1 Introduction

The Hickson compact group of galaxies 31 (HCG 31; Hickson 1982) has recently proven an interesting laboratory for studying various issues related to galaxy formation and merging. HCG 31 is physically compact. Six of its members fall within the dimensions of a typical giant galaxy, share a common H I envelope (Williams et al. 1991), and span a small range in radial velocities (Rubin et al. 1990; Hickson et al. 1992). It is therefore no surprise that HCG 31 shows a wide array of evidence for galaxy interaction including: tidal tails and irregular morphology, complex kinematics, and vigourous star formation, including starbursts with Wolf-Rayet spectral features (Rubin et al. 1990; Iglesias-Páramo & Vílchez 1997; Izotov & Thuan 1998; Johnson et al. 1999; Johnson & Conti 2000).

Rubin et al. (1990) conclude that HCG 31 will appear as a single normal elliptical galaxy within a few orbital periods. Here, we differ somewhat from this conclusion. Although we find abundant kinematical evidence for recent gravitational interaction among the group members, we suggest that there are other explanations besides imminent coalescence. Likewise, we find chemical abundances for the different components that are unlike those of normal dwarf irregulars, but again argue that their origins may be easily understood. Given the abundant evidence for gravitational interaction, a merger may yet occur in HCG 31, but we find that it need not be currently underway.

In the following section, we present our data and their reduction. In Sect. 3, we present our results. In Sect. 4, we discuss these results and their implications for the state of HCG 31. In Sect. 5, we present our interpretation of the past and future evolution of HCG 31. In Sect. 6, we summarize our conclusions. Throughout, we adopt the galaxy nomenclature of Rubin et al. (1990) and we assume a distance of 58 Mpc for HCG 31, based upon a mean redshift of 0.0137 (Hickson et al. 1992) and a Hubble constant of 71 km s^-1 Mpc^-1(Mould et al. 2000). A preliminary version of some of our results was presented by Rosado et al. (1999).

Figure 1: The field of HCG 31 in the V-band. Galaxies A, B, C, D, E, and F are shown; galaxies G and Q are to the south-east and north, respectively, as indicated by the arrows. Deeper, higher-resolution images may be found in Rubin et al. (1990), Iglesias-Páramo & Vílchez (1997), Johnson et al. (1999), and Johnson & Conti (2000). The ellipses are meant to illustrate schematically the positions of galaxies A, C, E, F and the tidal dwarf candidates (denoted by T"; Hunsberger et al. 1996). Objects within the component galaxies are labelled using the galaxy component letter and a sequence number, e.g., F1 and F2 are subcomponents of galaxy F. Where sub-components are labelled, the parent galaxies are not, to reduce the clutter. North is up and east is to the left, as indicated, and $1\hbox {$^{\prime \prime }$ }= 281$ pc. Unlabelled objects are foreground stars, except that north-east of galaxy D, which is a reflection. The filled circles indicate the locations of the apertures used for multi-object spectroscopy.