Black holes are ubiquitous in the nuclei of both active and non-active galaxies (e.g., Kormendy & Richstone 1995), and are thought to be at the direct origin of non-stellar nuclear activity (e.g., Lynden-Bell 1969; Begelman, Blandford & Rees 1984). The fact that the mass of the central supermassive black hole (SMBH) in a galaxy is correlated with the velocity dispersion of the bulge, and hence with its mass (Ferrarese & Merritt 2000; Gebhardt et al. 2000) provides the most tangible link between the nuclear regions and their host galaxies. But because not all galaxies with SMBHs have AGN characteristics, the presence of an SMBH in itself cannot be enough to make a galaxy "active", at least not continuously, and additional mechanisms must be considered which can ignite the nuclear activity.
In the case of starburst galaxies, defined rather loosely as galaxies which show abnormally enhanced massive star formation activity in their central regions (or in some more extreme cases throughout the galaxy), a similar question can be posed, namely what ignites the starburst. In both the AGN and the starbursts, the availability of fuel at the right place and at the right time must play a critical role. Such gaseous fuel is plentiful in the disks of galaxies, but must lose significant quantities of angular momentum in order to move radially inward. In fact, the "fuelling problem" is not the amount of fuel that is available, but how to get it to the right place, as graphically illustrated by Phinney (1994, his fig. 1). Estimates for the mass accretion rate needed to fuel AGN vary from around 10-4 M / year for low-luminosity AGN such as LINERs, up to around 10 M / year for high-luminosity AGN such as QSOs, or, over a putative lifetime of 108 years for the AGN activity, only 104 to 109 M.
Large stellar bars, as well as tidal interactions and mergers, have some time ago been identified as prime candidates to drive gas efficiently from the disk into the inner kpc (see next Section). In this review we will concentrate on the observational evidence, mostly statistical in nature, for the effectiveness of these gravitational mechanisms, concentrating on the effects of bars in Section 2, and on those of interactions in Section 3. Galactic rings are considered in Sections 4 and 5, and summarising remarks are given in Section 6. Related reviews considering the fuelling of primarily AGN include those by Shlosman, Begelman & Frank (1990), Beckman (2001), Combes (2001), Shlosman (2003), Wada (2004), and Jogee (2004).