2023A&A...678A..47B

The observations of all known major activity phases of the disks around the classical Be stars γ Cas and 59 Cyg with low-mass companions are comprehensively reviewed and purely qualitatively evaluated again, though taking advantage of new insights gained over the past 25 yr into the physics of Be disks. Both stars have exhibited activity cycles in the violet-to-red (V/R) flux ratio of emission lines with two peaks. This activity is indistinguishable from those of the vast majority of Be stars and so probably were caused by one-armed (m = 1) disk oscillations. The anomalous high-activity phases from 1932 to 1942 in γ Cas and between 1972 and 1976 in 59 Cyg were distinguished from m = 1 density waves by large variations in the separations of pairs of emission peaks. In two consecutive cycles, shell phases during which the emission peaks were maximally separated alternated with single (blended) emission peaks. The amplitude in peak separation of more than a factor of two implies a high-amplitude variation in the disk aspect angle. When the peaks were blended and the disk was viewed closest to face-on, local maxima in visual brightness probably occurred in γ Cas, and the visibility of the stellar absorption lines was reduced, as is expected from increased free-bound emission into the line of sight (there is no time-resolved photometry for 59 Cyg from the event in the 1970s). In y Cas, the pre-event V/R variability (pre-event observations of 59 Cyg do not exist) was practically identical to m = 1 variability. In spite of the subsequent rapid rise in amplitude (up to ∼4), the V/R variations connected smoothly in phase but may require an explanation involving the 3D structure of the disk. The phasing of single-peak and shell stages relative to the V/R activity was the same in both cycles of γ Cas, whereas this is not clear for 59 Cyg. During both high-activity cycles of γ Cas, but at different phases, transient additional pairs of emission lines appeared in γ Cas that were much sharper than the main ones and they also had different peak separations and V/R ratios. In the second instance, their velocities were up to ~+500 km s^–1. The extremely rapid excitation of the activity phases and their short duration of only two cycles in both stars may indicate a resonant behavior of an unidentified nature. In both stars, the line emission was strongly developed at the onset of the high-activity phases but it basically disappeared at the end of them, and the disks may have been dynamically destroyed. The atypical disk variations were presumably triggered by enhanced interactions between a disk and companion star. In both systems, there seems to be less evidence for a mass-loss outburst than for a reduced mass-injection rate into the disk. The resulting lower viscous coupling between a disk and star would have facilitated the tilting of the disk.