No CrossRef data available.
Published online by Cambridge University Press: 12 April 2016
I describe three eclipsing double-lined binaries whose fundamental data (spectra, masses etc.) are totally out of line with theoretical concepts. Accepting the data at face value, they seem to imply that (i) a massive star (about 10 – 30 M⊙) in a moderately wide binary (P ≈ 50 – 250 d) can eject almost its entire envelope to infinity, without shrinking its period substantially, (b) some formerly triple systems can become binaries through the merger of two of the three components, and (c) a red giant can be stimulated to lose half its mass in a stellar wind before reaching its Roche lobe.