Abstract:
A hypothesis is proposed according to which single runaway stars are released from close binaries in which the companion star explodes as a carbon detonation supernova and is totally disrupted. From the stars which can explode as supernova, it is only the lower mass stars that burn carbon explosively. If the proposed hypothesis is correct, lower mass runaways should be predominantly single, while the higher mass runaways should be predominantly binaries containing inactive compact stars. In an analysis of data provided by Bekenstein and Bowers (1974) it is found that of 55 runaways, 29 have masses less than 15 solar masses, out of which only 7 (24%) are deduced to be binaries on the basis of radial velocity variations. On the other hand, out of 26 runaways with masses equal to at least 15 solar masses, 14 (54%) are deduced to be binaries.
