7 Runaways and pulsars without parents?

The sample of runaways and pulsars we have analysed here is severely incomplete (see Sect. 2.1). Furthermore, we have only a limited knowledge of where massive stars form in the Solar neighbourhood. The case of $\zeta$ Pup described in Sect. 5 shows that some local parent groups may not yet have been found. Beyond $\sim$500 pc, only the large complexes such as Cygnus OB1/OB2 or Sco OB1 are well documented. The accuracy of the three-dimensional positions and velocities of these distant star-forming regions is poor, making determination of reliable orbits very difficult. Runaway stars of spectral types B0 and later are especially difficult to link to a parent group. These stars have main-sequence lifetimes of up to several tens of Myr and can therefore travel far (several kpc) from their places of origin. A similar argument holds for the pulsars. We suspect that most of the runaway stars and pulsars in our sample for which we were not able to identify a parent group originated outside the Solar neigbourhood.

An example is the B2.5V star 72 Col, HIP 28756 (van Albada 1961; the asterisk at $[\ell,b] \sim [238^\circ,-23^\circ]$in Fig. 2). It has a peculiar velocity of $\sim$200 km s^-1 and its parent association is Sco OB1, at a distance of $\sim$2 kpc (Humphreys 1978). Van Albada derived a kinematic age of $\sim$14 Myr for 72 Col, based on a simple model of Galactic rotation (Kwee et al. 1954). This star does not appear in Sects. 5 and 6 because its path did not carry it through one of the nearby associations. The Solar neighbourhood thus not only contains runaways for which the parent associations are also nearby, but it also contains runaways which originated far from the Sun.

Another star that immediately catches the eye in Fig. 2 is HIP 94899 (the asterisk at $[\ell,b] \sim [341^\circ,-26^\circ]$). This double star of spectral type B3Vn has a radial velocity of 151 km s^-1, and its path seems to cross the Per OB3 association. However, our simulations show that the runaway never comes within 40 pc of the association, implying that this system must have another, unknown, parent.

We have seen in Sect. 3 that PSR J1239+2453 most likely originates outside the Solar neighbourhood. We did not find a parent group for the remaining pulsars because (i) an unreasonably large radial velocity (>500 km s^-1) is necessary for the paths of the pulsar and parent group to intersect (PSR J1135+1551), just as we found for PSR J1239+2453, or (ii) the past orbit simply does not intersect any of the nearby young stellar groups (PSR J0953+0755 and PSR J1456-6845).