Discussion

Our Chandra observations have revealed three striking aspects of the 3C 401 system. Firstly, there is clear evidence for an ongoing interaction between 3C 401 and the ICM of its surrounding galaxy cluster. The most obvious manifestation of this interaction is the nuclear bar (with radius of $\sim 10\arcsec$/$35{\rm\thinspace kpc}$) that lies orthogonal to the axis of the MERLIN radio lobes. Guided by observations of radio-galaxy/ICM interactions in the local universe ($z<0.1$), it seems very likely that this structure results from the formation of ICM cavities by the expanding radio-lobes. The detailed anti-coincidence between the X-ray surface brightness and the radio surface brightness seen in 3C 401 supports this hypothesis. Using our estimates of the ICM pressure, we can make a crude estimate for the mechanical power required to inflate these radio lobes. The total energy required to inflate the two lobes is $E\sim 1.5\times 10^{59}{\rm\thinspace erg}$, where we have approximated each lobe as a sphere with radius $5\arcsec$ and have taken the required energy to be $E\sim 2pV$ where $V$ is the volume of the lobe. The lifetime of the source is likely to be of the order of the ICM sound crossing time of one radio lobe, $t\sim 1.7\times 10^{15}{\rm\thinspace s}$. Thus, the time-averaged power required to inflate these radio-lobes against the pressure of the ICM is $P=E/t\sim 1\times 10^{44}\hbox{${\rm\thinspace erg}{\rm\thinspace s}^{-1}\,$}$. This is very close to the measured X-ray luminosity of the ICM showing that the radio galaxy can have a major impact on the energetics of the ICM in this source if the mechanical energy can be thermalized efficiently. Taking the spatially integrated 1.4GHz flux to be $4.7\times 10^{-23}\hbox{${\rm\thinspace erg}{\rm\thinspace s}^{-1}\,$}$ (Kellerman, Pauliny-Toth & Williams 1969), we estimate the radio power to be $\nu L_\nu\approx 7\times 10^{42}\hbox{${\rm\thinspace erg}{\rm\thinspace s}^{-1}\,$}$. This places the 3C 401 cluster near the correlation between the mechanical energy inferred to be inside these cavities and the current radio power level (Bîrzan et al. 2004). Consistency with this correlation means that the current radio source contains sufficient mechanical energy to create these cavities (assuming the theoretical expectation of a 1-10% efficiency for converting the total radio source mechanical energy into radio frequency luminosity; Bicknell et al. 1997).

Secondly, we have noted a larger scale cross-like structure extending to distances of $\sim 50\arcsec$ ($170{\rm\thinspace kpc}$) from the centre of the cluster and also aligned with the radio-axis of 3C 401. While the reality of this feature appears to be robust, its interpretation is not clear. The coincidence between the orientation of this structure and the radio axis of 3C 401 suggests that this might also be due to radio plasma interaction with the ICM, although the possibility remains that the cross-like structure is caused by unrelated dynamical processes (e.g., subcluster mergers). If it is indeed due to interaction with 3C401, two possibilities arise. If this ICM structure is caused by two pairs of ``ghost cavities'', then they are amongst the largest known. Using the same assumptions as in the paragraph above to assess the energetics of these ghost cavities, we estimate that 3C401 had to have a period about 300Myr ago in which its mechanical power was $2-5\times 10^{44}\hbox{${\rm\thinspace erg}{\rm\thinspace s}^{-1}\,$}$ (i.e., a factor of a few greater than the present). While this explanation for the ICM cross has the appeal that ghost cavities are structures that are known to exist in some clusters, it does not naturally explain the four-fold symmetry of this structure (the two pairs of ghost cavities would have to lie at roughly the same distance from 3C 401 and have axes that are perpendicular).

This leads us to speculate that the ICM cross is actually due to a high amplitude global oscillation mode (most likely a low-$l$ internal gravity mode) excited by a previous outburst from 3C401. The theory of such oscillations has been developed by Balbus & Soker (1990) and Lufkin, Balbus and Hawley (1995), although these authors envisage the excitation of internal g-modes through a resonant interaction with orbiting galaxies, not through an explosive central event. A detailed theoretical investigation of this possibility, including predicted maps of ICM surface brightness and temperature for different modes, is beyond the scope of this paper. At this stage, we note that the oscillation period of such a mode will be a factor of a few longer than the sound crossing time of the region, and the energy of the mode will be of the same order as that estimated above for the ghost cavity scenario. However, it is likely that only a modest fraction of the total energy from the radio-galaxy outburst would end up in such a mode, with p-modes likely carrying away the majority of the energy of the initial blast. Hence, within this scenario, 3C 401 would likely have exceeded a mechanical luminosity of $10^{45}\hbox{${\rm\thinspace erg}{\rm\thinspace s}^{-1}\,$}$ during its past phase of activity. Deeper imaging of this field (most likely with XMM-Newton will be required to study the ICM cross structure in more detail and distinguish between (or disprove) the ghost cavity and global mode scenarios.

The final striking aspect of this system is the unusual surface brightness distribution of the cluster -- we measure a core radius of $r_0=36{\rm\thinspace kpc}$ and $\beta=0.46$, substantially flatter than the typical $\beta=0.67$ found in many clusters of a comparable or greater mass. However, this is quite similar to the value of $\beta$ found in low mass clusters and groups (Osmond & Ponman 2004 and references therein), a result that is taken as evidence for the enhanced importance of excess entropy in these low mass systems. Whether the 3C 401 cluster really is anomalous in having a flat profile for its mass requires further study with deeper X-ray imaging. In particular, one may be concerned that we are not obtaining a true measure of the value of $\beta$ given that the central region of the cluster is morphologically complex and that we cannot constrain the ICM surface brightness profile beyond about 170kpc. If the flat profile is confirmed, it is tempting to interpret this as signs of particular strong ICM heating and entropy injection, as might be expected for a cluster whose ICM is still in the process of forming. Indeed, we proposed to observe this radio galaxy/ICM system with Chandra because previous work (Harvanek et al. 2001; Harvanek & Stocke 2002) presented significant evidence linking intermediate FR I/FR II radio galaxies to the formation of a dense ICM in the cluster which surrounds them.