next previous
Up: Studies of dense

4. Discussion

 

4.1. Ammonia as a tracer of dense gas

The ammonia emission was detected in about half of the sources in the range of right ascension of tex2html_wrap_inline2313. This is comparable with the detection rate obtained in Paper I for the outer Galaxy (37%) but the number of sources in the present study is certainly insufficient for statistical conclusions.

Comparing the detected and non-detected sources we should mention two apparent differences between these groups: (1) the non-detected sources are significantly more distant on the average than the detected ones (the mean distances are tex2html_wrap_inline2315 and tex2html_wrap_inline2317, respectively) and (2) the detection rate for the sources with outflows is much higher than for the sources without reported outflows (8 from 9 and 3 from 8, respectively). However, the second item can easily be a consequence of the first one because it should be more difficult to detect outflows in more distant objects. Therefore, we cannot conclude whether the detection rate is related to the outflow activity. Perhaps the differences can be explained by the beam dilution effects. Further we shall concentrate mostly on the detected objects.

The masses derived from the column densities are remarkably close to virial masses in most cases. It shows in particular that our assumption of the ammonia abundance, tex2html_wrap_inline2319, is in general correct (at least for the detected clouds). The striking exception is S 88 B. The mass derived for this source is an order of magnitude lower than the virial mass. This may indicate that the ammonia abundance here is correspondingly lower than in most of the other detected cores. The map of the ammonia emission in S 88 B (Fig. 1 (click here)) shows a ``hole" at the IR source position where we detected no ammonia at all. This resembles the distribution of the CO emission which has a horseshoe-like structure (White & Fridlund 1992) with a significantly reduced intensity of the CO line in approximately the same region. However, it is worth noting that the CS emission towards the IR source observed with a similar beam size (though not mapped) is rather strong (Paper II).

For explanation of the apparent ammonia underabundance it might be important that the IRAS point source associated with S 88 B is very luminous (tex2html_wrap_inline2321), second in our sample by the luminosity. The most luminous source in the sample is S 100 (K 3-50) and we detected no ammonia emission in this object (earlier Churchwell et al. 1990) observed a weak ammonia emission here; the intensity was lower than our detection limit). However, K 3-50 is a strong source of molecular emission in the lines of other high density tracers, e.g. HCN J=1-0 (Burov et al. 1988) and CS J=7-6 (Plume et al. 1992).

For S 87 and S 255 the virial masses also exceed (by a factor of 2) the masses derived from the ammonia column densities. Both these sources have pronounced bipolar structure and can be really unbound. It is worth noting also that the estimates of the virial masses in the case of their highly elongated geometry should be made in a more refined way.

The difference between the ammonia distribution on the one hand and CS and dust distribution on the other hand in S 255 (which is also very luminous) can also be attributed to the ammonia underabundance in the neighbourhood of the central source.

Thus, we suspect that ammonia is underabundant in the vicinity of the most powerful IR sources. Of course, this is not a decisive conclusion because the number of the investigated objects is too limited and the beam is pretty large. Recently Davis & Dent (1993) suggested that ammonia is underabundant towards young stars being the central sources of molecular outflows but did not relate this effect with the IR luminosity. However, many of our sources demonstrate outflow activity but mainly the most luminous ones have signs of ammonia underabundance.

What can be the reason for the ammonia underabundance? Davis & Dent (1993) suggested that there is a mechanism which restricts ammonia formation in such places. In our opinion there are basically two ways to explain the possible ammonia underabundance:

(1) The time-dependent chemical models show that the ammonia abundance is relatively low at early times (e.g. Bergin et al. 1995; Taylor et al. 1996). Thus, the ammonia underabundance can be related to the evolutionary status of the clouds. However, there is no direct relation to the luminosity of the central source in this model.

(2) An enhanced radiation field in the vicinity of young stellar objects increases the thickness of the tex2html_wrap_inline2327 layer and tex2html_wrap_inline2329 is destroyed primarily by a reaction with tex2html_wrap_inline2331 (e.g. Turner 1995). At the same time the species like CO, CN, tex2html_wrap_inline2333, etc. which are formed via tex2html_wrap_inline2335 are still very abundant in these regions in spite of the high photodestruction rates (Bergin et al. 1995; Jansen et al. 1995). Probably this difference in their chemistry under the conditions of enhanced radiation field is the most plausible explanation for the relative ammonia underabundance.

4.2. Spatial structure and kinematics

About half of the mapped sources show elongated structures with a more or less significant velocity gradient along the major axis. The most prominent example is S 255. S 87 has a two-component structure with 2 distinct velocity components which overlap partly spatially. However, as it has been argued above ammonia probably does not reflect the dense gas distribution correctly especially in these cases. Anyway, there is a question what is the reason for this appearance. The most probable explanations are a rotating disk or an outflow. Barsony (1989) argued in favour of the latter model for S 87. We shall discuss the data on these sources in more detail in a separate paper.

Another important item in respect of the core kinematics is the variations of the velocity dispersion in the cores. The CS data for southern cores with an apparently spherical structure indicate that the velocity dispersion decreases outwards with radius (Zinchenko 1995). The present ammonia data are not so definite. In S 199, S 231, RNO 1B and BFS 48 the line widths increase significantly near the ammonia peaks. In S 87 the widths of the separate components also increase near their corresponding emission peaks. In the other cases there are either no regular variations (S 86) or gradients of the line widths along certain directions can be seen (S 76E, S 88B, S 255). But most of the cores in the present sample are far from the spherical symmetry.

4.3. Temperature

Considering the kinetic temperature distribution in the cores we see that in most cases it is centrally peaked reaching tex2html_wrap_inline2337 and drops to tex2html_wrap_inline2339 at the edges of the ammonia emitting regions. These values refer to the 40tex2html_wrap2343 beam of the Effelsberg antenna and do not exclude a presence of hotter compact clumps. Probably, in the outer layers of the clouds which are not traced by ammonia, the temperature rises again as indicated by the fact that the CO temperatures in many cases are higher than the kinetic temperatures derived from ammonia data.

4.4. Hyperfine intensity anomalies

Finally, we have looked for hyperfine intensity anomalies in the (1, 1) transition in our sources. In addition to S 87 where such anomalies have been seen by Stutzki et al. (1984) we found similar deviations in S 199 (Fig. 5 (click here)). These anomalies are a rather common phenomenon in warm clouds (Stutzki & Winnewisser 1985) and have been interpreted in the framework of clumpy models with the individual clumps having small line widths of tex2html_wrap_inline2345 (see Stutzki & Winnewisser 1985 and references therein).

  figure524
Figure 5: The tex2html_wrap_inline2347 (1, 1) line in S 199 showing hyperfine intensity anomalies


next previous
Up: Studies of dense

Copyright by the European Southern Observatory (ESO)
web@ed-phys.fr