2006ApJ...646.1043M

We revisit the problem of the star formation timescale and the ages of molecular clouds. The apparent overabundance of star-forming molecular clouds over clouds without active star formation has been thought to indicate that molecular clouds are ``short-lived'' and that star formation is ``rapid.'' We show that this statistical argument lacks self-consistency and, even within the rapid star formation scenario, implies cloud lifetimes ~10 Myr. We discuss additional observational evidence from external galaxies that indicate lifetimes of molecular clouds and a timescale of star formation of ~10⁷. These long cloud lifetimes, in conjunction with the rapid (~1 Myr) decay of supersonic turbulence, present severe difficulties for the scenario of turbulence-controlled star formation. By contrast, we show that all 31 existing observations of objects for which the line width, the size, and the magnetic field strength have been reliably measured are in excellent quantitative agreement with the predictions of the ambipolar-diffusion theory. Within the ambipolar-diffusion-controlled star formation theory, the line widths may be attributed to large-scale nonradial cloud oscillations (essentially standing large-amplitude, long-wavelength Alfvén waves), and the predicted relation between the line width, the size, and the magnetic field is a natural consequence of magnetic support of self-gravitating clouds.