2 Properties of the input sample

  We shall briefly summarise in this section the main properties which characterise the sources in our sample.

First, a word on the nomenclature (see e.g. Table 1) is useful. We have adopted that used by TFTH, but this should be treated only as indicative because several other names are often in use and also because these names may refer to objects much more extended than those studied by us. To avoid confusion, it is best to consult the coordinates given in Table 1. These refer to the strongest H₂O maser component found in the VLA observations of TFTH. NGC 2264 has been renamed NGC 2264-C to avoid confusion with other molecular outflows present in the same molecular cloud which bear similar names (it is component C in the original outflow survey of the Monoceros OB1 molecular cloud made by Margulis et al. 1988). Schreyer et al. (1997) call this source NGC 2264 IRS1.

The basic common property of the sources in our sample is that in all the fields, there is an H₂O maser (Comoretto et al. 1990; Brand et al. 1993) of star forming type (Palagi et al. 1993). All the H₂O masers were observed with the VLA and have positions with 0.1$^{\prime\prime}$ accuracy (TFTH). They were also observed with the VLA in the continuum at 8.4 GHz and no small diameter source (size $\leq$ 1$^{\prime\prime}$) with flux density greater than 0.3 mJy (3$\sigma$) was found within a radius of $\sim$10$^{\prime\prime}$ of the maser position (TFTH). The only exception is AFGL5142, where a 0.85 mJy (at 8.4 GHz) source is present at the position of the two brightest maser components. It was included because multiwavelength observations had already been obtained, showing a complex outflow morphology (Hunter et al. 1995).

In all fields, there is an IRAS source close to the H₂O maser (within $\sim$30$^{\prime\prime}$). The IRAS name, the four IRAS fluxes and the 25/12 and 60/12 colours are given in Table 1. All but two (GGD4 and NGC 2264-C) fall in the box of the 25/12-60/12 colour-colour plot that Wood & Churchwell (1989b) define as that occupied by UC HII regions ($25/12\geq0.57$, $60/12\geq1.3$). However, they are clearly not UC HII regions as normally defined in that their radio continuum luminosity is extremely small. One should also note that a direct physical relationship between the IRAS source and the H₂O maser (more specifically, with the young star that powers the maser) is far from being firmly established. This is partly because the accuracy of the IRAS position is at least two orders of magnitude worse than that of the H₂O masers. Secondly, the IRAS source is in most cases associated with a cluster of newly formed stars. The large IRAS beam collects the emission from the entire star forming complex (Hunter et al. 1994; Testi et al. 1994; Hunter et al. 1995; Felli et al. 1997) and the separation of the contribution of different stars is impossible. In other words, the presence of an IRAS source confirms that we are indeed looking at star forming regions, but the IRAS fluxes cannot be unambiguously attributed to the young stellar object (YSO) associated with the H₂O masers.

The sources of our sample are located in known molecular clouds and are associated with large scale CO molecular outflows (see Fukui et al. 1993 for the references). However, we do not aim in this work to better define the large scale outflows (which are not necessarily related to the YSO associated with the water masers) because this would require larger scale maps than we were able to carry out. Similarly, we have not attempted to provide large scale maps of the "ambient'' molecular cloud. Our aim has rather been to search for the presence of high density clumps in a small region around the maser positions using high density tracers. In Table 2, we give for further reference the peak velocity of the CO cloud ($V_{\rm p}$)and the two velocities that define the inner (min) and outer (max) ranges of the blue and red large scale outflows.

  

Table 2: Velocities of the large scale CO molecular outflows (see TFTH and references therein)

In Table 3 we give a list of parameters derived from the literature using the distance (d) adopted by TFTH: the IRAS (bolometric) luminosity ($L_{\rm FIR}$), the H₂O maser luminosity ($L_{\rm H_2O}$), the mechanical luminosity of the large scale CO outflow ($L_{\rm CO}$), the expected flux of Lyman continuum photons ($N_{\rm L}^{\rm exp}$) assuming a single ionising star with luminosity equal to the IRAS luminosity, and the expected radio flux density at 8.4 GHz ($S_{\rm exp}$) from an optically thin HII region. One sees from the range of $L_{\rm FIR}$($200-3~10^4~L_\odot$) that our sample is not restricted to proto O-B stars but also contains embedded stars of later spectral type. Indeed in these latter cases (GGD 4, IC 1396-N, L1204-G), it is not surprising that no radio continuum emission was observed by TFTH because $S_{\rm exp}$ was below the detection limits (see Table 3). Nevertheless, we thought it interesting to include them in this survey in order to see if they share the molecular morphology of the high luminosity objects.

Finally, it is important to be aware that all of these star forming regions have been imaged in the near IR (Hodapp 1994; Testi et al. 1999). In almost all cases, a cluster of $K\hbox{$^\prime$}$-band sources was found (Hodapp 1994), clearly revealing the complex nature of the star forming region and showing that high resolution observations are needed to distinguish different stars in the same cluster.

  

Table 3: Derived source parameters