The initial period function of late-type binary stars and its variation

The variation in the period distribution function of late-type binaries is studied. It is shown that the Taurus-Auriga pre-main-sequence population and the main-sequence G dwarf sample do not stem from the same parent period distribution with better than 95 per cent confidence probability. The Lupus, Upper Scorpius A, and Taurus-Auriga populations are shown to be compatible with being drawn from the same initial period function (IPF), which is inconsistent with the main-sequence data. Two possible IPF forms are used to find parent distributions to various permutations of the available data, which include Upper Scorpius B (UScB), Chameleon, and Orion Nebula Cluster pre-main-sequence samples. All the pre-main-sequence samples studied here are consistent with the hypothesis that there exists a universal IPF that is modified by binary-star disruption if it forms in an embedded star cluster leading to a general decline of the observed period function with increasing period. The pre-main-sequence data admit a log-normal IPF similar to that arrived at by Duquennoy & Mayor (1991, A&A, 248, 485) for main-sequence stars, provided the binary fraction among pre-main-sequence stars is significantly higher. However, for consistency with proto-stellar data, the possibly universal IPF ought to be flat along the log-P or log-semi-major axis and must be similar to the K1 IPF form derived by means of inverse dynamical population synthesis, which has been shown to lead to the main-sequence period function if most stars form in typical embedded clusters.