The Initial Mass Function of the Galactic Bulge Down to ~0.15 Msun

We present a luminosity function (LF) for lower main sequence stars in the Galactic bulge near (l,b)=(0,-6) to J=24, corresponding to M_J~9.3. This LF is derived from HST + NICMOS observations with the F110W and F160W filters. The main sequence locus in the infrared shows a strong change in slope at J~20.5 (M_J~5.75) which is well fit by new low mass models that include water and molecular hydrogen opacity. Our derived mass function (which is not corrected for binary companions) is the deepest measured to date in the bulge, and extends to 0.15 $M_\odot$ with a power law slope of $\alpha=-1.33\pm0.07$; a Salpeter mass function would have $\alpha=-2.35$. We also combine our J band LF with previously published data for the evolved stars to produce a bulge LF spanning ~15 magnitudes. We show that this mass function has negligible dependence on the adopted bulge metallicity and distance modulus. Although shallower than the Salpeter slope, the slope of the bulge IMF is steeper than that recently found for the Galactic disk ($\alpha=-0.8$ and $\alpha=-0.54$ from the data of Reid & Gizis, 1997, and Gould et al. 1997, respectively, in the same mass interval), but is virtually identical to the disk IMF derived by Kroupa et al. (1993). The bulge IMF is also quite similar to the mass functions derived for those globular clusters which are believed to have experienced little or no dynamical evolution. Finally, we derive the $M/L_J$ ratio of the bulge to be $\sim 0.9\pm0.1$ $M_\odot/L_\odot$, and briefly discuss the implications of this bulge IMF for the interpretation of the microlensing events observed in the direction of the Galactic bulge.