5. Proper motions
For the low Galactic latitude objects, the kinematic analysis is limited by the lack of proper motion data (eg. Rolleston et al. 1997; Heber et al. 1997). Recently, Thejll et al. (1997) presented proper motion measurements for a sample of 138 PG stars that were classified as being hot sub-dwarfs; these include four of our programme stars, viz. PG 0009+036, PG 2219+094, PG 2229+099 and PG 2345+241. The proper motion vectors (µ, ) in the equatorial coordinate system were converted to the Galactic coordinate system (, ) using the numerical routines within SLALIB (Wallace 1992). Velocity components perpendicular to, and parallel with the Galactic plane were calculated from the peculiar radial motion and proper motion. The `true' z-velocity was then used to compute time-of-flights; details of the revised kinematical analysis are given in Table 9.
Table 9. Stellar proper motions.
In particular, the proper motion of PG 2219+094 makes a significant contribution to the star's velocity perpendicular to the plane. Indeed, the revised velocity, km s-1, reveals the runaway nature of this star - the subsequent kinematical analysis confirming that PG 2219+094 could have been ejected from the star-formation regions of the Galactic disk. The inclusion of the proper motion information for the other three programme stars in the kinematical analyses, support the conclusions presented in Sect. 4.
Unfortunately, the errors of measurement yield a large spread in the possible ejection velocities (see Table 9), and hence it is difficult to make any conclusion about the ejection mechanism. Moreover, these proper motions have been derived using comparisons between absolute positions on AGK3/FK5 equinox systems, and may contain relatively large systematic errors for objects that are expected to have very small proper motions (eg. of order 10 mas yr-1). The problem of non-inertial pre-FK5 equinoxes is discussed in, for example, Fricke (1982) and references therein. Absolute proper motions measured with long time baseline plate pairs and with respect to galaxies may provide more accurate information (eg. the Lick Northern Proper Motion project, Klemola et al. 1987). However, the input catalogue for this project contains none of the stars discussed here, and the accuracy of the proper motions measured is mas yr-1 (Jones 1996). With the completion of the second epoch Palomar sky survey, a better approach may be to determine relative proper motions with subsequent corrections for absolute proper motions perpendicular to the Galactic disk - for the purposes of estimating velocities in the z-direction. Sky limited 48 inch Schmidt plates are not ideal for astrometry of objects brighter than m14 mag due to off-centred image halos as a function of plate position; this problem is compounded by the POSS-I (Minkowski & Abell 1963) and POSS-II (Reid et al. 1991) field centres being different. Despite this, a careful treatment of magnitude-dependent systematic effects coupled with a centroiding accuracy of 1 µm, a plate scale of 0.07 arcsec mm-1 and an epoch difference of yr may yield proper motions of accuracy mas yr-1 (see Hambly et al. 1998).
© European Southern Observatory (ESO) 1999
Online publication: June 18, 1999