We now intend to model other resolvable bright jets in FRI radio galaxies to determine the extent to which their observed brightness and polarization properties resemble those of 3C31. We expect to be able to infer their velocity, emissivity and magnetic-field distributions, building on the broad success of the jet-deceleration model in accounting for the statistical asymmetries of the B2 sample of FRI sources (Laing et al.1999). Other sources showing well-collimated inner jets and rapid flaring include NGC315 (Venturi et al. 1993), PKS1333-33 (Killeen, Bicknell & Ekers 1986) and 3C449 (Feretti et al. 1999), and it seems likely that the regimes of collimation behaviour we have identified in 3C31 are common in FRI sources. We aim to study a sample of sources with a range of angles to the line of sight, if possible distributed isotropically, in order to test the results of Section 6. We also plan to develop a more sophisticated error analysis in order to assess confidence levels with some degree of rigour.
3C31 has been cited as the archetypal FRI source, but is actually in the minority in having diffuse ``tails'' of emission extending to large distances from the core rather than confined bridges analogous to the lobes of FRII sources (De Ruiter et al. 1990). Significant differences in dynamics (especially entrainment of the surrounding medium) might be expected between the two classes. We also expect that the deceleration process should depend on the jet power and the external environment.
In Laing and Bridle (2002), we present a dynamical model for the jets in 3C31, based on the velocity field derived in the present paper, a description of the surrounding galactic atmosphere derived from Chandra and ROSAT observations (Hardcastle et al.2002) and application of conservation laws following Bicknell (1994). This approach should also be extensible to other sources. Our results favour entrainment across the boundary layer as the origin of the majority of the mass-loading of the jets in 3C31, but it will be important to explore this in other large-scale FRI radio galaxies. We should seek further evidence for the entrainment process, such as the reduced polarization near the boundaries of the flaring regions.
Our ultimate goal is to replace the empirical descriptions of velocity, emissivity and field structure with realistic physical models. Although this is some way off, we have developed a self-consistent adiabatic model which can handle arbitrary field configurations and (laminar) velocity fields in a relativistic jet, with the aim of establishing whether any of the flow regions we have identified can be described in this way.
If our interpretation of the emission from the inner region of 3C31's jets is correct, observations of the apparent brightness and motions of FRI jets on even smaller scales will not be sensitive to the properties of the underlying bulk flow, but only to those of its slowest-moving components, which may be essentially stochastic. Improved transverse resolution of the inner jets in such sources will be required to determine the origin and distribution of the slow-moving material, and the extent to which these innermost regions of FRI jets resemble the larger-scale jets in FRII sources, e.g. those in 3C353 (Swain, Bridle & Baum 1998). This will require greater sensitivity and longer baselines than are currently available with the VLA or MERLIN.
Finally, a number of FRI sources (including 3C31) have been detected at X-ray and/or optical wavelengths (e.g. Sparks et al. 2000 ; Hardcastle, Birkinshaw & Worrall 2001 ; Marshall et al. 2002 ,; Perlman et al. 2001 ,; Hardcastle et al. 2002 ). The radiation is most plausibly produced by the synchrotron process over the entire observed frequency range, and the shape of the spectrum therefore carries information about particle acceleration and energy loss. It will be important to incorporate descriptions of these processes into our models.