Like the radio, the X-ray luminosity is but a small fraction of the total bolometric luminosity of all starbursts and most AGN, albeit usually a few orders of magnitude in L larger than for the radio, but still several orders of magnitude less than the infrared luminosity of ULIGs. It is not clear that the observed X-ray spectrum can directly be related to the bolometric luminosity of ULIGs. However, by analogy with the mean X-ray properties of QSOs or starbursts, it may at least be possible to say whether a QSO-like X-ray nucleus is present.
The 2-10 keV luminosity of all five ULIGs in Table 2 is weak relative to the strong far-infrared emission (e.g. Nakagawa et al. 1999); however, the X-ray luminosity of ULIGs is not weak in relation to RQQSOs when comparing the X-ray luminosity to the luminosity in the optical (e.g. Ogasaka et al. 1997; Iwasawa 1999; Turner 1999). Strong absorption appears to affect both the optical/UV and soft X-rays in ULIGs, as might have been expected.
Absorption effects are minimized by considering only the hard X-ray emission, LHX L(5-10 keV), detected with ASCA. Three of the five ULIGs were detected in hard X-rays, all at a level consistent with the mean hard X-ray luminosity observed for RQQSOs [Note: Mrk 231 is compared here to the mean properties of BALQSOs which are somewhat weaker in their observed LHX than other RQQSOs (Turner 1999).]