2016ApJ...833....8G

We reexamine the dynamical evolution of the mass and luminosity functions of globular star clusters (GCMF and GCLF). Fall & Zhang (2001, FZ01) showed that a power-law MF, as commonly seen among young cluster systems, would evolve by dynamical processes over a Hubble time into a peaked MF with a shape very similar to the observed GCMF in the Milky Way and other galaxies. To simplify the calculations, the semi-analytical FZ01 model adopted the "classical" theory of stellar escape from clusters, and neglected variations in the M/L ratios of clusters. Kruijssen & Portegies Zwart (2009, KPZ09) modified the FZ01 model to include "retarded" and mass-dependent stellar escape, the latter causing significant M/L variations. KPZ09 asserted that their model was compatible with observations, whereas the FZ01 model was not. We show here that this claim is not correct; the FZ01 and KPZ09 models fit the observed Galactic GCLF equally well. We also show that there is no detectable correlation between M/L and L for GCs in the Milky Way and Andromeda galaxies, in contradiction with the KPZ09 model. Our comparisons of the FZ01 and KPZ09 models with observations can be explained most simply if stars escape at rates approaching the classical limit for high-mass clusters, as expected on theoretical grounds.