Search

Prior to ROSAT, separation of X-ray components in radio galaxies has been limited to a few well-known sources, e.g., M 87 and Cen A. Now, from ROSAT PSPC measurements of the first six objects in our study of low-power radio galaxies, we find that both resolved (thermal) and unresolved X-ray emission in a single source is typical (Worrall & Birkinshaw 1993; ApJ, submitted). The angular size of, and fraction of luminosity in, the resolved X-ray emission varies between objects. There is evidence to relate the unresolved X-ray emission with the inner radio jet.

A correlation of the form , with γ = 0.83, is found between 2 keV monochromatic X-ray luminosity and 1.65 μm monochromatic infrared luminosity for a sample of optically selected radio-quiet QSOs. This is equivalent to an average 1.65 μm to 2 keV spectral index of αir/x = 1.28+0.05 log(lir/1031 ergs s−1 Hz−1). The ratio of X-ray to infrared luminosity decreases with increasing infrared luminosity in a manner similar to the relationship between X-ray and optical luminosity. The standard deviations of the distributions of data about the best-fit functions are comparable for the X-ray versus infrared and X-ray versus optical fits. Thus, contrary to a previous claim, the infrared luminosity is not better than the optical luminosity at predicting the X-ray emission from QSOs.

A good correlation is known to exist between the X-ray and radio luminosity of flat-spectrum, core-dominated radio sources (e.g., Owen, Helfand and Spangler 1981). Worrall (1987) presents a logarithmic plot of spectral luminosity in the source frame at 2 keV versus that at 5 GHz for a variety of QSOs, Highly Polarized QSOs (HPQs), and BL Lac Objects. Friedmann cosmology with Ho = 100h km s−1 Mpc−1, qo = 0 is assumed. Exclusion of objects which are optically or X-ray selected, or in which the radio emission is not dominated by a flat-spectrum compact core, gives a sub-sample consisting of 50 QSOs, 20 HPQs, and 10 BL Lacs, of which 5,4,3, respectively, are known superluminals. The dispersion of these data about the log-log correlation (assuming a Gaussian distribution), is σobs = 0.44 ± 0.06 (90% confidence errors for one interesting parameter).

Quasars with similar core-compact radio properties can be classified by their differences at optical and infrared frequencies. Their X-ray properties might be expected to be similar if the synchrotron self-Compton mechanism relates their radio and X-ray emission. We have compared the 0.2–3.5 keV mean power-law energy spectral indices, , for 4 quasar classes: 12 Highly Polarized QSOs (HPQs), 19 Flat Radio Spectrum, core-compact, low-polarization, QSOs (FRS QSOs), 24 radio-selected BL Lac objects, and 7 X-ray-selected BL Lac objects.

Our statistical study finds that the X-ray luminosity, lx for an average QSO of given optical luminosity, lo is an increasing function of radio luminosity, lr. An average QSO of given lo and lr has higher lx if the radio spectrum is flat than if it is steep. The correlation between lx, lo and lr appears to be improved if lr refers only to core emission, lrcore. Assuming an average radio-loud QSO is a radio-quiet QSO with a separate component which produces lrcore and Ixcore, lxcore/lrcore is larger if the core is that of a QSO with extended radio structure than if that of a compact flat-spectrum QSO.