Annu. Rev. Astron. Astrophys. 1997. 35: 267-307 Copyright © 1997 by Annual Reviews. All rights reserved

1. INTRODUCTION

The Cosmological Principle states that the universe will appear homogeneous and isotropic to a typical observer, but one of the deeper implications of this principle is rarely examined. Observational cosmology is usually probed via catalogs of galaxies. Although much of the universe is dark, galaxies are the prime repositories of shining baryonic matter. Galaxy properties are used to measure the size and shape of the universe; deviations from Hubble flow and image distortions through lensing are used to map out the dark matter distribution. If galaxies are to be used as effective cosmological probes, then our catalogs must be complete and homogeneous, in accord with the Cosmological Principle. Yet the detectability of galaxies depends very much on the cosmic environment. An observer whose star was in a giant molecular cloud or near the center of an elliptical galaxy would have difficulty discovering external galaxies and so would perceive the universe quite differently from us. Put simply, we only catalog the galaxies we can see.

Observational bias in the selection of galaxies dates back to Messier and Herschel. Galaxies are diffuse objects selected in the presence of a contaminating signal: the brightness of the night sky. Below a certain percentage of the night sky brightness, no galaxy can be detected. Above this limiting isophote, a galaxy must present a large enough angular size to be distinguished from a star. For a given luminosity and radial profile, a galaxy will be visible to the maximum distance at a surface brightness level substantially higher than the limiting isophote. High surface brightness galaxies (HSB) are small because they are intrinsically compact, and low surface brightness galaxies (LSB) are small because they mostly fall below the limiting isophote. The night sky essentially acts as a filter, which when convolved with the true population of galaxies, gives the population of galaxies we observe. This censorship due to surface brightness was first commented on by Zwicky (1957), and was further investigated by Arp (1965) and Disney (1976).

This review deals primarily with low surface brightness (LSB) galaxies in the local universe (z 0.1). There is no convention for defining low surface brightness; discussion is mostly restricted to galaxies with central surface brightness fainter than 23 B mag arcsec^-2 . Stellar systems more luminous than M_B = - 14 are considered; for reviews of LSB dwarf galaxies in the Local Group and beyond, see Ferguson & Binggeli (1994), Irwin & Hatzidimitriou (1995), and Mateo (1996). First, the structural properties of galaxies are summarized, and the influence of surface brightness selection is described. Next, the potential incompleteness of galaxy luminosity functions is considered. Surveys for LSB galaxies are reviewed. We then establish the significance of LSB galaxies for the census of light and matter in the universe, for the formation and evolution of galaxies, and for the statistics of quasar absorption. This historically neglected population has important implications for virtually every aspect of observational cosmology. Unless otherwised noted, H₀ = 100 km s^-1 Mpc^-1 and q₀ = 0.5 is assumed, or results are expressed in terms of h₁₀₀ = H₀ / 100.