|Annu. Rev. Astron. Astrophys. 1999. 37:311-362
Copyright © 1999 by Annual Reviews. All rights reserved
The probes of physical conditions have been developed and are now fairly well understood. Some physical conditions have been hard to measure, such as the magnetic field strength, or hard to understand, notably the kinematics. Star-forming structures, or cores, within primarily sterile molecular clouds can be identified by thresholds in column density or density. Stars form in distinct modes, isolated and clustered, with massive stars forming almost exclusively in a clustered mode. The limited number of measurements of magnetic field leave open the question of whether cores are subcritical or supercritical, and whether this differs between the isolated and clustered mode. Cores involved in isolated star formation may be distinguished from their surroundings by a decrease in turbulence to subsonic levels, but clustered star formation occurs in regions of enhanced turbulence and higher density, compared to isolated star formation.
Evolutionary scenarios and detailed theories exist for the isolated mode. The theories assume cores with extended, power-law density gradients, leading to the mass accretion rate as the fundamental parameter. Detailed tests of these ideas are providing overall support for the picture, but also raising questions about the detailed models. Notably, kinematic evidence of gravitational collapse has finally been identified in a few cases. The roles of turbulence, the magnetic field, and rotation must be understood, and the factors that bifurcate the process into single or multiple star formation must be identified. Prospects for the future include a less biased census for cores in early stages and improved information on density gradients, both facilitated by the appearance of cameras at millimeter and submillimeter wavelengths on large telescopes. Antenna arrays operating at these wavelengths will provide more detailed information on the transition region between envelope and disk and study early disk evolution and binary fraction. Future, larger, antenna arrays will probe disk structure to scales of a few AU. Finally, a closer coupling of physical and chemical studies with theoretical models will provide more pointed tests of theory.
Theories and evolutionary scenarios are less developed for the clustered mode, and our understanding of the transition between isolated and clustered mode is still primitive. Current knowledge suggests that more massive cores have flatter density distributions and greater tendency to show substructure. At some point, the substructure dominates and multiple centers of collapse develop. With restricted feeding zones, the mass accretion rate gives way to the mass of clump as the controlling parameter, and some studies of clump mass spectra suggest that the stellar IMF is emerging in the clump mass distribution. The most massive stars form in very turbulent regions of very high density. The masses and densities are sufficient to form the most massive clusters and to explain the high stellar density at the centers of young clusters. They are also high enough to match the needs of coalescence theories for the formation of the most massive stars. As with isolated regions, more and better measurements of magnetic fields are needed, along with a less biased census, particularly for cool cores that might represent earlier stages. Larger antenna arrays will be able to separate clumps in distant cores and determine mass distributions for comparison to the IMF. Larger airborne telescopes will provide complementary information on luminosity sources in crowded regions. Observations with high spatial resolution and sensitivity in the mid-infrared will provide clearer pictures of the deeply embedded populations, and mid-infrared spectroscopy with high spectral resolution could trace kinematics close to the forming star. Deeper understanding of clustered star formation in our Galaxy will provide a foundation for understanding the origin and evolution of galaxies.
I am grateful to P André, J Bally, M Choi, F Motte, D Johnstone, and L Looney for supplying figures. Many colleagues sent papers in advance of publication and/or allowed me to discuss results in press or in progress. A partial list includes P André, R Cesaroni, R Crutcher, D Jaffe, L Mundy, E Ostriker, Y Shirley, J Stone, D Ward-Thompson, and D Wilner. I would like to thank R Cesaroni, Z Li, P Myers, F Shu, F van der Tak, and M Walmsley for detailed, helpful comments on an earlier version. This work has been supported by the State of Texas and NASA, through grants NAG5-7203 and NAG5-3348.