Photometrically distinct compact nuclei have been known for a while to reside in the centers of bulgeless, weakly active or inactive late-type disks (e.g., Kormendy & McClure 1993; Matthews & Gallagher 1997). Extensive surveys with the HST show, however, that distinct pointlike or compact nuclei are the rule rather than the exception in the centers of all sorts of disk galaxies.
An optical (V, WFPC2) and near-infrared (H, J, NICMOS) survey of ~ 100 intermediate-type, Sb-Sc spirals shows that ~ 70% of these systems host distinct nuclei with visual absolute magnitudes -8 MV -16, comparable at the bright end with young super star clusters in starbursting galaxies (Carollo et al. 1997a, 1998, 2002). Some of these nuclei appear pointlike in the HST images. However, many are marginally resolved with half-light radii ~ 0."1-0."2, corresponding to linear scales of a few to up to ~ 20 pc. These nuclei cover a large range of colors in the range -0.5 mag V - H 3 mag. Statistically, the distribution of V - H colors is broader for the nuclei than for the surrounding galactic structure; this suggests that star formation, AGN activity, dust reddening, or a combination of these is generally present in the nuclei embedded in intermediate-type hosts. Most of the nuclei, at HST resolution, are located at the galaxy centers and appear to be round, star-cluster-like, structures. However, in some cases the nuclei are offset from the isophotal/dynamical centers of the host galaxies or show some degree of elongation (as it is the case for, e.g., the nucleus of M33; Kormendy & McClure 1993; Lauer et al. 1998; Matthews et al. 1999). Both the displacement, which is typically of a few tens of parsecs, and the elongation appear to be uncorrelated with either the luminosity and color of the nucleus, or with the galaxy type. Searches for trends with other galactic subcomponents reveal no clear relationship between the distinct nuclei and nuclear (or even larger-scale) bars: bars are neither ubiquitous nor unusual in nucleated intermediate-type spirals. Some of the nuclei are embedded in exponential-type bulges; actually, every such bulge in this sample hosts a distinct nucleus in its isophotal center or slightly offset from it. These nuclei have low to moderate luminosities, -8 mag MV -12 mag. Selection effects may be present. Brighter nuclei are typically embedded in very complex circumnuclear structures that do not allow for the derivation of reliable bulge parameters. Moreover, exponential-type bulges lying under substantial circumnuclear structure but hosting no central nucleus would also, for similar reasons, drop from the sample. The nuclei that have been identified inside the exponential-type bulges have colors compatible with those arising from stellar populations. Under the assumption of no dust reddening, average stellar ages of ~ 1 Gyr are inferred for these relatively faint central star clusters; these ages, in turn, imply masses of about 106 to 108 M.
Studies of large samples of Scd and later-type disks also find compact, distinct nuclei, identified as "star clusters," in about 75% of the population (Böker et al. 2002). Their distribution of absolute luminosities has a FWHM of about 4 mag and a median value of MI = -11.5 mag. These luminosities, as well as the sizes of these star clusters, are comparable to those of the nuclei embedded in the earlier-type galaxies. For 10 nuclei in the Böker et al. (2002) sample, Walcher et al. (2003) report spectroscopic estimates for the stellar ages that are smaller than about 1 Gyr, and masses of the nuclei estimated from stellar velocity dispersions in the range 106 - 108 M. These masses are similar to the photometry-based mass estimates derived for the clusters embedded in the exponential-type bulges, and are one order of magnitude higher than expected from stellar synthesis models. These authors interpret this discrepancy as due to old(er) stellar generations contributing to the total mass of the nuclei. They conclude that the nuclei of late-type disks are grown in multiple star formation events, and suggest that this mechanism may contribute to the formation of Kormendy's pseudo-bulges.
At the other extreme of the Hubble sequence, the large majority of the early-type spiral galaxies that host massive bulges also require a central component, in addition to the bulge and the disk, to reproduce the observed light profiles (Balcells et al. 2003). In most cases, this additional component appears to be a point source at the resolution of NICMOS on the HST. Balcells et al. point out that in ground-based data these additional light contributions cannot be disentangled from the bulge light and may produce overestimates of the derived Sérsic n parameter (see Section 1.3).