Annu. Rev. Astron. Astrophys. 1999. 37:311-362 Copyright © 1999 by Annual Reviews. All rights reserved

1. INTRODUCTION

Long after their parent spiral galaxies have formed, stars continue to form by repeated condensation from the interstellar medium. In the process, parts of the interstellar medium pass through a cool, relatively dense phase with a great deal of complexity– molecular clouds. While both the diffuse interstellar medium and stars can be supported by thermal pressure, most molecular clouds cannot be thermally supported (Goldreich & Kwan 1974). Simple consideration would suggest that molecular clouds would be a very transient phase in the conversion of diffuse gas to stars, but in fact they persist much longer than expected. During this extended life, they produce an intricate physical and chemical system that provides the substrate for the formation of planets and life, as well as stars. Comparison of cloud masses to the total mass of stars that they produce indicates that most of the matter in a molecular cloud is sterile; stars form only in a small fraction of the mass of the cloud (Leisawitz et al 1989).

The physical conditions in the bulk of a molecular cloud provide the key to understanding why molecular clouds form an essentially metastable state along the path from diffuse gas to stars. Most of the mass of most molecular clouds in our Galaxy is contained in regions of modest extinction, allowing photons from the interstellar radiation field to maintain sufficient ionization for magnetic fields to resist collapse (McKee 1989); most of the molecular gas is in fact in a photon-dominated region, or PDR (Hollenbach & Tielens 1997). In addition, most molecular gas has supersonic turbulence (Zuckerman & Evans 1974). The persistence of such turbulence over the inferred lifetimes of clouds in the face of rapid damping mechanisms (Goldreich & Kwan 1974) suggests constant replenishment, most likely in a process of self-regulated star formation (Norman & Silk 1980, Bertoldi & McKee 1996), since star formation is accompanied by energetic outflows, jets, and winds (Bachiller 1996).

For this review, the focus will be on the physical conditions in regions that are forming stars and likely precursors of such regions. While gravitational collapse explains the formation of stars, the details of how it happens depend critically on the physical conditions in the star-forming region. The details determine the mass of the resulting star and the amount of mass that winds up in a disk around the star, in turn controlling the possibilities for planet formation. The physical conditions also control the chemical conditions. With the recognition that much interstellar chemistry is preserved in comets (Crovisier 1999, van Dishoeck & Blake 1998), and that interstellar chemistry may also affect planet formation and the possibilities for life (e.g. Pendleton 1997, Pendleton & Tielens 1997, Chyba & Sagan 1992), the knowledge of physical conditions in star-forming regions has taken on additional significance.

Thinking more globally, different physical conditions in different regions determine whether a few, lightly clustered stars form (the isolated mode) or a tight grouping of stars form (the clustered mode) (Lada 1992; Lada et al 1993). The star formation rates per unit mass of molecular gas vary by a factor > 10² in clouds within our own Galaxy (Evans 1991, Mead et al 1990), and starburst galaxies achieve even higher rates than are seen anywhere in our Galaxy (e.g. Sanders et al 1991). Ultimately, a description of galaxy formation must incorporate an understanding of how star formation depends on physical conditions, gleaned from careful study of our Galaxy and nearby galaxies.

Within the space limitations of this review, it is not possible to address all the issues raised by the preceding overview. I will generally avoid topics that have been recently reviewed, such as circumstellar disks (Sargent 1996, Papaloizou & Lin 1995, Lin & Papaloizou 1996, Bodenheimer 1995), as well as bipolar outflows (Bachiller 1996), and dense PDRs (Hollenbach & Tielens 1997). I will also discuss the sterile parts of molecular clouds only as relevant to the process that leads some parts of the cloud to be suitable for star formation. While the chemistry and physics of star-forming regions are coupled, chemistry has been recently reviewed (van Dishoeck & Blake 1998). Astronomical masers are being concurrently reviewed (Menten 1999); as with HII regions, they will be discussed only as signposts for regions of star formation.

I will focus on star formation in our Galaxy. Nearby regions of isolated, low-mass star formation will receive considerable attention (Section 4) because we have made the most progress in studying them. Their conditions will be compared to those in regions forming clusters of stars, including massive stars (Section 5). These regions of clustered star formation are poorly understood, but they probably form the majority of stars in our Galaxy (Elmegreen 1985), and they are the regions relevant for comparisons to other galaxies.

Even with such a restricted topic, the literature is vast. I make no attempt at completeness in referencing. On relatively non-controversial topics, I will tend to give an early reference and a recent review; for more unsettled topics, more references, with different points of view, will be given. Recent or upcoming publications with significant overlap include Hartmann (1998), Lada & Kylafis (1999), and Mannings et al (2000).