Recently it has been suggested that the hard x-ray power-law tails exhibited in the ASCA spectra of several nearby giant elliptical galaxies are caused by low-radiative efficiency accretion flows (ADAFs) onto a massive black hole in the centers of the galaxies. The estimated fluxes from these central ADAFs as derived from the ASCA x-ray spectral analyses indicate that they may be visible as x-ray point sources. We analyze archival ROSAT HRI images of three such galaxies, NGCs 1399, 4636, and 4696, to determine whether point x-ray sources consistent with ASCA-derived fluxes are present in the centers of the galaxies. We find that the upper limit for the flux of a central point x-ray source in each of these galaxies as determined by the ROSAT HRI data is only marginally consistent with the predicted flux inferred from the ASCA spectra. We suggest that although a central point source, such as a ADAF associated with a massive black hole, cannot be completely ruled out, they can only account for a fraction of the flux of the observed hard x-ray power law components.

Imaging of the Sunyaev-Zel'dovich (SZ) effect in galaxy clusters combined with cluster plasma x-ray diagnostics can measure the cosmic distance scale to high redshift. Projecting the inverse-Compton scattering and x-ray emission along the cluster line-of-sight introduces systematic errors in the Hubble constant, H0, because the true shape of the cluster is not known. I present a study of the systematic errors in the value of H0, as determined by the x-ray and SZ properties of theoretical samples of triaxial isothermal ``beta'' model clusters, caused by projection effects and observer orientation. I calculate estimates for H0 for each cluster based on their large and small apparent angular core radii and their arithmetic mean. I demonstrate that the estimates for H0 for a sample of 25 clusters have 99.7% confidence intervals for the mean estimated H0 analyzing the clusters using either their large or mean angular core radius are within 14% of the ``true'' (assumed) value of H0 (and enclose it), for a triaxial beta model cluster sample possessing a distribution of apparent x-ray cluster ellipticities consistent with that of observed x-ray clusters. This limit on the systematic error in H0 caused by cluster shape assumes that each sample beta model cluster has fixed shape; deviations from constant shape within the clusters may introduce additional uncertainty or bias into this result.

Constraints on the clustered mass density \Omega_m of the universe derived from the mean intracluster gas fraction of X-ray clusters may be biased by a single-phase assumption for the thermodynamic structure of the intracluster medium (ICM). We propose a descriptive model for multiphase structure in which a spherically symmetric ICM contains isobaric density perturbations with radially dependent variance \sigma^2(r)=\sigma_c^2 (1+r^2/r_c^2)^{-\eps}. The model extends the work of Gunn & Thomas (1996) which assumed radially independent density fluctuations thoughout the ICM. Fixing the X-ray emission profile and emission weighted temperature, we explore two independently observable signatures of the model in the {\sigma_c,\eps} space. For bremsstrahlung dominated emission, the central Sunyaev--Zeldovich (SZ) decrement in the multiphase case is increased over the single-phase case and multiphase X-ray spectra in the range 0.1-20 keV are flatter in the continuum and exhibit stronger low energy emission lines than their single-phase counterpart. We quantify these effects for a fiducial 10^8 K cluster and demonstrate how the combination of SZ and X-ray spectroscopy can be used to identify a preferred location in the {\sigma_c,\eps} plane. From these parameters, the correct value of mean ICM gas fraction in the multiphase model results, allowing an unbiased estimate of \Omega_m to be recovered. The consistency of recent determinations of the Hubble constant from SZ and X-ray observations with values determined by other methods suggests that biases in ICM gas fractions are small, \ltsim 20%.

Statistical uncertainties in determining the temperatures of hot (0.5 keV to 10 keV) coronal plasmas are investigated. The statistical precision of various spectral temperature diagnostics is established by analyzing synthetic ASCA Solid-state Imaging Spectrometer (SIS) CCD spectra. The diagnostics considered are the ratio of hydrogen-like to helium-like line complexes of $Z\ge14$ elements, line-free portions of the continuum, and the entire spectrum. While fits to the entire spectrum yield the highest statistical precision, it is argued that fits to the line-free continuum are less susceptible to atomic data uncertainties but lead to a modest increase in statistical uncertainty over full spectral fits. Temperatures deduced from line ratios can have similar accuracy but only over a narrow range of temperatures. Convenient estimates of statistical accuracies for the various temperature diagnostics are provided which may be used in planning ASCA SIS observations.