THE VERY SOFT-X-RAY EMISSION OF X-RAY-FAINT EARLY-TYPE GALAXIES

A recent reanalysis of Einstein data, and new ROSAT observations, have revealed the presence of at least two components in the X-ray spectra of X-ray faint early-type galaxies: a relatively hard component (kT > 1.5 keV), and a very soft component (kT approximately 0.2-0.3 keV). I n this paper we address the problem of the nature of the very soft com ponent, and whether it can be due to a hot interstellar medium (ISM), or is most likely originated by the collective emission of very soft s tellar sources. To this purpose, hydrodynamical evolutionary sequences for the secular behavior of gas flows in ellipticals have been perfor med, varying the Type Ia supernovae rate of explosion, and the dark ma tter amount and distribution. The results are compared with the observ ational X-ray data: the average Einstein spectrum for six X-ray faint early-type galaxies (among which are NG-C 4365 and NGC 4697), and the spectrum obtained by the ROSAT pointed observation of NGC 4365. The ve ry soft component could be entirely explained with a hot ISM only in g alaxies such as NGC 4697, i.e., when the depth of the potential well-o n which the average ISM temperature strongly depends-is quite shallow; in NGC 4365 a diffuse hot ISM would have a temperature larger than th at of the very soft component, because of the deeper potential well. S o, in NGC 4365 the softest contribution to the X-ray emission comes ce rtainly from stellar sources. As stellar soft X-ray emitters, we consi der late-type stellar coronae, supersoft sources such as those discove red by ROSAT in the Magellanic Clouds and M31, and RS CVn systems. All these candidates can be substantial contributors to the very soft emi ssion, though none of them, taken separately, plausibly accounts entir ely for its properties. We finally present a model for the X-ray emiss ion of NGC 4365, to reproduce in detail the results of the ROSAT point ed observation, including PSPC spectrum and radial surface brightness distribution. The present data may suggest that the X-ray surface brig htness is more extended than the optical profile. In this case, a stra ightforward explanation in terms of stellar sources could be not satis factory. The available data can be better explained with three differe nt contributions: a very soft component of stellar origin, a hard comp onent from X-ray binaries, and a approximately 0.6 keV hot ISM. The la tter can explain the extended X-ray surface brightness profile, if the galaxy has a dark-to-luminous mass ratio of 9, with the dark matter v ery broadly distributed, and a SN Ia explosion rate of approximately 0 .6 the Tammann rate.