We have analyzed the spatial distribution of galaxies from the release of the Sloan Digital Sky Survey of galactic redshifts (SDSS DR7), applying the complete correlation function (conditional density), two-point conditional density (cylinder), and radial density methods. Our analysis demonstrates that the conditional density has a power-law form for scales lengths 0.5-30 Mpc/h, with the power-law corresponding to the fractal dimension D = 2.2+-0.2; for scale lengths in excess of 30 Mpc/h, it enters an essentially flat regime, as is expected for a uniform distribution of galaxies. However, in the analysis applying the cylinder method, the power-law character with D = 2.0+-0.3 persists to scale lengths of 70 Mpc/h. The radial density method reveals inhomogeneities in the spatial distribution of galaxies on scales of 200 Mpc/h with a density contrast of two, confirming that translation invariance is violated in the distribution of galaxies to 300 Mpc/h, with the sampling depth of the SDSS galaxies being 600 Mpc/h.

Gravitational lensing of background compact objects like active galactic nuclei and quasars, by extended intermediate mass lenses such as globular clusters and and dark matter clumps with masses 10^5 - 10^8 M_sun, is considered. It is shown that observational study of the galaxy-quasar's associations is a powerful direct observational test of the nature of massive galaxy haloes. Optical interferometric observations with VLTI and Keck instruments are able to constrain masses and number of substructure halo objects. Evidence of gravitational lensing by globular clusters in haloes of spiral and elliptical galaxies is presented.

Based on the SDSS catalog, we have found new close quasar-galaxy pairs. Quasars projected onto the halos of nearer galaxies are encountered among the multitude of quasars observed at various distances from us. Among them there are quasars that are close to the galaxies not only in angular separation, but also in redshift. Such quasar-galaxy pairs are called close pairs. We developed further the hypothesis that such pairs appear, because the fluxer on the nucleus of the more distant galaxy passes through halo globular clusters of the nearer galaxy,resulting in magnification and splitting of the image of the source that we interpret as a quasar. To corroborate this hypothesis, we analyzed the distribution of quasars in the plane of the halos of these galaxies. The quasars from close pairs were found to follow the density profile of globular clusters in the halos of elliptical and spiral galaxies with slopes of -1.5 and -2.4 for elliptical and spiral galaxies, respectively. This suggests that quasars do not appear near galaxies by chance and that quasars are associated with galaxies via halo globular clusters.

We suggest a new method for estimating the fractal dimension of the spatial distribution of galaxies, the method of selected cylinders. We show the capabilities of this method by constructing a two-point conditional column density for galaxies with known redshifts from the LEDA database. The fractal dimension of a sample of LEDA and EDR SDSS galaxies has been estimated to be D = 2.1 for cylinder lengths of 200 Mpc. Amajor advantage of the suggested method is that it allows scales comparable to the catalog depth to be analyzed for galaxy surveys in the form of conical sectors and small fields in the sky.

We argue that globular clusters (GCs) are good candidates for gravitational lenses in explaining quasar-galaxy associations. The catalog of associations (Bukhmastova 2001) compiled from the LEDA catalog of galaxies (Paturel 1997) and from the catalog of quasars (Veron-Cetty and Veron 1998) is used. Based on the new catalog containing 8382 pairs, we show that one might expect an increased number of GCs around irregular galaxies of types 9 and 10 from the hypothesis that distant compact sources are gravitationally lensed by GCs in the halos of foreground galaxies. The King model is used to determine the central surface densities of 135 GCs in the Milky Way. The distribution of GCs in central surface density was found to be lognormal.

The distribution function of quasars with respect to apparent brightness is given, found under the assumption that quasars are, at least partially, the gravitationally enhanced images of the active nuclei of distant galaxies. A Schechter law and a two-power law for the luminosity function of the sources are used, as well as a probabilistic law of image enhancement for various models of gravitational lenses. To find the theoretical distribution function of quasars with respect to apparent brightness we use a theorem on the probability density of a product of random quantities. It is shown that the slope of this function ranges from -1 to -2 for faint quasars, like that for ordinary galaxies. In the case of bright quasars, the slope of the apparent brightness distribution function is determined mainly by the lensing effect and has a lower limit of -3. The good agreement between theory and observations suggests that statistically quasars are gravitationally enhanced images of the active nuclei of distant galaxies. If the initial assumptions are correct, then the luminosity function of galaxies and the apparent brightness function for quasars are not independent but are related by means of the differential lensing probability.

A new catalog of 8382 close quasar-galaxy pairs is presented. The catalog was composed using published catalogs of quasars and active galactic nuclei containing 11358 objects, as well as the LEDA catalog of galaxies, which contains on the order of 100 thousand objects. The search for pairs was carried out in such a way that the linear distance between the galaxy and projected quasar does not exceed 150kpc. Based on these new data, the dependence of the number of pairs on a=z_G/z_Q is analysed, where z_G and z_Q are the redshifts of the galaxy and quasar, respectively, revealing an excess of pairs with a<0.1 and a>0.9. This means that the galaxies in pairs are preferably located close to either the observer or the quasar and avoid intermediate distances along the line of sight to the quasar. Computer simulations demonstrate that it is not possible to explain this number of pairs with the observed distribution in a as the result of chance positional coincidences with a uniform spatial distribution of galaxies. Data on globular clusters show that the excess of pairs with a<0.1 and a>0.9 is consistent with the hypothesis that we are observing distant compact objects that are strongly gravitationally lensed by transparent lenses with a King mass distribution located in the halos of nearby galaxies. The Hubble diagram for galaxies and quasars is presented. Observational tests of the mesolensing hypothesis are formulated.

The probability of strong gravitational lensing of distant compact objects by transparent lenses with King mass distributions is calculated. Examples of such lenses are globular clusters, dwarf galaxies, and clusters of hidden mass. The lens masses thus range from 10^3 to 10^9 Solar mass, intermediate between stellar and galactic masses. Taking account of the conic caustic of King lenses and the fractal large-scale distribution of matter along the line of sight, it is possible to explain the correlation between quasars and galaxies as the effect of strong gravitational lensing of the active nuclei of distant galaxies by globular clusters in the halos of nearby galaxies. Observational tests for verifing this mesolensing hypothesis are described.