Conclusions

The polarization "rails'' at the edges of 3C353's jets result from vector cancellation between polarized jet emission and orthogonally polarized lobe emission. The flat-topped I profiles and the apparently parallel () magnetic field in the outer layers of 3C353's jets resemble features of other well-resolved FRII jets (Cygnus A: Carilli et al. 1996; 3C219: Bridle, Perley, & Henriksen 1984, Perley, Bridle, & Clarke 1994). We therefore support the suggestion (Bridle & Perley 1984) that much of the emission from FRII jets originates in a thick shear layer. The emissivity distribution deduced for 3C353 could be explained by in situ particle acceleration in the shear layer (as proposed for M87 by Owen, Hardee, & Cornwell 1989) or by Doppler hiding of emission from the jet "spine". Doppler hiding would fit the following view of relationships between FRI and FRII jets:

1. All AGN radio jets initially contain a relativistically-moving spine surrounded by a slower-moving shear layer (Laing 1993, Laing 1996).
2. The shear layer contains and roughly in equipartition, but little ; the spine contains little .
3. The spine decelerates to sub-relativistic speeds a few kpc from the nucleus in FRI sources, but not in FRII sources (Bicknell 1995).
4. In FRII sources, the observed "jet'' emission comes mostly from the shear layer, but in FRI sources the spine emission becomes visible where the spine decelerates, a few kpc from the AGN.

If this view is correct, then similarly well-resolved radio jets in FRII quasars will be more strongly center-brightened, and their counterjets more strongly limb-brightened, than those in FRII radio galaxies, because the Doppler asymmetries are greater in jets that are closer to the line of sight.

We thank Robert Laing and the referee for helpful advice and comments.