Annu. Rev. Astron. Astrophys. 1997. 35: 357-88 Copyright © 1997 by Annual Reviews. All rights reserved

1. INTRODUCTION

As their name suggests, compact groups are small systems of several galaxies in a compact configuration on the sky. The first example was found over one hundred years ago by Stephan (1877) who observed it visually using the 80-cm Foucault telescope of the Observatoire de Marseille. Stefan's Quintet is a small group of five galaxies, three of which show strong tidal distortions due to gravitational interaction. A second example was found 71 years later by Seyfert (1948) from a study of Harvard Schmidt plates. Seyfert's Sextet (Figure 1) is one of the densest groups known, having a median projected galaxy separation of only 6.8h^-1 kpc (the Hubble Constant H₀ = 100h km s^-1 Mpc^-1).

The Palomar Observatory Sky Survey (POSS) provided a new and extensive resource for the systematic investigation of small groups of galaxies. Two catalogs, the Atlas of Interacting Galaxies (Vorontsov-Velyaminov 1959, 1975) and the Atlas of Peculiar Galaxies (Arp 1966), contain galaxies or galaxy groups selected on the basis of visible signs of interaction or peculiar appearance. In addition to Stefan's Quintet and Seyfert's Sextet, these include many new compact groups, including a striking chain of five galaxies, VV 172. Prior to these, Shakhbazian (1957) had discovered a small dense cluster of 12 faint red galaxies that appeared so compact that they were initially mistaken for stars. Over the next two decades, Shakhbazian and collaborators examined over 200 POSS prints covering 18% of the sky and cataloged 376 additional "compact groups of compact galaxies" (Shakhbazian 1973, Shakhbazian & Petrosian 1974, Baier et al. 1974, Petrosian 1974, 1978, Baier & Tiersch 1975, 1976, 1978, 79). Apart from occasional photographic or spectroscopic observations (eg. Mirzoian et al. 1975, Tiersch 1976, Massey 1977, Shakhbazian & Amirkhanian 1979, Vorontsov-Velyaminov et al. 1980, Vorontsov-Velyaminov & Metlov 1980) these systems initially received little attention. However, interest in them is growing. Although the majority seem to be small clusters, they share some of the properties, and pose some of the same questions, as do compact groups.

When redshifts were measured for galaxies in the first compact groups (Burbidge & Burbidge 1959, 1961a), surprises were found. Both Stefan's Quintet and Seyfert's Sextet contain a galaxy with a discordant redshift. It seemed unlikely that a foreground or background galaxy would appear so often projected within such compact systems (Burbidge & Burbidge 1961a). This impression was further reinforced with the discovery of yet another discordant redshift in VV 172 (Sargent 1968). Are these examples of physical association between objects of widely different redshifts, as has been advocated for many years by Arp (1987)?

Figure 1. Seyfert's Sextet. Discovered in 1948, this group of five galaxies is one of the densest known. The sixth object appears to be a tidal plume. The small face-on spiral galaxy has a redshift that is more than four times larger than those of the other galaxies.

Even if the discordant galaxies are ignored, the velocity dispersions of these systems are generally higher than would be expected given their visible mass (Burbidge & Burbidge 1959, 1961b, 1961c, Burbidge & Sargent 1971). It was argued that such groups must be unbound and disrupting (eg. Ambartsumian 1961), although Limber and Mathews (1960) showed that the virial theorem could be satisfied for Stefan's Quintet, given the uncertainties in the projection factors, if the individual galaxy masses were considerably larger than those of isolated galaxies. The observations can of course also be explained if the bulk of the mass is in a non-visible form. In hindsight, this was one of the earliest indications of the possible existence of dark matter in galactic systems.

A new problem emerged with the realization that bound groups would be unstable to orbital decay resulting from gravitational relaxation processes (Peebles 1971). A simple calculation indicated that the dynamical-friction timescale was much shorter than the Hubble time (Hickson et al. 1977), as was soon confirmed by numerical simulations (Carnevali et al. 1981). At the same time, it became increasingly clear that mergers played an important role in the evolution of many, if not all, galaxies (Press & Schechter 1974, Ostriker & Tremaine 1975). Compact groups emerged as prime locations for investigations of the dynamical evolution of galaxies.

Motivated by the desire for a homogeneous sample that could be subject to statistical analysis, Rose (1977) and, later, Hickson (1982) produced the first catalogs of compact groups having specific, quantitative, selection criteria. Subsequent detailed investigation, at many wavelengths, has produced a large body of observational data for the Hickson catalog. As a result, it has now become possible to address some of the outstanding questions concerning the nature of compact groups and their role in galaxy evolution. Not surprisingly, new questions have been raised and new controversies have appeared. However, much progress has been made in resolving both old and new issues.

This review is organized as follows: In Section 2, the definition of a compact group is discussed, along with methods of identification and surveys that have been made. In Sections 3, 4, 5, observed properties of these systems are summarized and discussed. Sections 6 and 7 focus mainly on interpretation of the observations, and on implications of these results. All work on compact groups of galaxies cannot possibly be discussed in this short paper, although an attempt is made to touch upon most current topics. Other recent reviews of compact groups and closely related subjects include those by White (1990), Hickson (1990, 1997), Whitmore (1992), Kiseleva & Orlov (1993), Sulentic (1993), and Mamon (1995).