Results

1. We detect at least a single component in all the observed objects. Examples with multiple components are shown in Fig. 1. We find no systematic difference in compactness between Sy 1s and 2s unlike the result of Roy et al.$\;$(1994). The distributions of the detected pc-scale luminosities are also similar.

Figure: VLBI images: Sy 1s (top panels); Sy 2s (bottom panels). Synthesized beams are boxed at bottom-left. The contour levels are given in % of the peak.

3pt

2. The distributions of the ratios of the pc- to kpc-scale flux densities are similar for the two Seyfert subclasses. If their radio jets were relativistically beamed, and if the detected pc-scale feature is the nuclear jet, then, the purportedly pole-on Sy 1s would have a beamed or enhanced radio nucleus relative to the presumably unbeamed kpc-scale emission, and therefore systematically higher values of these ratios.

3. In at least one case, $viz.,\;$NGC5929, the detected compact feature certainly corresponds to one of the two kpc-scale hotspots of the triple (Su et al.$\;$1996) rather than the nucleus, implying that the compact features that we see might sometimes be the termination points of the jet.

4. The projected kpc-scale linear sizes of the Sy 1s and 2s are not systematically different. However, there is a significant correlation between linear size and total radio luminosity (Fig. 2). It is likely that this correlation swamps observable size differences between the two subclasses for this sample. It is also possible that caveat (2) above contributes to this effect in that, the radio jets from the Sy 2s in the sample encounter systematically longer path-lengths through the ISM of the host galaxy than the radio jets of the Sy 1s in the sample, leading to a relatively higher quenching effect in these Sy 2s.

Figure: The correlation of linear size with radio luminosity.