1 Erratum to: Astron Astrophys Rev (2016) 24:10 DOI 10.1007/s00159-016-0094-x

Following comments by a colleague, the author wishes to reformulate the last paragraph on page 25 (extending to page 26) of his review article. It should read:

Until recently, most studies of hot-mode accretion assumed that the hot gas accretes as a spherical, unperturbed flow at the Bondi rate (Bondi 1952). However, recent simulations that include realistic prescriptions for cooling, heating and turbulence, provide evidence that much of the hot gas in the nuclear regions on a scale of \({\sim }\)0.1 to 1 kpc can condense into filaments of cool gas that fall near-radially towards the nucleus, with cloud–cloud collisions helping to cancel any angular momentum in the gas (e.g. Gaspari et al. 2013, 2015). As a result, the accretion rates associated with such chaotic cold accretion (CCA) can be orders of magnitude higher than the Bondi rate. The CCA mechanism—which blurs the distinction between hot- and cold-mode accretion—is capable of providing more than enough fuel to power WLRG/FRI activity, but it is unlikely that the associated mass accretion rates would be sufficient to power the luminous quasar-like activity observed in many SLRG/FRII sources (M. Gaspari, private communication).