In order to the test the hypothesis that radio sources in elliptical galaxies are fuelled by a fraction of accreted X-ray gas, a sample of 81 cD galaxies in clusters and 23 cD galaxies in poor groups is studied. Various subsamples have been defined (reviewed in Table t) according to the origin of the cD galaxy classification (optically, radio or X-ray selected). A catalogue is presented, listing the measured optical, radio and X-ray luminosities from various origins, but all transformed to a uniform and homogeneous system: optical Mv (38 kpc metric diameter), radio P1.4 (1.4 GHz monochromatic total radio power) and Lx (1 Mpc metric diameter 0.5-3.0 keV X-ray band). The three luminosity parameters are investigated for cross- correlations by studying power-power plots and by analysing how the integral radio luminosity function, expressed in fractions of radio detections (F(> P1.4)), depend on Mv and Lx. All three parameters are found to correlate with each other. F(> P1.4) increases with both increasing Lx and brighter Mv and Lx also increases with brighter Mv. The determinations of the different regression relations are internally consistent. The empirical conclusions from the analysis are: (i) The mean Mv of poor group cDs is 0.m4 fainter than the mean Mv of cluster cDs. (ii) The bivariate radio luminosity functions of both samples confirm, both in shape and in their dependence on Mv, those of normal and giant ellipticals. (iii) cD galaxies have an increasing probability to contain a central (≲ 28 kpc) radio source when the X-ray luminosity of their halo (˜1 Mpc diameter) increases. 50 ± 9% of Lx ≧ 1044 erg s-1 cDs have a central radio source with P1.4 ≧ 1024WHz-1, while 12+l2-5% of Lx <1043 ergs-1 cDs have a radio source of that power. This important conclusion is summarised in Fig. 5. (iv) Comparing rich cluster cDs and poor group cDs a relation between Mv and Lx is found. This relation holds among the rich cluster cDs as well. The physical origins of the relations between cluster richness R, Lx, Mv, and P1.4 are discussed in detail. Several lines of evidence suggest gravitational binding of the X-ray gas to the cD galaxies. It is argued that the observed relation between Lx and P1.4 is not caused by confinement of the radio emitting plasmas by the X-ray gas, but instead by gas accretion into the galaxies. A gravitationally driven radiative accretion flow of the X-ray gas is fuelling the central non-thermal radio source. The observed correlations are interpreted according to the following scheme, (see also Fig. 7): (i) R - Mv: more massive cDs are formed in richer clusters. (ii) Mv - Lx: more massive galaxies are able to bind gravitationally more intergalactic gas and hence produce more luminous X-ray haloes. (iii) Lx - P1.4: in galaxies with more luminous X-ray babes the central engine, which generates the radio source, is fuelled at a higher rate. Other relations between the four parameters are thought to originate dominantly from these three principal relations. E.g. the result of Auriemma et al. (t977) that optically more luminous elliptical galaxies have a higher probability of becoming a radio source, is an indirect result of the combination of relations (ii) + (iii). Our study shows on three different and fully independent occasions an agreement between observed correlations and the predictions from Bondi's solution for spherical accretion of gas on consequently massive galaxies, where a fraction of that mass is fuelling radio sources (the Mv - Lx relation, the dependence of the RLF on Mv, the P0.6 - Vgal relation for Coma cluster galaxies). Finally, it is suggested that gas accretion into normal and giant elliptical galaxies might be a very general phenomenon, which provides a natural explanation of the radio, optical and X-ray properties of ellipticals.
|Journal||Astronomy and astrophysics|
|Publication status||Published - 1-Sep-1983|