2004MNRAS.349.1419C

We have analysed the optical and radio properties of a sample of 3CR FR I radio galaxies which have Hubble Space Telescope (HST) imaging capable of detecting optical cores. The jet powers of the FR I radio galaxies are estimated from their low-frequency radio luminosities, and the optical core luminosity is taken as an upper limit on the emission from any unobscured accretion disc. We argue that if the accretion discs in these sources are assumed to be advection-dominated accretion flows (ADAFs), or adiabatic inflow-outflow solution (ADIOS) flows, then the Blandford-Znajek mechanism provides insufficient power to explain the high radio luminosities of at least a third, and perhaps all, of the sample. We suggest instead that a significant fraction (the `high-jet-power' third), and perhaps most, of the 3CR FR I radio galaxies have normal accretion discs, but that their optical cores can be hidden, with any HST -detected optical synchrotron emission coming from jets on scales larger than the obscuring material. A normal accretion disc hypothesis, at least for the high-jet-power third of the 3CR FR Is, explains why narrow-line luminosity correlates with radio luminosity. It also explains why one object in the sample (3C 386) has an observed broad-line nucleus. We conclude that there is no evidence to suggest that there is a difference in accretion mode between FR I and FR II radio galaxies.