Normal and Abnormal Binary Frequencies

It is becoming apparent that most normal main-sequence stars are members of both short-period and long-period systems. Let us cite some evidence. Petrie ( 1960) showed from a sample of 1752 stars that 52%, independent of type from 09 to M, have variable radial velocities, mostly due to spectroscopic duplicity. But not all short-period binaries can be detected spectroscopically because of rotationally broadened lines and finite measuring accuracies. Therefore it seems likely that somewhere between 50 and 100% of the normal stars are in short-period binaries. Visually, Heintz (1969) showed that 42% of the brighter stars have one or more visual companions, averaging 0.77 companions per star, and Poveda et al. (1982) showed that after allowance for unseen companions (too close to resolve, too faint to have been measured), nearly 100% of the stars are in long-period binaries. The two techniques are nearly independent, over­ lapping only for some of the stars within 25-50 pc of the Sun. There­ fore we conclude that most stars are in both short-period and long-period binaries. On the other hand, we note differences from one group of stars to another: the novae, metallic-line, and Ba II stars seem always to be in spectroscopic binaries, whereas the Be and high-velocity stars seldom are spectroscopically double. How can one group of stars show nearly double the normal spectroscopic binary frequency when the normal frequency is already nearly 100%? This article is a discussion ofthe recent data · on duplicity and multiplicity for (a) the normal main-sequence stars as a basis for comparison and (b) a variety of classes of stars that have abnormal frequencies. The goals are to assess the occurrence of multiplicity among normal stars and to identify