3 Colour-magnitude diagram analysis

Since the selected objects were first identified in old catalogues, we decided to use our CCD finding charts (Fig. 1) to examine their star distributions. The angular sizes for Melotte 105, Hogg 15, Pismis 21 and Ruprecht 140 given by Alter et al. (1970) are 4 $.\mkern-4mu^\prime$0, 2 $.\mkern-4mu^\prime$0, 2 $.\mkern-4mu^\prime$0 and 3 $.\mkern-4mu^\prime$5, respectively, so that our CCD frames should cover almost their entire fields. Indeed, Melotte 105 seems to extend over the whole CCD area, while Hogg 15 appears as a somewhat elongated group of bright and faint stars in the upper-left part of the chart. Pismis 21 seems to be a handful of concentrated stars dominated by some bright stars at the centre of the image. This would not appear to be the case for Ruprecht 140, for which we did not recognize any clear star concentration. According to its coordinates (Lyngå 1987), its centre is placed at ($X_{\rm c}$, $Y_{\rm c}$) $\approx$ (350, 100), where there are some faint stars with a halo of bright stars at $\sim$2 $\hbox{$^\prime$ }$ around them. We presume that these stars might have appeared like an open cluster in old field photographic plates. Although Ruprecht (1966) classified the star richness of all these objects as poor, Fig. 1 shows that the star density considerably varies from one field to another. In fact, in the fields of Melotte 105 and Hogg 15, 356 and 343 stars have been measured, respectively, while in the fields of Pismis 21 and Ruprecht 140 nearly 50% fewer stars have been measured. Thus, Melotte 105 is the richest cluster of the present sample.

Figure 4 shows the resulting CMDs for Melotte 105.

Figure 4: (V, B-V) and (V, V-I) colour magnitude diagrams for stars observed in the field of Melotte 105

The Main Sequence (MS) of a moderately young open cluster is clearly visible, which extends down to V $\sim$ 18.5 mag and covers a range of $\Delta V$ $\sim$ 7 mag. Our (V, B-V) CMD, even when obtained using a telescope with an aperture 2.5 times smaller than that used by Kjeldsen & Frandsen (1991), results in $\sim$1.5 mag deeper than their CMD. Notice also that the present photometry almost triples the number of stars measured by Kjeldsen & Frandsen, thus superseding any previous photometric study on this cluster. The cluster seems to be affected neither by a high reddening nor by field star contamination, as judged from the similar limiting magnitude reached in both CMDs and the scarce number of field stars lying outside the MS. The star with V = 11.291, B-V = 0.351 and V-I = 0.491 in the upper-left of the MS is suspected to be a blue straggler. Ahumada & Lapasset (1995) included this star in the third class of their catalogue of blue stragglers in open clusters, since it is located in the blue straggler area but has no membership indicator.

The CMDs for stars measured in the field of Hogg 15 are shown in Fig. 5. Like Melotte 105, the cluster MS extends through a large magnitude range. The upper MS looks tighter than its lower part, and it is mostly composed of the brightest stars in the field, which are clearly concentrated within a circle $\le$22 $.\!\!^{\prime\prime}$5 (50 pixels) around the adopted cluster centre ($X_{\rm c}$, $Y_{\rm c}$) = (380, 230). The lower MS becomes broader than the upper MS, particularly for V $\le$ 16 mag. Furthermore, the lower MS remains somewhat wider than the upper MS, even if we removed some probable field stars placed towards the reddest side of the cluster's lower MS. This broadness could be caused by differential reddening, since the cluster is near the centre of the Coalsack dark cloud. NGC 4609 is a well-known open cluster located $\sim$5 $\hbox{$^\prime$ }$ northeast following Hogg 15, with a spread of the stars around the MS larger than the typical dispersion in other open clusters, probably due to the fact that the distribution of the absorption material inside the Coalsack is far from being uniform (Feinstein & Marraco 1971; Straizys et al. 1994). According to Orsatti et al. (1998), the magnetic field in the direction to Hogg 15 follows the general trend of the polarization directions in the region.

Figure 5: (V, B-V) and (V, V-I) colour magnitude diagrams for stars observed in the field of Hogg 15

The field of Pismis 21 is rather different. Some different structures can be identified in Fig. 6, none of which are similar to an open cluster MS. Ng et al. (1996) studied the distribution of stars from the disc population in CMDs of star fields located at low Galactic latitudes towards the Galactic centre. According to their findings, the slightly tilted star sequence following the direction $V = 8\times(V-I) + 7.2$ in Fig. 6 is composed of MS disc stars. Note that each feature in the (V, V-I) CMD has its counterpart in the (V, B-V) CMD. This sequence becomes approximately one magnitude fainter than in the (V, B-V) CMD as a result of the contribution of redder stars. Moreover, the number of stars fainter than $V \sim 18.0$ mag in the (V, V-I) CMD is practically the same as that for the total magnitude range in the V vs. B-V CMD. On the other hand, the broad and somewhat dispersed group of stars centered at (V, V-I) $\sim$ (14.5, 2.0) would seem to correspond to Horizontal Branch (HB) young disc stars, whereas those stars distributed along the sequence placed below this group would be HB and Red Giant Branch (RGB) old disc stars. Thus the field of Pismis 21 does not correspond to an open cluster field, but to field stars belonging to different disc components.

Figure 6: (V, B-V) and (V, V-I) colour magnitude diagrams for stars observed in the field of Pismis 21

Carraro & Patat (1995) concluded that Ruprecht 46, also catalogued as an open cluster, does not form a physical system. Their CMDs reflect the characteristics of the Galactic field across the line of sight. The star sequence in the CMDs of Ruprecht 46 nearly follows the direction of the tilted sequence in the Pismis 21 CMDs. Notice that the position and dispersion of the field star sequences also depend on interstellar absorption through the various Galactic directions. However, not every loose open cluster candidate always results in a random fluctuation of the star density in a region of the sky. This is the case, for example, for NGC 7762: a scarcely populated open cluster whose appearance in the V image is not substantially different from that of Pismis 21 (Patat & Carraro 1995). This result demonstrates that the apparent visual identification of cluster candidates is not at all a sufficient criterion to assess their physical reality. Based on the assumption that Pismis 21 is a real cluster, Moffat & FitzGerald (1977) estimated a distance from the Sun of 1.7 kpc and Ahumada et al. (2000) derived a reddening value, age and metallicity for the stars located at the centre of Fig. 1 from integrated spectroscopy. However, we want to stress that this technique is very effective when it is used as a complementary tool, preferably of confirmed compact clusters.

Ruprecht 140, from the inspection of its CMDs in Fig. 7, is not an open cluster and, perhaps, some haziness produced by faint stars in low spatial resolution photographic plates with some bright stars around it could have caused its identification as an open cluster candidate.

Figure 7: (V, B-V) and (V, V-I) colour magnitude diagrams for stars observed in the field of Ruprecht 140