5 Conclusions

Based on spectrophotometric ISO imaging with the LWS and the CAM-CVF of significant parts of the active star forming ${\rm Serpens ~ cloud ~ core}$ our main conclusions can be summarised as follows:

$\bullet$

We find the emission in the [O  I] 63 $\mu$m and [C  II] 157 $\mu$m fine structure lines to be extended in our $8^{\prime}\times 8^{\prime}$ map. The absolute intensities and their ratios can be explained in terms of PDR models, where a UV field of $G_0 = 15 \pm 10$ is falling onto the outer layers of the dark cloud, where densities are of the order of (10⁴-10⁵) cm^-3.

$\bullet$

Also the emission in rotational lines of H₂O and high-J CO appears (slightly) extended, but we cannot exclude the possibility that it arises from point sources in the field, viz. the Class 0 objects SMM 9/S 68 and SMM 4. The maximum of the emission is observed toward SMM 1, the dominating far infrared and submm source in the ${\rm Serpens ~ cloud ~ core}$.

$\bullet$

The spectrum of SMM 1 contains numerous lines of CO, H₂O and OH. These lines are generally subthermally excited and optically thick and trace regions of dimensions $\sim$10³ AU ( $\sqrt{1500 \times 600}$ AU), where temperatures are above 300 K and densities above 10⁶ cm^-3. The derived abundances, relative to H₂, are for CO, H₂O, OH and ¹³CO, respectively, $X_{\rm mol}=(1,~0.1,~0.02,~\ge\!\!0.025) \times 10^{-4}$. The ortho-to-para ratio for H₂O is consistent with the high temperature equilibrium value (ratio of the statistical weights of the nuclear spins), i.e. H₂O-o/p =3.

$\bullet$

The relatively high OH abundance is indicative of an elevated level of ionising flux in the ${\rm Serpens ~ cloud ~ core}$, causing the ionisation rate to be $\zeta \gg 10^{-18}~{\rm s}^{-1}$, i.e. significantly higher than the average rate prevailing in dark clouds. Strong and active X-ray sources, known to exist in the cloud, could be responsible for this radiation.

$\bullet$

The observed SED of SMM 1 is consistent with a model of a dusty torus with an outer radius of 14 000 AU (45 $^{\prime \prime }$). The torus is heated by a central stellar source, with a (somewhat arbitrarily) adopted effective temperature of 5000 K. A luminosity of 140 $L_{\odot}$ is required to explain the observations. The total mass of the toroidal core of SMM 1 is 33 $M_{\odot}$. The derived visual extinction through the torus exceeds 2000 mag and the torus is optically thick up to mm wavelengths.

$\bullet$

2D modelling of the radiative transfer through the circumstellar torus of SMM 1 reveals that it is highly unlikely that the observed molecular emission arises in the torus. The same conclusion can be drawn for models of dynamical collapse ("inside-out'' infall).

$\bullet$

The H₂ data have been obtained at significantly higher spatial resolution than that offered by the LWS. These CAM-CVF observations trace the regions of pure rotational H₂ line emission. The H₂-maxima are observed to be displaced from SMM 1 and are situated toward the northwest, along the jet of outflowing material from this Class 0 source. The temperature of $\sim$10³ K of this H₂ gas is indicative of the heating by relatively slow shock waves.

$\bullet$

The comparison of our molecular line data with shock models suggests that shock heating with $v_{\rm s}$ $\sim$ (15-20) km s^-1 along the outflow of SMM 1 is the most likely mechanism of molecular excitation, although the details of these shocks are less clear.

Acknowledgements

We are grateful for the help with the data reductions of the CVF observations by Stephan Ott. We also thank Ewine van Dishoeck for making avalailable to us the collision rate coefficients for OH in electronic form. The support of this work by Rymdstyrelsen (Swedish National Space Board) is acknowledged.