Introduction

The 21-cm line of neutral hydrogen gives us almost our only view of galaxies that is independent of stars. Stars directly generate nearly all of the visible emission, of course, and they are also directly or indirectly responsible for almost all of the emission at other wavelengths. For example, far-infrared emission from interstellar dust grains is powered by starlight heating the grains, and the grains would not exist but for the nucleosynthesis in former generations of stars. Likewise, virtually every other form of continuum or line radiation is powered by stars in some stage of their evolution and/or depends on the byproducts of stellar evolution. By contrast, neutral hydrogen is primordial, and the 21 cm line has such a low excitation temperature that the ionizing extragalactic background radiation is sufficient to maintain its excitement.

Until recently, 21 cm receivers were not sensitive enough to easily detect the HI line from galaxies. Since a large investment of time was needed to detect a galaxy, observers pointed their radio telescopes at known galaxies--known from their optical emission. Unfortunately, this teaches us little about HI independent of stars. In recent years, system temperatures have improved to the point that a galaxy could be detected rapidly, often in as little as a few seconds. Coupled with the advent of large multi-beam systems like the 13-beam Parkes receiver, it has become reasonable to start surveying the sky ``blindly'' to determine how neutral hydrogen behaves in its own right.

Several single-beam surveys at Arecibo, Green Bank, the VLA, and smaller telescopes have been and are currently being conducted. The Parkes Multibeam Survey will surpass all of these surveys combined as it examines an entire hemisphere of the sky. The multibeam receiver should provide the ideal equipment for a blind survey, with good sensitivity and the potential for excellent interference rejection by cross-comparison of the signals.

My focus in this paper will be on the steps needed to translate HI survey observations into a galaxy luminosity function. I apply tools developed for understanding the completeness of optical samples to two Arecibo surveys, and discuss some of the problems associated with interpreting the results. These problems suggest that we need a much more thorough understanding of how observing procedures and line-identification techniques impact on the detection rate, and I suggest some strategies for the Parkes survey and analysis procedures to clarify the interpretation.

© Copyright Astronomical Society of Australia 1997