2. The source list
2.1. Target selection
Here, association means that the IRAS position is located within the optically visible (opaque) boundaries of the globule. Although it was not a strong selection criterion, most of the selected globules have high extinction values (optically thick in the centre; opacity class A in the nomenclature of Hartley et al. or class 5 or 6GAMMA in the nomenclature of Feitzinger & Stüwe).
In total, we selected 35 globules for our mm continuum and line study. Of these, 15 globules are also contained in the infrared study of Persi et al. (1990) and 20 globules are contained in the list of BHR. Table 1 lists the selected globules together with the names and coordinates of the associated IRAS point sources. For the dark cloud (DC) names, we use the nomenclature of Hartley et al. (1986).
Table 1. Coordinates and distances of the selected globules
In order to simplify the understanding of the target selection and the interpretation of the data, we subdivided the globules into sub-groups according to the IR broad-band SEDs of the associated IRAS point sources, following the classification scheme of Paper I. Here, we use the IR spectral index . According to the definition of , the colour temperature increases with increasing spectral indices. One has to keep in mind, however, that the FIR colour temperatures may systematically overestimate the true effective dust temperature of the dense cores if very small, transiently heated grains contribute to the 60 µm emission. Fig. 1 shows the selected sources in the IRAS colour-colour diagram. The sample breaks up into the following two groups:
The group membership and the IRAS point source flux densities are listed in Table 3. Three globules could not be classified within this scheme. They are marked by "?" in Table 3. Fig. 2 shows the averaged broad-band SEDs of the two groups compiled from the IRAS point source fluxes. Here, we excluded the far Carina sources ( kpc, see Sect. 2.2) from the averaging procedure.
In contrast to our list of northern globules (Paper I), this sample does not contain any group 3 sources (SEDs which are falling from 12 to 25 µm, candidates for T Tauri stars) nor "star-less" globules (group 4). The selected sample is obviously biased towards globules with deeply embedded YSOs, and therefore is not representative of Bok globules in general nor of the entire star-forming phase.
There are four objects in our sample which were already studied in more detail by other authors. The IRAS source 08242-5050 in the globule DC 267.4-7.5 is associated with the spectacular HH jet HH 46-47 and a bipolar molecular outflow (Schwartz 1977; Dopita et al. 1982; Chernin & Masson 1991; Olberg et al. 1992). The cometary globule DC 253.3-1.6 contains the infrared object IRAS 08076-3556 which has a steep SED and is located very close to the Herbig-Haro object HH 120 (Persi et al. 1994). NIR images (H2) show evidence for two outflows in this globule oriented almost perpendicular to each other (Hodapp & Ladd 1995). The sources in DC 253.3-1.6 and DC 267.4-7.5 were already measured at 1.3 mm continuum by Reipurth et al. (1993). The globule DC 297.7-2.8 (BHR 71) which has a deeply embedded YSO driving a collimated bipolar outflow was recently studied in detail by Bourke et al. (1997). The IRAS source embedded in the globule DC 303.8-14.2 was found to drive a bipolar molecular outflow and to show spectroscopic evidence for gravitational collapse (Lehtinen 1997).
For most of the globules of our list the distances were not known. Therefore, we checked the association of the globules with molecular cloud complexes with well-known distances using the method described in Paper I. Distinct groups of globules are associated, e.g., with the Vela-Gum complex (Zealey et al. 1983) or with molecular clouds in the Carina arm of the Galaxy (Grabelsky et al. 1987). The distances derived in this way are listed in Table 1 together with the associated molecular cloud structures. For some of the selected globules, we adopted the distances derived by BHR from stellar reddening which are consistent with the distances derived by our method.
Following Paper I, we assigned three reliability classes to the derived distances with A being the most reliable distances and C the most uncertain ones. Class B was assigned to most of the globules which are associated with molecular clouds in the Carina arm (photometric distances, Brand & Wouterloot 1988). Four of the class C globules which are located at galactic latitudes between and and which have many foreground stars were associated with the inner edge of the Carina arm. Two other C globules at galactic latitudes of and which have only very few foreground stars were assigned to the average distance of 200 pc of the local globules in this direction.
Figs. 3 and 4 show the galactic distribution of all globules of our sample together with the boundaries of selected molecular cloud complexes. The distance distribution (Fig. 5) has two dominant peaks at 200 and 400 pc which are related to Lindblad's expanding ring of early-type stars and dark clouds ("Gould's Belt"; e.g., Lindblad et al. 1973; Sandqvist & Lindroos 1976) and to the Vela-Gum complex. All globules of this sample which are further away than 500 pc are related to the Carina arm.
The objects in the Carina arm are more distant ( 0.7 kpc) in accordance with the fact that they have more foreground stars than the nearby globules in the local spiral arm. Since we are looking into the relatively empty inter-arm region towards the Carina arm at galactic longitudes between and (see Fig. 3), also relatively distant globules can be seen in this direction. While the most distant objects at longitudes between and (2.2 - 4.4 kpc) are difficult to identify as globules because of their diffuse appearance, the appearance of the more nearby Carina globules at longitudes between and (0.7 - 1.3 kpc) is not very different from that of the local globules. In other directions, the higher stellar density within the local spiral arm prevents the identification of such distant globules from optical surveys.
The average distance of the local globules (excluding all Carina sources) is 300 pc. This value is somewhat smaller than the average distance of 500 pc which we derived for the northern globule sample (Paper I). This discrepancy can be understood if one considers that the Sun is located at the inner ("southern") edge of the local spiral arm. Excluding only the "far" Carina sources ( kpc), the average distance of the selected globules ( kpc) amounts again to 500 pc.
© European Southern Observatory (ESO) 1998
Online publication: September 8, 1998