Astron. Astrophys. 340, 402-414 (1998)

9. Roslund 5

Roslund 5 was among the seven possible galactic clusters discovered by Roslund (1960) on objective-prism plates. It is located in Cygnus at (). It was studied photographically by Nelson (1969), who found a relatively well-populated main sequence for this cluster. Lee & Perry (1971) obtained UBV magnitudes for 46 stars in the cluster area. They found a considerable scatter in the colour magnitude diagram of these stars. They therefore concluded that Roslund 5 is not an open cluster, but only a slightly obscured area with a background of relatively early-type stars.

Looking at Fig. 2 of Lee & Perry, it seems possible that at least some of their stars are located at a common distance. A cluster may therefore still be present in this area. We have selected all stars in a by field centered on the cluster position from the Hipparcos Catalogue. Fig. 15 shows their proper motions. There is a conspicuous group of stars with proper motions close to = 3 mas/yr, = -1 mas/yr. This may be a hint for a star cluster.

Fig. 15. Proper motions of Hipparcos stars in the field of Roslund 5.

Fig. 16 shows proper motions, positions, parallaxes, and colour-magnitudes of the stars which may form a cluster. They show a concentration around . In addition, our hypothetical members may be at a common distance (Fig. 16c) and their magnitudes are sufficiently close to each other to be consistent with a common distance (Fig. 16d). Their colours show some scatter, but this may be due to the fact that the extinction is varying over the region studied (Lee & Perry 1971). We obtain a mean parallax of mas, corresponding to a distance of about 500 pc. If we assume a mass of for Roslund 5, we obtain a tidal radius of 5 pc, which corresponds to an angular diameter of roughly one degree. At least three, possibly even four of the Hipparcos stars may therefore be physically connected to each other.

Fig. 16a-d. Proper motions (a ), positions (b ), parallaxes (c ) and colour-magnitudes (d ) of the possible members of Roslund 5.

We next searched the ACT Catalogue for possible members. If Roslund 5 is a real cluster, one would expect to see a density enhancement of ACT stars around the cluster position. However, we do not see such an enhancement if we consider all ACT stars in the field (see the left panel of Fig. 17). But Roslund 5 may be sparsely populated, so this does not necessarily exclude a cluster. We therefore apply some criteria to the ACT stars, to pre-select members from non-members. Within the magnitude limits of the ACT, a main-sequence star surely has if it is a cluster member. In addition, we require that the proper motion of a star must be sufficiently close to mas/yr, mas/yr, so that the -value of its proper motion deviation

is lower than a certain threshold . We found that is a good compromise between bad statistics due to a low number of stars and no signal due to too many stars considered. The distribution of the remaining stars is shown in the right picture of Fig. 17. Again, there appears to be a slight clustering near .

Fig. 17. Distribution of all ACT stars (left) and those stars fulfilling the membership criteria (right) in the field of Roslund 5.

We consider all stars that fulfill the membership criteria and lie within a circle of 0.5 degree radius around as possible cluster members. They can be found in Table 9 together with the possible members from Hipparcos. Table 9 surely does not include all members of Roslund 5 and it may contain several non-members also.

Table 9. Possible members of Roslund 5
Notes:
Column 1: Running, 2: HIP, 3: HD/BD number, 4-6, 7-9: Right ascension and Declination (J2000), 10+11: V and B-V, 12-13: Parallax, 14-17: Proper motion from Hipparcos or ACT, 18+19: Radial velocity and error from the WEB Catalogue

The last two columns give the radial velocity and a quality flag for the stars which could be found in the WEB Catalogue (Duflot et al. 1995). We note that the quality flag of the WEB Catalogue is on the system of the GCRV (Wilson 1953), which means that the error of the radial velocity rises from A to E. For example a C corresponds to a mean error of less than 2.5 km/sec, while a D corresponds to a mean error of less than 5 km/sec. Keeping this in mind, it seems possible that most stars of Table 9 have the same radial velocity. The exception may be HIP 99299 with a velocity of -37 km/sec, which seems to be too different from the mean cluster velocity of about -18 km/sec.

The data presented so far is compatible with the assumption that Roslund 5 is a cluster, although we cannot rule out the possibility that our stars are part of an unbound association. Precise radial velocities for the stars of Table 9 would help to decide whether they form a bound system or not.

© European Southern Observatory (ESO) 1998

Online publication: November 9, 1998
helpdesk.link@springer.de