1.1. A Short History of AGN Search Techniques
From a historical perspective (e.g., Osterbrock 1991), the first indications of "non-stellar" activity in the nuclei of galaxies came from the discovery of strong, broad emission lines in NGC1068 (Fath 1913; Slipher 1917) and the discovery of the jet in M87 (Curtis 1917). The spectral features found in NGC1068 are almost never found in stars or supernovae remnants and thus were an indication of some new phenomena occurring. It took 50 years before the morphology of the "jet" in M87 was related to non-stellar processes. 3 The first "sample" of non-stellar activity was that of Seyfert in 1943, who found a wide variety of strong "broad" lines in the nuclei - but not elsewhere - of several otherwise "normal" galaxies. It was clear from this early paper that there was something quite unusual about these sources and that they must be fairly common, but it took almost 20 more years for significant progress to be made.
This next step occurred with the discovery of extragalactic radio sources and attempts to find optical counterparts for them. The discovery of low to moderate redshift "active galaxies" as the optical counterparts to several bright radio sources (Baade & Minkowski 1954) showed a new type of "active galaxy". It was clear even in 1958 (Minkowski) that there was tremendous scatter in the optical properties of the identifications at a fixed radio flux. Schmidt (1963), Greenstein & Matthews (1963), and Schmidt & Matthews (1964) discovered that the optical counterparts of several luminous radio sources were stellar-looking sources at large redshifts, and thus were very luminous, compact, extragalactic sources with non-stellar spectra. They were subsequently named "quasars" for quasi-stellar radio sources. The optical properties of these radio sources were very similar to each other, indicating that a class of objects had been found. Sandage (1965) realized that there were sources with the same general optical properties as quasars that were not radio sources. These sources had blue colors, meaning a large ultraviolet (UV) flux relative to the classical optical band, fairly high variability in their continuum intensity, and most had strong, broad emission lines over a wide range of ionization.
It was rapidly realized that the nuclei of some "Seyfert galaxies" had similar properties to quasars (Woltjer 1959; Burbidge, Burbidge, & Sandage 1963), and for the last 30 years, these sources have been grouped under the name Active Galactic Nuclei or AGN (I think that the first use of this name in the literature is from Burbidge 1970). However, even early on, it was clear that not all AGN resembled quasars. There were Seyfert 2 galaxies, which do not have broad lines or strong non-stellar continua but do have strong, narrow forbidden lines that could not be produced by ionization from normal stars. Also, there were BL Lacertae objects, which usually do not have optical or UV emission lines but do have a very strong non-thermal continuum and a wide variety of optical colors. It was thus clear from the start that identifying complete samples of AGN would require a wide variety of techniques and criteria.
Outside of trying to identify the optical counterparts to the radio sources (it took over 30 years to completely identify all of the sources in the 3CR survey, the first large radio survey; Spinrad et al. 1985), the first systematic search for AGN that I can find in the literature was the realization (Arp et al. 1968) that the Markarian survey of compact galaxies with blue stellar colors contained a large number of sources with the properties of Seyfert galaxies. At almost the same time, Sargent (1970) used similar critieria for sources from the Zwicky survey and also found numerous AGN. These two early works, combined with the radio surveys, set the standard for AGN surveys: using photographic techniques to find sources with compact, blue nuclei and following up with optical spectroscopy (see Weedman 1977 for an early review). However, these searches were completely empirical; i.e., they were looking for sources that had the properties of sources that one already knew were active galaxies.
It is clear that AGN have a very wide range of relative parameters, from the line strengths and line widths (ranging down to sources without any emission lines at all; e.g., BL Lacertae objects), to the continuum colors, to the amplitudes and timescales of variability. Thus, having an inclusive definition is very difficult. It is fair to say that most workers have had a difficult time coming up with an AGN definition that is totally complete and not subject to noise and strong selection effects. Recent large optical surveys, such as the Two Degree Field (2dF) and the Sloan Digital Sky Survey (SDSS), have focused on well-defined color or line strength criteria that allow them to be well defined but clearly incomplete.
3 We now know that jets can also occur in stellar processes (for example, in Herbig-Haro objects, young neutron stars [e.g., see the Chandra image of the Vela pulsar], supernovae remnants [e.g., see the Chandra image of Cassiopeia A], and the X-ray and radio emission from some luminous galactic X-ray sources [e.g., Cygnus X-3]). It seems as if the jet phenomenon is related to the emission of large amounts of energy over a short period of time and into a restricted volume. Back.