Previous
abstract 
Next
abstract
The star formation rate in molecular clouds may result from an equlibrium between core collapse and cloud support by (proto-)stellar winds (McKee 1989). This feedback between the formation of stars and cloud support may even determine the stellar IMF (Silk 1994). O stars photoionize, disrupt, and eventually destroy clouds: the massive star formation rate may therefore determine a cloud's lifetime (Williams \& McKee, 1994).
Observed at sub-parsec scale in CO and its isotopes, molecular clouds appear highly fragmented and can be characterized as consisting of a number of discrete density enhancements or clumps. We have developed an algorithm to determine and analyze the clumpy structure of a spectral line data cube (Williams, de Geus, and Blitz, 1994), and have applied it to a number of molecular clouds with a variety of star formation histories (clouds that are forming massive stars, forming only low mass stars, absent of star formation).
As first found by Larson (1981), we demonstrate that scaling relations between clump mass, size and linewidth are constant from cloud to cloud. The clump mass spectrum is also observed to be universal, $dN/dM\propto M^{-1.5}$, and the same as the Galactic mass spectrum of molecular clouds. Clumps that are associated with star formation (via their coincidence with color-selected IRAS point sources) tend to be the most massive and most bound clumps in the parent cloud. However, the clumps in clouds that are forming massive (O) stars are, as a whole, more bound than those in other clouds. The degree of dynamical evolution, as measured, for example, by the clustering of high mass clumps relative to low mass clumps toward the center of a cloud is also observed to be greatest in the high mass star forming clouds. We discuss these similarities and differences in the context of cloud evolution and star formation.
