Emission lines provide clues to the nature and properties of the objects which emit them. Much of what we know about H II regions and planetary nebulae depends, in one way or another, on such data. Concerning active galactic nuclei, I have been quoted in print as saying "If you can't stick your head under the hood and see the engine, you better look at what's coming out of the tail pipe," and I suppose that this is as good a way as any of expressing the thought that clues from their emission-line spectra are also important for these objects.
My own research in recent years has been devoted almost entirely to the Seyfert galaxies, and in this review I will deal chiefly with subjects I know and understand myself. Continuity suggests and the information we have on luminosity functions seems to confirm quantitatively that Seyfert galaxies and QSOs form a continuous sequence. This is indicated schematically in Figure 1, reprinted with permission from Morgan and Dreiser (1983). In recent years various authors have made many apparent detections of galaxies around objects formerly described as "quasistellar," from images and spectra obtained with CCDs and other current detectors which have high quantum efficiency, digital readout, and subtraction capabilities. Two recent excellent reviews of this subject, containing many references to the original papers, are by Hutchings (1983) and Miller (1985). Schmidt and Green (1985) in their Bright Quasar Survey simply adopted absolute magnitude MB = - 23 as the division between Seyfert 1 galaxies and QSOs as a practical working definition.
Figure 1. One-dimensional sequence of four AGNs, illustrating the range from Seyfert galaxies to QSOs (W.W. Morgan and R.D. Dreiser, 1983).
Continuity is important, but it must always be remembered that it does not necessarily imply that all AGNs are alike, any more than continuity of stellar properties implies that all stars are alike. For instance an O star and a K star are both in gravitational equilibrium, and both produce energy by nuclear reactions, but beyond these common properties of stars they are quite different in their structures and in their specific nuclear-energy-production reactions. The same is true of a K giant and a K dwarf. We must not assume that anything we learn from one AGN is true for all AGNs, any more than we assume that all stars have mass 1 M, or produce energy by the proton-proton cycle.