A tight mass-temperature relation, M(r)/r ~ T_X, is expected in most cosmological models if clusters of galaxies are homologous and the intracluster gas is in global equilibrium with the dark matter. We here calibrate this relation using 8 clusters with well-defined global temperatures measured with ASCA and masses inferred from weak and strong gravitational lensing. The surface lensing masses are deprojected in accordance with N-body simulations and analytic results. The data are well-fit by the mass-temperature relation and are consistent with the empirical normalisation found by Evrard et al. (1996) using gasdynamic simulations. Thus, there is no discrepancy between lensing and X-ray derived masses using this approach. The dispersion around the relation is 27 per cent, entirely dominated by observational errors. The next generation of X-ray telescopes combined with wide-field HST imaging could provide a sensitive test of the normalisation and intrinsic scatter of the relation resulting in a powerful and expedient way of measuring masses of clusters of galaxies. In addition, as M(r)/r (as derived from lensing) is dependent on the cosmological model at high redshift, the relation represents a new tool for determination of cosmological parameters, notably the cosmological constant Lambda.

Properties of high redshift clusters are a fundamental source of information for cosmology. It has been shown by Oukbir and Blanchard (1997) that the combined knowledge of the redshift distribution of X-ray clusters of galaxies and the luminosity-temperature correlation, $L_X-T_X$, provides a powerful test of the mean density of the Universe. In this paper, we address the question of the possible evolution of this relation from an observational point of view and its cosmological significance. We introduce a new indicator in order to measure the evolution of the X-ray luminosity-temperature relation with redshift and take advantage of the recent availability of temperature information for a significant number of high and intermediate redshift X-ray clusters of galaxies. From our analysis, we find a slightly positive evolution in the $L_X-T_X$ relation. This implies a high value of the density parameter of $0.85\pm0.2 $. However, because the selection of clusters included inour sample is unknown, this can be considered only as a tentative result. A well-controlled X-ray selected survey would provide a more robust answer. XMM will be ideal for such a program.

We discuss the implications of two possible, recent Sunyaev-Zel'dovich (SZ) detections for which no optical or X--ray counterparts have been found. This suggests that the objects reside at high redshift, which is difficult to reconcile with a critical cosmology. We develop this argument and find that an open model with $\Omo=0.2$ remains consistent with what we currently know about the two objects. The reasoning also demonstrates the utility of SZ cluster searches.

We present a self-consistent approach to the modeling of X-ray clusters of galaxies in a flat universe. Employing the Press and Schechter (1974) formalism to derive the mass function and relating the observable properties of clusters to their virial mass allows us to study the cluster X-ray distribution functions and their evolution with redshift. This approach differs from some results based on the X-ray luminosity function, which we argue is subject to modeling uncertainties. We obtain stringent constraints on the power spectrum of the initial density perturbations assumed to seed galaxy formation by comparing the theoretical temperature function with available data: the amplitude $\sigma_8$ is found to be $0.57 \pm 0.05$, while the local spectral index $n$ is falling in the range $-2.4 \leq n \leq -1.5$ on scales between 5 and 15 $h^{-1}$ Mpc: the standard CDM model is clearly ruled out, independently of the normalization. We further examine the evolutionary properties of X-ray clusters. Our approach greatly clarifies the situation: contrary to some previous claims, we find that flat models are in reasonable agreement with the observed number of clusters at high redshif ts. We also examine the contribution of clusters to the X-ray counts and to the soft X-ray background and compare the expected values in the case of the $\Omega_0=1$ and $\Omega_0=0.2$ universes. The number counts are in agreement with the observations, further confirming the relevance of our modeling. We conclude that although clusters are not the primary source of the soft X-ray background, their contribution is nevertheless non-negligible. This is particularly important for they could escape the constraints imposed on possible point sources contributing to the background. Finally, we briefly

We use a self-consistent modeling of x-ray cluster properties to constrain cosmological scenarios of structure formation in the case of open cosmological models. We first show that an unbiased open model can reproduce present day observations, provided that the density parameter is in the range 0.15-0.4. Although this estimate is derived in a rather different way, it is very close to dynamical measurements. We also obtain constraints on the shape of the initial fluctuation power spectrum: we find that the local spectral index n falls in the range -1.1 < n < -1.6 on scales between 10 and 30h^{-1}Mpc. We focus on differences between open models and the Einstein--de Sitter model, and we exhibit a test which depends only on the cosmological parameter and is insensitive to the shape of the initial fluctuations power spectrum: we show that, contrary to what is generally admitted, the redshift distribution of X-ray clusters selected on the basis of their apparent temperature is almost independent of the power spectrum and that it is only sensitive to the value of density of the universe. Since the relation between the virial mass and the X-ray temperature seems rather robust, we expect very little theoretical uncertainty in the modeling. This allows us to show that the combined knowledge of the redshift distribution of X-ray flux selected clusters and that of the luminosity-temperature correlation at high redshifts offers a new and efficient way to measure the mean density of the universe. This appears to be a clean test not suffering of the problems attached to local density estimators. We have tentatively applied our test to the redshift distribution of clusters detected within the Einstein medium sensitivity X-ray survey.

We examine how the Sunyaev-Zel'dovich (SZ) effect of clusters can probe $\Omega_{o}$. Using self-consistent models of X-ray clusters in the context of hierarchical models of structure formation, we show that both the mean Compton {\it y} parameter and the number of clusters observed via the SZ effect strongly depend on $\Omega_{o}$. These quantities are higher in open cosmological models due to the earlier epoch of structure formation. Specifically, we compare two models which are able to reproduce the present abundance of X-ray clusters, one for $\Omega_{o}=0.2$ and the other $\Omega = 1$. For $\Omega_{o}=0.2$, ${\ymean}$ exceeds the current FIRAS limit of $1.5\times10^{-5}$. Also, the SZ redshift source distribution considerably differs in the two cosmological models. Thus, these results show that both an improvement in our knowledge of CMB spectral distortions and the realization of millimeter surveys to look for SZ sources over a large area of the sky should produce interesting constraints on $\Omega_{o}$ and on the evolution of the baryonic fractionof virialized structures.

We consider the Sunyaev-Zel'dovich (SZ) effect as a probe of Omega_o: Using a self-consistent modeling of X-ray clusters, we examine the dependence of both the mean Compton y parameter and the SZ source counts on Omega_o. These quantities increase with decreasing Omega_o due to the earlier epoch of structure formation in low-density cosmogonies; the results depend only on the quantity of gas heated to the virial temperature of collapsed objects and are independent of the spatial distribution of the gas in the potential wells. Specifically, we compare two models which reproduce the present-day abundance of clusters - a biased, critical universe and an unbiased, open model with Omega_o=0.3. We find that the mean y parameter approaches the current FIRAS limit of y<2.5 10^{-5} for the open model, demonstrating the importance of improving this limit on spectral distortions, and that the SZ source counts and corresponding redshift distribution differ significantly between the two cosmogonies; millimeter surveys covering a large area should thus provide interesting constraints on the density parameter of the Universe and on the evolution of the heated gas fraction in virialized objects.