1996A&A...313..929M

We present 21cm line observations of two galactic cirrus clouds, the Polaris Flare and LVC127, and compare the HI emission with that of CO and with emission in the IRAS bands at 100µm and 60µm. The 100µm emission per hydrogen nucleus is constant outside, but lower inside the CO emitting volume. We also find an infrared excess emission over and above what can be related to HI or CO emission. We conclude that apart from the diffuse atomic hydrogen and the denser molecular hydrogen traced by CO emission, there is a third gas component of diffuse molecular hydrogen. In places the diffuse H₂ has a column density rivalling that of HI. In general it does not contain a denser CO emitting interior. The fact that the diffuse H₂ is not traced by CO emission is due to the lower gas density, either because CO is not collisionally excited, or because the H₂ is not effective enough in shielding CO molecules from photo-dissociation. We find a conversion factor X between CO line integral and H₂ column density similar to other values from cirrus clouds, but lower than values for galactic plane clouds derived from gamma ray studies. If we account for the lower 100µm emission per hydrogen in the CO emitting volume, the difference is reduced, and our X may be consistent with the lower values found in a recent gamma ray study.