1997A&A...318..812D

Using a detailed model of massive close binary evolution and accounting properly for the effects of asymmetric supernova explosions (SN) where we use recent observations of pulsar runaway velocities, we determine the theoretically expected number of post-SN O-type stars with and without a compact companion (CC), the number of O-type runaways, the number of WR+CC systems and the formation rate of binary pulsars in our Galaxy. We conclude that o at least 50% of the O-type runaways are formed through the binary scenario, o less than 3% of all WR stars may hide a CC, o the formation rate of binary pulsars in our Galaxy ≃0.003-0.01 times the formation rate of massive stars; this corresponds roughly to a binary pulsar formation rate of the order of 10^–5/year in agreement with the observations. Our results reveal a significant fraction of single WR stars but with a binary history. We also predict the existence of 'weird' WR stars, i.e a WR star with a CC in its centre (descendants of Thorne-{dot}(Z)ytkow objects).