Seyfert galaxies were originally noted for the strength and broadening of their emission lines, and as a class were later characterized by the high ionization states of many of the atomic and ionized species producing these lines. This composite spectrum of the archetypal Seyfert NGC 4151 shows the wide variety of emission lines present, from the Lyman limit at 912 A to the mid-infrared at about 9 microns. It uses spectra taken with apertures several arcseconds in size, so as to reproduce the usual spectrum mixing broad and narrow-line components. From 912-1800 A, the data come from the Shuttle-borne Hopkins Ultraviolet Telescope; from 1800-3200 A, from the mean of three measurements by the International Ultraviolet Explorer (IUE) taken at similar brightness levels; from 3200-4000A, from an observation at Kitt Peak National Observatory, with the continuum rescaled to match the adjacent spectra; from 4000-8000 A, a CCD observation obtained at the Lick Observatory 3-m Shane telescope by Alexei Filippenko; from 8000 A to 1 microns, an observation using the same telescope by Donald Osterbrock and collaborators, carefully corrected for atmospheric absorption; from 0.9-2.4 microns, measurements by Rodger Thompson at Steward Observatory's 2,3-m Bok telescope, and on into the infrared, from the Infrared Space Observatory provided by Eckhard Sturm. Because NGC 4151 is irregularly variable, some of the spectral components have been scaled to make the various pieces match for this presentation (so the relative strengths of lines in very different spectral regions may not be accurate). Even so, I may not have gotten the IUE data spliced in quite right between the HUT and ground-based sections.
Some of the most prominent emission lines are marked for reference. The permitted lines - those that can be produced at high densities by astronomical standards - show both brad and narrow components. The strongest of these are the hydrogen recombination lines, such as Lyman alpha at 1216 A, H-beta at 4861, and H-alpha at 6563, plus the strong ultraviolet lines of C IV at 1549 and Mg II at 2800. Other features produced only by very rarefied gas at densities of 1000 atoms per cubic centimeter or so - the forbidden lines, denoted by brackets - arise in regions with less velocity structure and are narrower. Some strong examples are [O III] at 4959 and 5007 A, [O II] at 3727, [Ne V] at 3426, and [S III] at 9060 and 9532.
NGC 4151 is a bit unusual in showing strong absorption in several lines, especially Lyman alpha and C IV. The absorption is blueshifted with respect to the line centers, so that it arises in some kind of wind or other gaseous outflow.
The spectra of active galactic nuclei are noteworthy in showing species with a large range in ionization at once, from neutral ions such as [O I] and [N I] to highly ionized cases such as [Ne V] and [O VI]. Even hot stars such as light up gaseous nebulae in our galaxy cannot ionize gas as highly as these ions require, so that both a strong source of hard radiation and a wide range in gas density must be present to see such spectra.
The deep-UV HUT data were provided by Gerry Kriss, as described by Kriss et al. 1995, ApJL 454, L7. Between 1800 and 2500 A, the mean of three IUE spectra at similar brightness levels was used, while an HST FOS spectrum is shown between 2500 and 3300 A. From there to 4250 A, I used a spectrum I took at Kitt Peak National Observatory adjusted to the mean brightness of neighboring pieces, and from 4250-8000 A the plot shows an observation by Alex Filippenko taken at the 3-m Shane telescope of Lick Observatory. From 8000-9900 A, data from the same telescope by Osterbrock, Shaw, and Veilleux (1990 ApJ 352, 561) are shown, adjusted to match the variability stage of the neighboring optical data; their reduction included careful accounting for the complex absorption that our own atmosphere produces in this range. So did the infrared echelle spectrum by Thompson (1995 ApJ 445, 700) using the 2.3-m Bok telescope at Steward Observatory, shown from 1.0-2.4 microns, which I have done a grave injustice by averaging and degrading in resolution to make this plot more easily legible. Finally, the remaining infrared data are from an ISO observation provided by Eckhard Sturm. ISO scanned the wavelengths of strong emission lines in great detail, but skipped most of the intervening wavelengths, accounting for the blocky appearance of the deep infrared spectrum.