1999A&A...341..867H

The current situation in the research of rapid line-profile and light variations of early-type stars is critically reviewed. It is suggested that the ultimate understanding of the physical processes causing these variations can only come from an open-minded and complex approach to the problem and from systematic observational effort. It is argued that the results of the search for periodicities in the complicated variations of these objects depend critically on whether the method used is appropriate to the real physical situation. The danger of detection of a false multiperiodicity is pointed out for two particular situations: (i) a single-periodic signal which undergoes slow periodic change, e.g., due to the light time effect in a binary system, and (ii) a single-periodic signal with a complicated phase curve (a model of not exactly equidistant corotating spokes). It is argued that the observed rapid variations need not be due to classical non-radial pulsations but may arise from more complicated velocity fields in the stellar atmospheres and/or mantles. Two early-type stars, ε Per and ζ Oph, are discussed in detail. It is argued that both may be the cases where the variations are caused by corotating structures slightly above the stellar photosphere. For ε Per, the pattern of the variations can also be affected by the motion of the star in a binary orbit. For ζ Oph, a double-wave light curve with the corotation period of 0.64° (suggested by the author earlier for the line-profile variations) was found from Hipparcos V photometry and its presence can also be suspected in other existing photometric data and in the recurrence times of the narrow features seen in the UV spectra.