In contrast to the radio continuum emission from Mkn 6, which is extended and highly structured (e.g., Kukula et al. 1996; Fig. 1), H I absorption is detected only towards component 6, a compact source located at the northern end of the arcsecond-scale radio jet (Fig 1; component numbering following Kukula et al. 1996). Discussed further below, the linewidth is very narrow in comparison with H I absorbed radio jets in other Seyfert galaxies; formally, the linewidth (FWHM) is km s-1 (corrected for the instrumental resolution) and the maximum opacity is . The integrated absorption profile corresponds to a foreground column of
where is the spin (excitation) temperature of the ground state. This column is not unusual for a sight-line through an inclined disk galaxy. However, we note that a similar column, detected in NGC 4151, was interpreted as absorption in a nuclear torus (Mundell et al. 1995).
The limits placed by non-detections better define the localisation of the H I absorption around component 6. Towards the brighter regions of the southern jet, components 2-4, the () limit is , corresponding to a foreground column density
The absorbing gas would easily have been detected had the gas completely covered the jet. On the other hand, component 5, which is the nearest neighbor to the absorbed component, is much fainter, and so the limits are less stringent: , or
We can conclude is that the H I absorbing gas covers a region including component 6 and extending no further south than component 5, or roughly 0:0075 (280 pc in projection). However, we can place no limits on the extent of the absorbing gas in other directions.
The centroid velocity of the absorption line is km s-1, blue-shifted relative to systemic by km s-1. For comparison, the position-velocity curve is plotted in Fig. 2. The details of the rotation curve within the inner few arcseconds are unknown, but the velocity of the 21 cm absorption line does not appear significantly displaced from any plausible rotation curve. We conclude that the absorption line arises in otherwise normally rotating gas, and there is no evidence for streaming motions greater than km s-1. Furthermore, we do not detect any velocity gradients across component 6. Assuming that the absorbing gas completely covers the background source (Sect. 4), the upper limit for the velocity gradient is approximately the width of the absorption line divided by the component size ( ; Kukula et al. 1996), or km s-1 pc-1. For comparison, the projected velocity gradient of the H I absorption seen towards NGC 4151 is km s-1 pc-1 (Mundell et al. 1995).
© European Southern Observatory (ESO) 1998
Online publication: April 15, 1998