7 Summary

The paper discusses the carbon line emission detected in the low-resolution (Paper I) and the high-resolution surveys (Paper II) of recombination lines near 327 MHz made in the galactic plane using the ORT. The observed carbon line parameters in the high-resolution survey are presented. The carbon lines observed in the surveys arise in diffuse C  II regions unlike the high frequency ($\nu > 1$ GHz) recombination lines which arise in the photo-dissociation regions associated with H  II regions. Most of our carbon line detections are in the longitude range l = 358$^\circ$ $\rightarrow$ 20$^\circ$ with a few detections between l = 20$^\circ$ to 89$^\circ$. At longitudes l = 0$^\circ$ and 13 $.\!\!^\circ$9, observations with a 2$^\circ$ $\times$ 2$^\circ$ beam along galactic latitude indicate that line emission extends at least up to $b \sim \pm$ 3$^\circ$.

The l-v diagram and radial distribution constructed using the carbon line data near 327 MHz show similarity with those obtained from the hydrogen RRLs near 3 cm from H  II regions. The radial extent of carbon line emission also resembles that of "intense'' ¹²CO emission. These similarities suggest that the distribution of the carbon line forming regions resembles the distribution of star-forming regions in our Galaxy. The l-v diagram of carbon line emission near 327 MHz is similar to those obtained from the carbon lines detected in absorption at frequencies near 76 MHz (Erickson et al. 1995) and 35 MHz (Kantharia & Anantharamaiah 2001), indicating that the latter are the absorption counterparts of the carbon lines observed in emission near 327 MHz.

We estimated the [C  II] 158 ${\mu }$m intensity expected from low frequency carbon RRL forming regions coexistent with the CNM and low-density PDRs. The estimate was made using a subset of physical parameters ($T_{\rm e}$ = $20 \rightarrow 80$ K) determined by Kantharia & Anantharamaiah (2001) for the diffuse C  II regions. Significant fraction ($\sim$95%) of the observed FIR line emission can arise in CNM/low-density PDR if the temperature of the diffuse C  II region is close to $\sim$80 K whereas only a small fraction (0.4%) of the observed FIR line emission can be produced in regions with temperature near 20 K. We then compared the longitudinal distribution of the two tracers of ionized carbon in the inner Galaxy to investigate their common origin. Available data do not rule out the possibility that diffuse C  II regions contribute significantly to the FIR line emission in the inner Galaxy. However, data with comparable angular and spectral resolutions are required to further investigate this possibility.

Analysis of our high resolution data shows that the diffuse C  II regions exhibit structure over angular scales ranging from $\sim$6$^\prime$ to $\sim$5$^\circ$. Toward l=35$^\circ$ and several other directions, the diffuse C  II regions exhibit structure over $\sim$6$^\prime$, manifested as different central line velocities and line widths. At l = 35$^\circ$  we estimated a linear size of $\sim$6 pc for a clump when placed at the near kinematic distance. Thus, it is likely that there exist C  II regions of this size (or smaller) or there exist clumps of this size embedded in a larger diffuse C  II region. Toward l = 13 $.\!\!^\circ$9, a narrow ($\sim$7 km s^-1) carbon line emitting region with an angular extent $\ge$2$^\circ$ in galactic longitude and $\sim$$\pm$ 3$^\circ$ in galactic latitude has been observed. A similar extended diffuse C  II region was identified toward l = 4 $.\!\!^\circ$25, b= 0$^\circ$. The angular extent of this region is at least 5$^\circ$ along l and 2$^\circ$ along b. Such large angular sizes translate to physical sizes perpendicular to the sight-lines of >200 pc. These are fairly large diffuse C  II regions. Thus, our data shows that the diffuse C  II regions observed in the inner Galaxy display structure on scales ranging from a few parsecs to a couple of hundred parsecs.

The association of some of the observed carbon line emission near 327 MHz in the surveys with H  I self-absorption features will be discussed in Roshi et al. (2002).

Acknowledgements

DAR thanks F. J. Lockman and D. S. Balser for the many stimulating discussions and helpful suggestions during the course of this work. NGK thanks Rajaram Nityananda for useful suggestions. We are grateful to late K. R. Anantharamaiah for his guidance and support, to which we owe much of the work we have accomplished in our careers. We thank the referee F. Wyrowski for his comments and suggestions which have improved the paper.