NASA NTRS1990
The N-galaxy 3C 390.3 has been monitored spectroscopically since 1974 (Osterbrock, Koski and Phillips 1975; Oke 1988). From various archives and literature, it is found that the Balmer lines change their intensities and profiles in a dramatic manner. The H alpha profile is very broad and peculiar, and the relative intensities of its two humps changes consistently with time, possibly periodically. Before 1980, the blue hump was significantly stronger than the one in the red. From 1980 to 1983 the blue hump became stronger (see Oke 1988). After 1983 the H alpha profile has returned to its early shape and seems to have completed a full circle. Unlike the rapid (on the order of a month or even less) and aperiodic variation in the continuum and integrated line intensities, the change in broad profile seems slow and consistent. Taking the analogy of cataclysmic variables, the double-horn profiles have been observed in cases of interacting stars. For example, the emission lines, both in He II and hydrogen Balmer lines in GD 552 (Stover 1985) show double-horn profiles and periodical changes in their line profiles, including the change in ratios of two humps. It is understood that the D-wave components (Smak 1976) are the signature of an emitting disk and the S-wave component is from the emission at a hot spot which rotates and results in a moving component in the velocity space. The mass flow from the nearby interacting star provides the stream toward the core of a neutron star or white dwarf. Therefore, it is proposed that the variation of broad line profiles observed in 3C 390.3 may be the result of a pair of interacting massive cores. The rotational velocity dominates and produces a variable double-horn profile. However, the line widths observed in broad line radio galaxies are one order larger than that in interacting stars. The Balmer decrements imply a much smaller density (10(exp 10-12) cm(exp-3)) than that in the cataclysmic variables. The much larger velocity and much thinner density make it unlikely that the broad line emission is simply formed in an accretion disk. The authors postulate that a significant rotational motion is involved. If the observed squared profiles are indeed due to the rotational velocity field, one can naturally explain their rare occurrence.