5. and towards background stars
Our map covers only a small portion of the Northern Streamer observed by LLCB94. Many of the stars whose colors indicate high extinction lie outside the region mapped by us. We therefore decided to make pointed observations at the positions of 94 stars with , inferred from their H-K colors, larger than 10 magnitudes (Fig. 10).
Here we compare CO measurements at and angular resolution with extinction measurements which sample a pencil beam through the cloud at the center of each CO beam. The mean ratio of integrated (21) and (10) intensities towards the stellar positions is or very similar to that derived for the region we have mapped. The scatter is however larger by a factor of 2 and at some positions the ratio is significantly higher than 1. We conclude that we can use the same techniques applied in Sect. 3.6 to derive the column densities towards each of our target stars (i.e. we assume (10) to be thin, K, and compute ). The low optical depth assumption in this case was tested by observing at the positions of 7 background stars showing more than 20 mag of extinction . The (10) optical depth was found to be less than 0.5 at all positions (Table 1) - as in the case of the mapped region. The average ratio is , again indicating that is in fact optically thin at all positions.
In Fig. 12, we first show a comparison of (10) integrated intensity with visual extinction . The relationship (Eq. 6) derived by Alves et al. (1998) from the LLCB94 data set is shown for comparison. Surprisingly, measured integrated intensities are in general, especially at mag, lower than this relationship - in contrast also to the results we obtained for the mapped region (Fig. 11). Optical extinctions sampled with a pencil beam therefore appear to be an overestimate of the average optical extinctions within the IRAM 30m telescope beam, indicating structure at scales below (0.05 pc).
The basic conclusion one draws from Fig. 12a is that with this data set, there is little or no correlation between integrated intensities and . The mean uncertainty of integrated (10) intensities is 0.07 Kkms-1 () while the (10) dispersion of intensities within bins of 5 mag width is 0.46, nearly a factor of 7 larger. Thus the pointed observations show a large intrinsic dispersion in the apparent CO-to-dust ratio which cannot be attributed to the observational uncertainty. This is similar to the situation in the mapped region (Fig. 11a) where we found a factor of 4, and the LLCB94 study.
The ratio of LVG column densities to for each pointed observation is shown in Fig. 12b,c. A linear least squares fit to the abundance ratios results in
which indicates a slight drop of the abundance ratio from cm-2mag-1 at 10 mag to at extinctions of more than 25 mag. This relation is more shallow than the equivalent relation derived for the mapped core region of IC 5146 (Eq. 7). However, the largest abundance ratio found, 1.65, is consistent with the canonical ratio while the smallest ratio found, 0.2, indicates depletion by up to a factor of (Fig. 12b, Eq. 8).
© European Southern Observatory (ESO) 1999
Online publication: December 22, 1998