Annu. Rev. Astron. Astrophys. 1988. 26: 245-294 Copyright © 1988 by Annual Reviews. All rights reserved

1. EARLY EXPLORATIONS

1.1. Voids in Astronomy

In the fifth century B.C., Democritus postulated the existence of immutable atoms characterized by size, shape, and motion. Changes in the combination of these atoms were postulated to result in the qualitative changes observed in everyday matter. In order to make the motions of atoms possible, the existence of the void (unoccupied space, empty space) was required [cf. 15th edition of the Encyclopaedia Britannica (1985)].

In astronomy, it is likely that individual voids were first observed by prehistoric man, who saw distinct dark regions within the prominent luminous band (now called the Milky Way) stretching across the sky. Although to the unaided eye these dark regions may seem devoid of light, on well-exposed photographs they appear dark only in contrast to their luminous starry neighborhoods. Counts, magnitudes, and colors of the few stars that lie within the boundaries of the dark regions imply the presence of vast quantities of obscuring particles (dust). The Coalsack is an example of a dust void in Galactic stellar astronomy (29, pp. 166-78; 141).

A typical "void" in the surface distribution of a sample of stars is not empty; it is a region where the stellar surface density is substantially smaller than that of (a) its surrounding (large local density contrast) or (b) an average over the entire sample of stars (large global density contrast), where the calculation of the average includes corrections for the selection functions of the sample. In Galactic stellar astronomy, a typical void in the surface distribution of stars corresponds to a region of three-dimensional space with an atypically large density of obscuring material. In extragalactic astronomy the three-dimensional region that constitutes a void is transparent and empty or nearly empty of galaxies. Voids were not immediately recognized in the surface distribution of galaxies because they tend to be overlaid by superimposed background and especially foreground galaxies. Voids were recognized only after Doppler velocities were measured for statistically homogeneous samples of galaxies in selected solid angles of the sky, which provided (through application of the Hubble relation between distance and redshift) direct information on the three-dimensional distribution of galaxies.

In extragalactic astronomy, the basic material unit is a galaxy with a characteristic length D_c ~ 20 kpc that typically moved within a group or cluster (D_c ~ 50 kpc-5 Mpc) in a supercluster (D_c ~ 50 Mpc or larger). (Herein, a length calculated from an angular extent and Hubble distance has been normalized with an adopted Hubble constant H₀ = 50 km s^-1 Mpc^-1.) (Work on detecting and deciphering the structure of groups and clusters of galaxies (including discussion of the problem of the "missing mass") has been reviewed by Bahcall (10), Rood (156), and Sarazin (169).] The latter two reviews briefly discuss superclusters. A comprehensive review of research on superclusters is given by Oort (128). Superclusters and voids are often discussed together because (a) they are identified and studied with a common data base, and (b) it is likely that they share a common origin and represent two complementary effects of related physical evolutionary processes. Therefore, the present review refers frequently to Oort (128), and figures therein.

If the physical universe is modeled as one vast system of individually distinct galaxies, then these galaxies are immersed in one void, an analog of the void of Democritus. We obtain a similar result if the Universe is modeled as one vast system of groups and clusters of galaxies. However, research on superclusters is in such an early stage that it is not yet clear whether they are typically (a) individual entities immersed in the void (i.e. analogs of Democritus' atoms- galaxies and groups and clusters of galaxies), or (b) parts of one vast system of galaxies that is polka-dotted in three-space with individual voids [analogous to the voids in two-space of Galactic stellar astronomy, but with a much larger fraction of the volume occupied by the voids (large polka dots)], or (c) parts of a connected structure of galaxies intertwined with a connected structure of voids. Models (a) and (b) are discussed in depth by Shandarin & Zel'dovich (176), and model (c), a sponge-like universe, is described in detail by Gott et al. (80).

Pending clarification of the topological nature of superclusters and voids, which may require the completion of the Center for Astrophysics (CfA) redshift survey of galaxies with apparent photographic magnitude m_p 15.5 [where the adopted m_p is generally the Zwicky (220) magnitude; see (93a) for discussion of comparisons between Zwicky magnitudes and other, more accurate but less abundant blue magnitudes], it is often useful to think of a void as a discrete entity - a region containing significantly fewer galaxies than predicted by the appropriate Poisson distribution. The occurrence of physical groupings of galaxies evidently requires the presence of voids defined in this way.

There are two ways that an individual void can be studied observationally: (a) The void can be probed with telescopic sensors in attempts to detect something within it. (b) The structure and content of the contiguous shell of superclusters surrounding the void can be studied. [The term "shell" has been adopted from de Lapparent et al. (56). M. Postman (private communication, 1987) suggests that "shell of galaxies" may be a more appropriate term, as it is not yet clear that superclusters are physically correlated.]