|Annu. Rev. Astron. Astrophys. 1989. 27:
Copyright © 1989 by . All rights reserved
The extragalactic X-ray background was discovered in 1962 in the same rocket flight that led to the discovery of the first extrasolar source of X-rays, Sco X-1 (Giacconi et al. 1962). Since then a great deal of effort has been spent trying to understand if this radiation is due to the integrated contributions of different classes of discrete sources or if diffuse processes are responsible for it (see Boldt 1987, Giacconi & Zamorani 1987). Based on the four galaxies then known to emit X rays (Milky Way, M31, and the Magellanic Clouds), Silk (1973) estimated that normal galaxies would contribute ~ 10% to the X-ray background. This estimate was then revised downward to less than ~ 1% (Rowan-Robinson & Fabian 1975, van Paradijs 1978, Worrall et al. 1979; see Fabian 1981) for a variety of reasons. One is the assumption that the X-ray luminosity of normal galaxies. which is dominated by the low-mass binary sources, would not be correlated with the optical luminosity but with some function of the galaxy mass (Rowan-Robinson & Fabian 1975); however, we now know that this is not so, as discussed earlier in this review. Other reasons are the effect of the redshift on the spectrum of the galactic sources (van Paradijs 1978), and the failure to detect a volume-limited sample of galaxies with the HEAO-1 satellite (Worrall et al. 1979). However, this survey observed galaxies that are all of very low optical luminosity and as such cannot be considered representative of normal galaxies as a whole (Fabian 1981, Elvis et al. 1984).
Based on the X-ray to optical ratios of normal galaxies observed with the Einstein Observatory (Fabbiano & Trinchieri 1985, Trinchieri & Fabbiano 1985), Giacconi & Zamorani (1987) estimated instead that the integrated emission of normal galaxies contributes ~ 13% of the 2-keV extragalactic X-ray background (see also Setti 1985). If one includes in this estimate the contribution of low-activity nuclei present in a fraction of these galaxies and the effect of starburst activity, it could be significantly larger: Elvis et al. (1984) estimate that the low-activity nuclei could contribute ~ 20% of the X-ray background in the absence of evolution; and Weedman (1987), using X-ray to optical ratios measured in starburst galaxies (Fabbiano et al. 1982) and the 60-µm luminosity function derived from the IRAS survey, finds that in the absence of evolution these galaxies could account for ~ 13% of the 2-keV background (see also Giacconi & Zamorani 1987). If one assumes that starburst activity was much more pronounced in the past (Bookbinder et al. 1980, Stewart et al. 1982), starburst galaxies could be responsible for the bulk of the background (Giacconi & Zamorani 1987, Weedman 1987). Although this possibility is rejected by Giacconi & Zamorani because it would predict a surface density of 21-23 mag galaxies inconsistent with optical searches, the presence of a large amount of dust in these systems, suggested by the IRAS data, leaves this a still viable option (Weedman 1987).
Even if galaxies contribute substantially to the 2-keV X-ray background, their contribution to the X-ray background in a harder spectral range is uncertain and rests upon the spectral characteristics of their X-ray emission (van Paradijs 1978, Giacconi & Zamorani 1987, Weedman 1987). The spectra of spiral galaxies are consistent with a relatively hard X-ray emission (kT > 2 keV; Fabbiano & Trinchieri 1987), and X-ray binaries can have hard X-ray spectra (with kT ~ 20 keV). Moreover, a hard spectral component may be present in the starburst galaxy M82 (Fabbiano 1988b, Schaaf et al. 1989). However, it is unlikely that the galactic contribution would be substantial above 10 keV, especially considering that the spectra of faraway galaxies would be redshifted. The soft X-ray spectrum of the nucleus of M81 (Fabbiano 1988a) introduces an additional source of uncertainty to the estimate by Elvis et al. (1984) of the contribution of low-activity nuclei to the 2-10 keV X-ray background, since most of the X-ray luminosity of these sources could be emitted in a softer energy range.
On the other hand, we do not know how common and how bright the optically quiet X-ray-active nuclei, such as those of M33 and NGC 1313, are (see earlier in this paper). Their inclusion could raise the estimate of the galactic contribution to the X-ray background, although their X-ray spectra (Fabbiano & Trinchieri 1987, Trinchieri et al. 1988) suggest that even this type of source should contribute mainly in the soft energy range. Future X-ray observations will help to constrain the spectral range in which galaxies may contribute to the extragalactic X-ray background and will give us a better estimate of the contribution of low-activity nuclei.