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

4.4. Morphological Content

Correlations between the presence of X-ray emission and the morphological composition of groups were suggested from the earliest ROSAT studies. Ebeling et al (1994) were the first to claim such an effect, noting that all but one of the X-ray detected HCGs in the ROSAT All-Sky Survey data had spiral fraction less than 50%. Subsequent studies of small samples appeared to support this trend (Henry et al 1995, Pildis et al 1995, Mulchaey et al 1996a). However, Ponman et al (1996) came to a very different conclusion based on their much larger survey of the HCGs. They detected several groups with high spiral fractions, including the extreme example HCG 16, a compact group that contains only spirals.

Figure 7 shows the distribution of early-type fraction for all the groups with published pointed observations with ROSAT. For the purposes of this plot, a group is considered "X-ray detected" only if there is evidence for an extended intragroup medium component. As is apparent from this figure, a significant number of spiral-rich groups do contain diffuse X-ray emission, which confirms the conclusion of Ponman et al (1996). In fact, in contrast to the earlier studies, the distribution of early-type fractions is surprisingly flat for the X-ray detected groups. The apparent contradiction with the earlier results can be explained by the fact that the majority of the groups in the current sample were selected from optical redshift surveys and were serendipitously observed by ROSAT (Helsdon & Ponman 2000, Mulchaey et al 2000), whereas the earlier studies were biased toward X-ray luminous groups, which tend to have higher early-type fractions (Mulchaey & Zabludoff 1998).

Figure 7. Distribution of early-type fraction for all groups (open histogram) and groups with diffuse X-ray emission (shaded histogram). The top panel shows the result for all published PSPC pointed-mode observations, whereas the lower panel contains only groups selected from optical redshift surveys.

A closer examination of Figure 7 reveals that while many spiral-rich systems are X-ray sources, spiral-only groups tend not to contain a diffuse X-ray component. The one exception in Figure 7 is HCG 16. However, the true nature of the X-ray emission in HCG 16 is unclear. The ROSAT image of the group indicates that the emission is very clumpy and concentrates around the brightest group members (see Figure 2, top). Some authors have attributed all of the X-ray emission to individual galaxies (Saracco & Ciliegi 1995; see also an earlier Einstein observation by Bahcall et al 1984), whereas others have claimed the existence of intragroup gas (Ponman et al 1996). Dos Santos & Mamon (1999) have reanalyzed the ROSAT PSPC data on HCG 16, paying special attention to the removal of emission associated with galaxies. Although Dos Santos & Mamon (1999) derived a lower luminosity for the diffuse gas than Ponman et al (1996), they still found evidence for some diffuse gas. However, the presence of diffuse emission does not necessarily mean that HCG 16 contains a diffuse intragroup medium. One possibility is that the emission is related to the unusually high number of active galaxies in the group (HCG 16 contains one Seyfert galaxy, two LINERs, and three starburst galaxies; Ribeiro et al 1996). The X-ray to infrared luminosity ratio of this system is much higher than one would expect if the X-ray emission is related to the galaxies' activity, however (Ponman, private communication). Alternatively, the X-ray emission may be associated with shocked gas, as appears to be the case in Stephan's Quintet (Pietsch et al 1997).

With the possible exception of HCG 16, all X-ray detected groups studied to date contain at least one early-type galaxy. There are several possible explanations for why spiral-only groups do not contain diffuse X-ray emission. One possibility is that all spiral-only groups are chance superpositions and not real, physical systems. This possibility seems unlikely, given the existence of our own spiral-only Local Group (see Section 5.10 for a discussion of the intragroup medium in the Local Group). Another possibility is that the intragroup gas in spiral-only groups is too cool to produce appreciable amounts of X-ray emission (Mulchaey et al 1996b). Based on velocity dispersions, the virial temperatures of spiral-only groups do tend to be lower than those of their early-type dominated counterparts (Mulchaey et al 1996b). While a cool (i.e. several million degrees K) intragroup medium would be difficult to detect in X-ray emission, such gas might produce prominent absorption features in the far-ultraviolet or X-ray spectra of background quasars (Mulchaey et al 1996b, Perna & Loeb 1998, Hellsten et al 1998). In fact, several such groups may have already been detected as OVI 1031.93, 1037.62 Å absorption systems (Bergeron et al 1994, Savage et al 1998). A third possibility is that the gas densities in spiral-only groups are too low to be detected in X-rays. Low gas densities in spiral-only groups are in fact consistent with recent prediction of preheating models for groups (Ponman et al 1999; see Section 5.9).