1999MNRAS.304..622S

Several OH/IR sources were reported by David et al. and Dickinson & Turner to show a slight overshoot of the 1667-MHz OH line relative to the presumably terminal expansion velocity indicated by the 1612-MHz line.

From the David et al. data, we selected a representative sample of 21 objects which we observed with the Nanay radio telescope with a good velocity resolution in both lines simultaneously. Our data clearly confirm the overshoot phenomenon. Shifts range from a few tenths to a few km.s^–1. Prominent overshoots, by at least 3 velocity channels, are found in 65 per cent of the measurements; only 17 per cent do not show a clear overshoot. Overshoots in the front and back parts of the envelope occur independently. No strong correlation was found with the front/back asymmetry, expansion velocity, IRAS colours, pump efficiency or line ratios.

It is argued that in normal envelopes, slight changes in the medium can broaden main lines, or can quench satellite line components, so that main lines appear to be the most extended. Overshoots can also be produced, however, by departure from the currently accepted circumstellar processes. Zeeman broadening, clumpiness, velocity fluctuations or axisymmetric winds appear as the most likely processes able to produce overshoots. Monitoring, mapping and observations of other maser species are required in order to identify firmly the nature and the importance of these overshoots.