The transverse widths of both jets grow only slowly, if at all, with increasing distance from the nucleus, after they first brighten. Their full opening angles, ), are 0.85 ± 0.26° (jet) and 1.12 ± 0.27° (counterjet). These estimates are the averages of the regression slopes against of five different measures of : FWHM, and isophote separation, equivalent rectangular width, and the separation of the steepest transverse brightness gradients (Swain 1996).
Neither jet is prominent in polarized emission , but there are distinct, linear minima in P near both edges of both jets, giving an appearance of parallel dark "rails'' in the P images.
Figure 2 shows how these rails are related to total intensity (I) profiles taken transverse to the jets. Because neither jet nor the rail separation expands significantly between the bright knots J1 and J4, the jet profiles are averaged over this region. The counterjet profiles are averaged over the short feature CJ2. The "noise'' is less in the jet average because the averaging region is fourteen times longer than for the counterjet.
Figure 2: Transverse profiles of the total intensity I (solid line) and polarized intensity P (dotted line) of the jets in 3C353 at 8.4 GHz with 0.44" FWHM resolution, averaged along the jet (upper panel) in the region between J1 and J4 and in the counterjet (lower panel) in CJ2. NR and SR mark the positions of the north and south polarization minima ("rails'') associated with the jets.
Note the following features of Figure 2:
We also imaged 3C353 using the VLA at 1.4, 1.7 and 4.9 GHz at 13" resolution (Swain 1996). These images showed that the location and shape of the polarization rails, and their depth (as a fraction of the jet emission in their vicinity) change little between 1.4 and 8.4 GHz.