Annu. Rev. Astron. Astrophys. 2000. 38: 289-335 Copyright © 2000 by Annual Reviews. All rights reserved

1. INTRODUCTION

Redshift surveys of the nearby universe indicate that most galaxies occur in small groups (e.g. Holmberg 1950, Humason, Mayall & Sandage 1956, de Vaucouleurs 1965, Materne 1979, Huchra & Geller 1982, Geller & Huchra 1983, Tully 1987, Nolthenius & White 1987). Despite diligent work in this area over the last two decades, the nature of poor groups is still unclear. Dynamical studies of groups are generally hampered by small number statistics: a typical group contains only a few luminous galaxies. For this reason, the dynamical properties of any individual group are always rather uncertain. In fact, many cataloged groups may not be real physical systems at all (e.g. Hernquist et al 1995, Frederic 1995, Ramella et al 1997), but rather chance superpositions or large-scale structure filaments viewed edge-on. Given the small number of luminous galaxies in a group, the prospects for uncovering the nature of these systems from studying the galaxies alone seem rather bleak.

The discovery that many groups are X-ray sources has provided considerable new insight into these important systems. X-ray observations indicate that about half of all poor groups are luminous X-ray sources. In many cases, the X-ray emission is extended, often beyond the optical extent of the group. X-ray spectroscopy suggests the emission mechanism is a combination of thermal bremsstrahlung and line emission from highly ionized trace elements. The spatial and spectral properties of the X-ray emission suggest the entire volume of groups is filled with hot, low-density gas. This gas component is referred to as the intragroup medium, in analogy to the diffuse X-ray emitting intracluster medium found in rich clusters (e.g. Forman & Jones 1982).

To first order, groups can be viewed as scaled-down versions of rich clusters. Many of the fundamental properties of groups, such as X-ray luminosity and temperature, are roughly what one expects for a "cluster" with a velocity dispersion of several hundred kilometers per second. However, some important physical differences exist between groups and clusters. The velocity dispersions of groups are comparable to the velocity dispersions of individual galaxies. Therefore, some processes such as galaxy-galaxy merging are much more prevalent in groups than in clusters. Other mechanisms that are important in the cluster environment, such as ram-pressure stripping and galaxy harassment, are not expected to be important in groups. The spectral nature of the X-ray emission is also somewhat different in groups than in clusters. At the typical temperature of the intracluster medium, almost all abundant elements are fully ionized, and the X-ray emission is dominated by a thermal bremsstrahlung continuum. At the lower temperatures of groups, most of the trace elements retain a few atomic electrons, and line emission dominates the observed X-ray spectrum. Thus, while the cluster analogy is a useful starting point, detailed studies of groups as a class are also important. Although no strict criterion exists for separating groups from poor clusters, for the context of this article I focus on systems with velocity dispersions less than about 500 km/s.

The idea that poor groups might contain diffuse hot gas dates back to the classic Kahn & Woltjer (1959) paper on the "timing mass" of the Local Group. Kahn & Woltjer (1959) found that the mass of the Local Group far exceeded the visible stellar mass and suggested the bulk of the missing mass was in the form of a warm, low-density plasma. Although it is now generally believed that the Local Group is dominated by dark matter, Kahn & Woltjer's estimates for the properties of the intragroup medium are remarkably similar to more recent estimates. More than a decade after Kahn & Woltjer, the idea of diffuse gas in the Local Group and other groups was revisited by Oort (1970), Ruderman & Spiegel (1971), Hunt & Sciama (1972), Silk & Tarter (1973).

The earliest claims for X-ray detections of groups came from the non-imaging X-ray telescopes Uhuru, Ariel 5, and HEAO 1 in the 1970s. Cooke et al (1978) produced a catalog (known as the 2A) of 105 bright X-ray sources from the Leicester Sky Survey Instrument on Ariel 5. Based on positional coincidences, Cooke et al (1978) suggested the identification of seven X-ray sources in the 2A catalog as groups of galaxies. Subsequent observations showed that several of these X-ray sources were variable, indicating they were actually active galaxies within the group (Ricker et al 1978, Ward et al 1978, Griffiths et al 1979). However, several of the remaining objects in Cooke et al (1978) were later shown to be poor clusters (Schwartz et al 1980).

X-ray studies of lower-mass systems received a major boost with the launch of the Einstein Observatory in November 1978. Einstein observations firmly established that some poor clusters with bright central galaxies (i.e. MKW and AWM clusters; Morgan et al 1975, Albert et al 1977) were X-ray sources (Kriss et al 1980, 1983, Burns et al 1981, Price et al 1991, Dell'Antonio et al 1994). The X-ray luminosities of these poor clusters range from several times 10⁴¹ ergs s^-1 h^-2₁₀₀ up to several times 10⁴³ ergs s^-1 h^-2₁₀₀. The X-ray emission in these poor clusters was shown to be extended (out to radii as great as 0.5 h^-1₁₀₀ Mpc) with temperatures in the range T ~ 1-5 keV. Although most of these systems are somewhat richer than the typical groups considered in this review, these Einstein observations clearly demonstrated that diffuse X-ray emission was not restricted to rich clusters.

Several attempts were also made to study even poorer galaxy systems with Einstein. Biermann and collaborators detected extended emission in two nearby elliptical-dominated groups (Biermann et al 1982, Biermann & Kronberg 1983). In both cases, the X-ray emission was centered on the dominant galaxy. For the NGC 3607 group, Biermann et al (1982) concluded that the X-ray emission most likely originated from a hot, intergalactic gas because it was extended on scales larger than the galaxy. (Biermann et al estimate a Gaussian width for the X-ray emission of 4.7' 13 h^-1₁₀₀ kpc.) From a rough fit to the X-ray spectrum, a temperature of 5 × 10⁶ K and an X-ray luminosity of 2 × 10⁴⁰ h^-2₁₀₀ ergs s^-1 was found. Following their discovery of X-ray emission in the NGC 3607 group, Biermann & Kronberg (1983) found a similar component in the NGC 5846 group. The Einstein Observatory was also used to study the X-ray properties of compact groups. Bahcall et al (1984) studied five compact groups, including four from Hickson's (1982) catalog. Three of the compact groups were detected with Einstein. The Einstein exposure times for these groups were very short, resulting in only ~ 20-60 net counts in the X-ray detected cases. Bahcall et al (1984) noted that the X-ray luminosities of two of the groups were of order ~ 10⁴² ergs s^-1 h^-2₁₀₀, much higher than the X-ray emission expected from the member galaxies alone. The emission was also extended in these two groups, and in the case of Stephan's Quintet, the shape of the X-ray spectrum was unlike that expected from individual galaxies. These X-ray properties led Bahcall et al (1984) to conclude that the X-ray emission likely originated in a hot intragroup gas in at least two of the five groups they studied. Thus, although it was not possible to unambiguously separate a diffuse component from galaxy emission with Einstein, there were strong indications that intragroup gas was likely present in some groups.