Finding black holes with microlensing

The MACHO and OGLE collaborations have argued that the three longest duration bulge microlensing events are likely caused by nearby black holes, given the small velocities measured with microlensing parallax and nondetection of the lenses. However, these events may be due to lensing by more numerous lower mass stars at greater distances. We find a posteriori probabilities of 76%, 16%, and 4% that the three longest events are black holes, assuming a Salpeter initial mass function (IMF) and a 40 M-circle dot cutoff for neutron star progenitors; the numbers depend strongly on the assumed mass function but favor a black hole for the longest event for most standard IMFs. The longest events (>600 days) have an a priori similar to26% probability of being black holes for a standard mass function. We propose a new technique for measuring the lens mass function using the mass distribution of long events measured with the Advanced Camera for Surveys on the Hubble Space Telescope, the Very Large Telescope Interferometer, the Space Interferometry Mission, or the Global Astrometric Interferometer for Astrophysics.