UPPER LIMITS ON THE MASSES OF 105 SUPERMASSIVE BLACK HOLES FROM HUBBLE SPACE TELESCOPE/SPACE TELESCOPE IMAGING SPECTROGRAPH ARCHIVAL DATA

Based on the modeling of the central emission-line width measured over subarcsecond apertures with the Hubble Space Telescope, we present stringent upper bounds on the mass of the central supermassive black hole, M-center dot, for a sample of 105 nearby galaxies (D < 100 Mpc) spanning a wide range of Hubble types (E-Sc) and values of the central stellar velocity dispersion, sigma(c) (58-419 km s(-1)). For the vast majority of the objects, the derived M-center dot upper limits run parallel and above the well-known M-center dot-sigma(c) sc relation independently of the galaxy distance, suggesting that our nebular line-width measurements trace rather well the nuclear gravitational potential. For values of sigma(c) between 90 and 220 km s(-1), 68% of our upper limits falls immediately above the M-center dot-sigma(c) relation without exceeding the expected M-center dot values by more than a factor 4.1. No systematic trends or offsets are observed in this sc range as a function of the galaxy Hubble type or with respect to the presence of a bar. For 6 of our 12 M-center dot upper limits with sigma(c) < 90 km s(-1), our line-width measurements are more sensitive to the stellar contribution to the gravitational potential, either due to the presence of a nuclear stellar cluster or because of a greater distance compared to the other galaxies at the low-sigma(c) end of the M-center dot-sigma(c) relation. Conversely, our M-center dot upper bounds appear to lie closer to the expected M-center dot in the most massive elliptical galaxies with values of sigma(c) above 220 km s(-1). Such a flattening of the M-center dot-sigma(c) relation at its high-sigma(c) end would appear consistent with a coevolution of supermassive black holes and galaxies driven by dry mergers, although better and more consistent measurements for sigma(c) and K-band luminosity are needed for these kinds of objects before systematic effects can be ruled out.