter ZW84), and the scanner spectroscopy by Bica & Alloin (1986a). Spectroscopic studies of individual giants in globular clusters provide fundamental abundance determinations, see e.g. the recent results for the metal-poor end of the metallicity scale by Geisler et al. (1995) and Minniti et al. (1996). All these studies have concentrated on the less reddened clusters. The more reddened globular clusters in crowded fields, found in the bulge and low-latitude disk directions, have been studied in a more systematic way only recently. In terms of optical CMDs, several good-quality seeing studies with CCD imaging are providing reliable cluster parameters for difficult objects, e.g. Terzan5 and HP1 in the bulge (Ortolani et al. 1996; Ortolani et al. 1997b), Lyngå7 and Palomar10 in the disk (Ortolani et al. 1993b; Kaisler et al. 1997). Also in the infrared, where reddening effects are minimised, efforts are being undertaken to explore such clusters by means of CMDs, e.g. Minniti et al. (1995); Frogel et al. (1995) and Guarnieri et al. (1998).

For globular clusters more reddened than ${E(B-V)}\approx1.0$, the visible flux becomes too low to provide useful spectra and longer wavelengths should be observed. Bica & Alloin (1987, hereafter BA87) carried out near-infrared CCD integrated spectroscopy of globular clusters in a wide range of metallicity employing the CaII triplet as metallicity indicator. Armandroff & Zinn (1988, hereafter AZ88) carried out near-infrared CCD spectroscopy for a sample of reddened bulge clusters. They derived metallicities from the CaII triplet and estimated reddening from an interstellar band at $\lambda8621$Å. However, they have not estimated reddening from the continuum distribution, since the spectra were not flux calibrated.

The goal of the present paper is to derive reddening and metallicity for a sample of 20 clusters projected on the bulge ($\vert l\vert<13\mbox{$^{\circ}$}$, $\vert b\vert<11\mbox{$^{\circ}$}$) and 5 others projected on the disk ($\vert l\vert\gt 30\mbox{$^{\circ}$}$ and $\vert b\vert<5\mbox{$^{\circ}$}$), by means of flux-calibrated near-infrared spectra. We have 8 clusters in common with AZ88's sample. The objects studied here are part of a systematic spectroscopic survey of globular clusters (and candidates). The first results dealt with 3 candidates which turned out to be emission nebulae (Bica et al. 1995). These objects are: TJ5 and TJ23 (Terzan & Ju 1980) which are planetary nebulae, and the one found by Bica (1994) which appears to be a supernova remnant.

This paper is structured as follows: in Sect.2 we present the observations. In Sect.3 we derive reddening from the continuum distribution as compared to reddening-free globular cluster template spectra, and metallicity from equivalent widths (W) of the CaII triplet. Individual objects are discussed in Sects.4 and 5 respectively for the bulge clusters and the ones projected on the disk. Finally, the conclusions of this work are given in Sect.6.