Annu. Rev. Astron. Astrophys. 1999. 37: 445-486 Copyright © 1999 by Annual Reviews. All rights reserved

7.2. The Discovery of Quasars and the AGN Phenomena

The discovery of quasars is too tangled a story to yet be described fully here. A summary of the principal events leading to the discovery of the redshift of 3C 273 is given in the important account by Schmidt (1975, from an original manuscript of 1969). Fuller accounts are given in the monograph by Burbidge & Burbidge (1967) and the reviews by Sullivan (1982, 1984), Hartwick & Schade (1990).

Soon after the 3C and 3CR Cambridge catalogs of the brightest radio sources were published, efforts were made using interferometry to determine the angular sizes of the various radio sources. Particularly important for the quasar history was the long base-line interferometry made by the Jodrell Bank radio astronomers. The group, led by Henry Palmer, moved three portable cylindrical parabolic antennas with effective diameters of 28 feet over a number of baselines in England. They performed interferometry with the Jodrell Bank 250-foot antenna, measuring the fringe visibilities of more than 350 3CR sources. With baselines ranging from 2200 to 61,100 wavelengths (Allen et al 1962, Rowson 1962), they compiled a list of unresolved 3C sources whose angular diameters were smaller than one arcsec, and therefore had "brightness temperatures" larger than 10⁷ K, a telling indication of nonthermal radiation (Palmer 1961). Also from these data, Allen, Hanbury Brown, & Palmer (1962) surmised that many resolved 3CR sources were double, following the previous demonstration by Hanbury Brown & Das Gupta that Cygnus A is double.

T.A. Matthews, a recent addition to Bolton's Caltech radio astronomy group, was fascinated with the possibilities of optical identifications of the 3C sources using the Owens Valley radio interferometer, composed of two 90-foot movable radio dishes. Matthews also had good communications with the Manchester group of Palmer, and he also knew of the long baseline work they were doing.

Palmer had given an account of the Jodrell Bank work at one of the first Texas symposia on high energy astrophysics, which Matthews also attended (circa 1959; my records are incomplete). At that meeting (or shortly thereafter) Palmer gave more details privately to Matthews, including a rather complete list of 3C sources that were unresolved at 61,100 wavelength separation. Palmer's list was a subset of the complete list published later by Allen et al (1962).

In late 1959 Matthews began a program of position determinations in RA, and supervised a Ph.D. thesis program by Read (1963) for declination positions. He also made a working list of positions determined by others, including those by the Palmer group. In 1960 Matthews suggested a collaboration with optical astronomers with access to the 200-inch to begin the final phases of an optical identification program. It was through the invitation of Matthews that I became involved in the optical identification program that was to last far beyond the quasar discovery era, until most of the 3CR radio sources had in fact been identified.

At first we concentrated on Palmer's list of unresolved sources at 61,000-wavelength baseline. There were at least 20 such sources on our initial observing list (which has been lost), all of which later turned out to be radio-loud quasars. I began to take plates with the 200-inch in 1959, centered on Matthews' radio positions. The plates were measured by Matthews in Pasadena, from which the first three star-like identifications of 3C 48, 3C 196, and 3C 286 (Matthews & Sandage 1963) had been made by early 1960. All three were abnormally blue in the ultraviolet as determined from photoelectric photometry begun with the 200-inch in 1960, showing that 3C 48 varied in intensity by 0.4 mag over a one year period. The object also had a totally abnormal spectrum that I could not decipher. The rest (almost) is history, as set out by Schmidt (1975).

"Almost" refers to the crucial radio position determined from a series of lunar occultations by Hazard, Mackey, & Shimmins (1963) for 3C 273. This position permitted Schmidt to identify the optical image on one of the 200-inch plates that was in Matthews' possession, which Matthews had evidently turned over to Schmidt. The Hazard et al position of the two radio components (one centered on the optical star-like object and the other at the end of the obvious jet) were accurate to 1 arcsec, making the optical identification secure.

The heroes of the quasar discoveries were clearly Palmer for his brilliant measurements with his Jodrell Bank colleagues of the radio angular diameters, Matthews for all aspects of the identification work and for the organization of the joint Owens Valley-Palomar collaboration, Hazard for the position of 3C 273, and Oke (1963), Schmidt (1963) for their joint discovery of the redshift of 3C 273. ^{(5), (6)}

Beginning in 1963 the quasar program became frenzied with the 3C 273 redshift discovery, not only at Palomar, but also at Kitt Peak, Lick, and Hawaii, with rivalry between all groups and within each group often leading to severe tension.

Quasars come in all radio and optical luminosities. The most luminous in both optical and radio absolute magnitudes turned out, in fact, to be those identified from the 3CR catalog, which is not surprising when understood based on the ideas of bias in flux-limited catalogs (e.g. Teerikorpi 1997).

After 1963, radio weak (or quiet) quasars were discovered (Sandage 1965). The least luminous of these in both optical and radio luminosity blend with the N galaxies discussed in an earlier section. The continuum continues to the Seyfert galaxies, and finally to the LINERS mentioned earlier.

After the first few identifications of objects in the 3CR radio catalog had been made, the program developed into an attempt to identify all objects in the catalog. It was known by 1970 that perhaps 25 percent of the 3CR sources were not visible to the plate limits of the Palomar 48-inch Schmidt, and a number of these were also not seen on routine exposures at the known radio position with the 200-inch. A review of the remaining problems up to 1974 was given by Kristian & Minkowski (1975).

To complete the work, a cooperative program was begun with Campbell Wade using the new three-element interferometer composed of three movable 85 foot dishes at the National Radio Astronomy Observatory (NRAO) in Greenbank, West Virginia, to obtain highly precise radio positions, from which Kristian and the writer would measure deep 200-inch plates to attempt to complete the 3CR identifications. The program was highly successful (Wade 1970, Wade, Sandage, & Kristian 1970, Brosche, Wade, & Hjellming 1973, Kristian & Sandage 1970, Kristian, Sandage, & Katem 1974, 1978), providing exceptionally faint galaxy candidates for the spectrographic observers of the 1980s, mostly at Lick and Kitt Peak. The redshifts and subsequent photometry extended the Hubble diagram to the new limits that were achieved in the 1980s.

⁵ The breaking of the redshift code in the spectrum of 3C 273 was done by Oke and Schmidt together when they combined their wavelength data and realized that they were seeing the Balmer series. Back.

⁶ The joys of doing science are boundless. Sometime during the long subsequent campaign with Kristian to optically identify all the remaining 3C sources, Iwas in communication with D.O. Edge, lead author of the 3CR Cambridge Catalog. Although we had never met, we had written in earlier letters about something concerning 3C 273 (the correspondence is lost). Edge, who is a devout churchgoer, confided that indeed 3C 273 was the radio source he liked best. The reason was that 273 is the number of his favorite anthem in the Methodist Hymnal. Back.