[O-III]-LAMBDA-5007 WING VARIATIONS AND CHANGES OF FE-II OPTICAL MULTIPLETS IN NGC-5548

The variations of the [O III] lambda 5007 line profile in NGC 5548 are examined from CCD spectra obtained at the Ohio State University and C rimean Astrophysical Observatory in December 1988 - October 1994. We h ave found the total Bur of [O III] lambda 5007 line to be variable, so the spectral calibration based on the assumption of non-variability o f [O III] lambda 5007 line is not absolutely correct. The variations a re much more pronounced in the far wings of the line profile (similar to 1100 km s(-1)). Assuming the central part of the line profile is no n-variable, we analyzed the variations of the red and blue wings relat ive to the profile center. We have found that these variations can be explained in terms of a reaction to the ionizing radiation changes wit h a transfer function spreading up to several hundred days. Perhaps, t he ''real'' constant flux is localized in the narrow [O III] line whic h belongs to the NLR and has a transfer function of about several hund red years, while the broad wings are emitted from the outer BLR. These wings may belong to Fe II lambda 5018 and/or [O III] lambda 5007. The transfer function of the blue wing is narrower, with a peak shifted t o the zero lag by similar to 200 days relative to the transfer functio n of the red wing. Both wings, and especially the blue wing, show weak reaction near the zero lag. These facts must exclude both pure radial and round motion, but may suggest a lack of emitting gas near the lin e of sight and the presence of an outflow component. The Fe II lambda 5169 line shows similar behavior to the [O III] lambda 5007 wings. It confirms Fe Ir lambda 5018 is the main contributor to the [O III] wing s or that the broad [O III] is emitted at the same distance from the c entral engine as the Pelt. The assumption about the presence of ''shor t'' and ''long'' lag spectral components, one of which varies similarl y to the H beta line and the other similarly to the [O III] lambda 500 7 wings, permits us to separate these components. The decomposition of the spectra shows that all observed Fe II multiplets have a high lag value of the order of several hundred days, pointing to the typical si ze of the region which emits these lines. The decomposition also shows that there is a short lag component in the red wing of [O III] lambda 5007 line. According to the centroid of this line (approximate to 503 2 Angstrom) it may be identified with the Si it lambda 5041, 5056 or H e I lambda 5016, 5048.