M32 ± 1

Multicolor HST WFPC2 images are used to study the central structure of the three Local Group galaxies M31, M32, and M33. PSF deconvolution and modeling of image aliasing are required to recover accurate brightness profiles within r < 0″.5. The data present a study in contrasts that suggests different evolutionary histories. In M31, the nucleus is double-peaked, as found by WFPC1 and confirmed by WFPC2. The dimmer peak, P2, is closely centered on the bulge isophotes to 0″.1, implying that it is the dynamical center of the galaxy. Directly on P2 lies a UV-bright compact source that was discovered by King, Stanford, & Crane at 1700 Å. WFPC2 images now show that this source is resolved, with r_1/2 ≈ 0.2 pc. It dominates the nucleus at 3000 Å, and its spectral energy distribution is consistent with late B-early A stars. This probable nuclear star cluster may consist of young stars and be an older version of the central cluster of hot stars that now sits at the center of the Milky Way, or it may consist of heavier stars built up from collisions in a possible cold disk of stars orbiting P2. Aspects of its formation remain highly problematic. In M32, new images show that the central cusp continues to rise into the HST limit with slope y ≈ 0.5 and the central density ρ₀ > 10⁷ M_⊙ pc^-3. The V-I and U-V color profiles are essentially flat, and there is no sign of an inner disk, dust, or any other structure. This total lack of features seems at variance with a nominal stellar collision time of 2 × 10¹⁰ yr, which implies that a significant fraction of the light in the central pixel should come from blue stragglers. This discrepancy is eased but not completely removed if the stellar population is young (2 × 10⁹ yr). The stubborn normalcy of M32 at tiny radii may be emerging as an important puzzle. In M33, the nucleus has an extremely steep γ = 1.49 power-law profile for 0″.05 < r < 0″.2 that appears to become somewhat shallower as the HST resolution limit is approached. The profile for r < 0″.04 can be described as having either a γ ≈ 0.8 cusp or a small core with r_c ≈ 0.13 pc. The central density is ρ₀ > 2 × 10⁶ M_⊙ pc^-3, and the implied central relaxation time is only ∼3 × 10⁶ yr, indicating that the nucleus is highly relaxed. The accompanying short collision time of 7 × 10⁹ yr predicts a central blue straggler component that is quantitatively consistent with the strong V-I and B-R color gradients seen with HST and from the ground. When combined with the Galaxy, the nuclei of the Local Group show surprisingly similar radial density profiles but divide into two classes according to velocity dispersion and black hole content: M31, M32, and the Milky Way are dominated dynamically (and stabilized against relaxation) by massive central black holes, while M33 lacks a dominant black hole. An obvious hypothesis is that a sizable stellar spheroid (which M33 lacks) is necessary to grow a massive black hole. A further implication is that the black hole growth in M31, M32, and the Milky Way was accompanied by evolution in the stellar density profiles, stellar populations, and dynamical structure of these nuclei such that their past appearance may have differed significantly from what they look like today. In short, HST observations are taking us to scales where understanding the central structure of galactic nuclei is intimately connected to the detailed interactions among their central stellar populations.