At present there are in use three different models to characterize the large scale structure of the universe. The clustering model (Soneira and Peebles, 1978) assumes that the superclusters are high density islands in a low density sea. The void model (Joeveer and Einasto, 1978), on the other hand, assumes that the voids are isolated low density islands in a high density sea. The sponge model (Gott et al., 1986) assumes that high and low density regions occupy equal volumes, and that the high and low density regions are both connected. The straightforward way to decide among these three models is the direct investigation of the spatial distribution of the galaxies. Nevertheless, there is an essentially different observational method that may also be useful to obtain some information about these models. The X-ray background radiation (XRB) is due either to the bremsstrahlung of hot intergalactic gas, or to the sum of the radiation of unresolved discrete sources (E.G. Boldt 1987). If the “discrete” origin is correct, then obviously the actual number of sources, and hence their total intensity, may vary from one part of the sky to another. Thus, in this case one has the possibility to estimate the number of sources in a given volume from the observed isotropy of the XRB. For example, Hamilton and Helfand (1987) suggest that the number of sources must be larger than 5000/(degree)2. Any such estimate needs several assumptions. In the previous works one usually assumed that the sources were distributed completely randomly; see, e.g. Fabian (1972). Nevertheless, if the XRB is generated by young galaxies (Bookbinder et al. 1980), it is not excluded that the sources of the SRB are also grouped similarly to galaxies. Because in this case the distribution of sources of the XRB is not completely random, one may expect a different type of fluctuations in the intensity of the XRB. In addition, since the grouping may be quite different for the three structure models, the expected fluctuations may also be different. There is a chance to discriminate among them using the observed isotropy of XRB. The basic observational datum concerning the isotropy of the XRB is well-known: the fluctuations in the intensity are smaller than 3%, if 3° × 3° pixels are used Shafer (1983).