Hubble Space Telescope ultraviolet spectrum of ARP 102B, the prototypical double-peaked emission-line AGN

UV spectra of the nucleus of the elliptical galaxy Arp 102B were obtained with the HST's Faint Object Spectrograph in order to investigate the UV emission-line counterparts of its unusual double-peaked Balmer lines. Broad Mg II λ2798 is present with nearly the same profile as the Balmer lines (peaks separated by ≈ 12,000 km s^-1), and a typical Mg II/Hβ ratio of 1. But there is little, if any C III] λ1909 or C IV λ1550 emission corresponding to the displaced Balmer-line peaks. Most important, there is no double-peaked component detected in Lyα; the Lyα/Hβ ratio is less than 0.12 in the displaced peaks. However, there is an "ordinary," nondisplaced broad-line component with FWHM ≈ 3500 km s^-1 in all of the permitted lines, demonstrating the need to invoke different locations and different physical conditions for double-peaked and single-peaked line components in the same object. The striking absence of displaced peaks in Lyα probably cannot be explained solely by reddening. Rather, it indicates that high density and large optical depth in Lyα are required to destroy the line photons by collisional de-excitation and possibly by bound-free absorption out of the n = 2 level of hydrogen. These results strongly support the application, at least to Arp 102B, of the accretion-disk model of Dumont & Collin-Souffrin, in which the disk produces only low-ionization lines and a Lyα/Hβ ratio that agrees with our observed upper limit. Also present is an extraordinary system of absorption lines at the systemic redshift of Arp 102B, in which metastable levels of Fe n up to 1.1 eV above the ground state participate in addition to the more common resonance transitions. Absorption from metastable levels of Fe II have been seen previously only in two unusual, low-ionization broad absorption-line QSOs, Q0059 - 2735 and Mrk 231. Temperatures and densities needed to excite these levels are similar to narrow-line region conditions. Why they are seen in absorption in Arp 102B, a relatively unobscured active galactic nucleus, but in no other Seyfert or radio galaxy, is a mystery.