Next Contents Previous


Early in the twentieth century, Fath (1909) undertook at Lick Observatory a series of observations aimed at clarifying the nature of the "spiral nebulae". A major question at the time was whether spirals were relatively nearby, gaseous objects similar to the Orion nebula, or very distant collections of unresolved stars. Fath's goal was to test the claim that spirals show a continuous spectrum consistent with a collection of stars, rather than the bright line spectrum characteristic of gaseous nebulae. He constructed a spectrograph designed to record the spectra of faint objects, mounted it on the 36-inch Crossley reflector, and guided the long exposures necessary to obtain photographic spectra of these objects. For most of his objects, Fath found a continuous spectrum with stellar absorption lines, suggestive of an unresolved collection of solar type stars. However, in the case of NGC 1068, he observed that the "spectrum is composite, showing both bright and absorption lines". The six bright lines were recognizable as ones seen in the spectra of gaseous nebulae.

The bright and dark lines of NGC 1068 were confirmed by Slipher (1917) with spectra taken in 1913 at Lowell Observatory. In 1917, he obtained a spectrum with a narrow spectrograph slit, and found that the emission lines were not images of the slit but rather "small disks", i.e., the emission was spread over a substantial range of wavelengths. (However, he rejected an "ordinary radial velocity interpretation" of the line widths.) During the following years, several astronomers noted the presence of nuclear emission lines in the spectra of some spiral nebulae. For example, Hubble (1926) mentioned that the relatively rare spirals with stellar nuclei show a planetary nebula type spectrum, notably NGC 1068, 4051, and 4151.

The systematic study of galaxies with nuclear emission lines began with the work of Seyfert (1943). Seyfert obtained spectrograms of 6 galaxies with nearly stellar nuclei showing emission lines superimposed on a normal G-type (solar-type) spectrum: NGC 1068, 1275, 3516, 4051, 4151, and 7469. The two brightest (NGC 1068, 4151) showed "all the stronger emission lines ... in planetary nebulae like NGC 7027." Seyfert attributed the large widths of the lines to Doppler shifts, reaching up to 8,500 km s-1 for the hydrogen lines of NGC 3516 and 7469. The emission-line profiles differed from line to line and from object to object, but two patterns were to prove typical of this class of galaxy. The forbidden and permitted lines in NGC 1068 had roughly similar profiles with widths of ~ 3000 km s-1. In contrast, NGC 4151 showed relatively narrow forbidden lines, and corresponding narrow cores of the permitted lines; but the hydrogen lines had very broad (7500 km s-1) wings that were absent from the profiles of the forbidden lines. Seyfert contrasted these spectra with the narrow emission lines of the diffuse nebulae (H II regions) seen in irregular galaxies and in the arms of spiral galaxies. Galaxies with high excitation nuclear emission lines are now called "Seyfert galaxies". However, Seyfert's paper was not enough to launch the study of AGN as a major focus of astronomers' efforts. The impetus for this came from a new direction - the development of radio astronomy.

Jansky (1932), working at the Bell Telephone Laboratories, conducted a study of the sources of static affecting trans-Atlantic radio communications. Using a rotatable antenna and a short-wave receiver operating at a wavelength of 14.6 m, he systematically measured the intensity of the static arriving from all directions throughout the day. From these records, he identified three types of static: (1) static from local thunderstorms, (2) static from distant thunderstorms, and (3) "a steady hiss type static of unknown origin". The latter seemed to be somehow associated with the sun (Jansky 1932). Continuing his measurements throughout the year, Jansky (1933) observed that the source of the static moved around in azimuth every 24 hours, and the time and direction of maximum changed gradually throughout the year in a manner consistent with the earth's orbital motion around the sun. He inferred that the radiation was coming from the center of the Milky Way galaxy. After further study of the data, Jansky (1935) concluded that the radiation came from the entire disk of the Milky Way, being strongest in the direction of the Galactic center.

Few professional astronomers took serious note of Jansky's work, and it fell to an engineer, working at home in his spare time, to advance the subject of radio astronomy. Reber (1940a, b) built a 31 foot reflector in his backyard near Chicago. He published a map of the radio sky at 160 MHz showing several local maxima, including one in the constellation Cygnus that would prove important for AGN studies (Reber 1944). He also noted that the ratio of radio radiation to optical light was vastly larger for the Milky Way than the sun.

With the end of World War II, several groups of radio engineers turned their efforts to the study of radio astronomy. Notable among these were the groups at Cambridge and Manchester in England and at CSIRO in Australia. The study of discrete sources began with the accidental discovery of a small, fluctuating source in Cygnus by Hey, Parsons, and Phillips (1946) in the course of a survey of the Milky Way at 60 MHz. With their 6 degree beam, they set an upper limit of 2 degrees on the angular diameter of the source. The intensity fluctuations, occurring on a time scale of seconds, were proved a few years later to originate in the earth's ionosphere; but at first they served to suggest that the radiation "could only originate from a small number of discrete sources". The discrete nature of the Cygnus source was confirmed by Bolton and Stanley (1948), who used a sea-cliff interferometer to set an upper limit of 8 arcmin to the width of the source. These authors deduced a brightness temperature of more than 4 × 106 K at 100 MHz and concluded that a thermal origin of the noise was "doubtful". Bolton (1948) published a catalog of 6 discrete sources and introduced the nomenclature Cyg A, Cas A, etc. Ryle and Smith (1948) published results from a radio interferometer at Cambridge analogous to the optical interferometer used by Michelson at Mt. Wilson to measure stellar diameters. Observing at 80 MHz, they set an upper limit of 6 arcmin to the angular diameter of the source in Cygnus.

Optical identifications of discrete sources (other than the sun) were finally achieved by Bolton, Stanley, and Slee (1949). Aided by more accurate positions from sea cliff observations, they identified Taurus A with the Crab Nebula supernova remnant (M 1); Virgo A with M 87, a large elliptical galaxy with an optical jet; and Centaurus A with NGC 5128, an elliptical galaxy with a prominent dust lane. The partnership of optical and radio astronomy was underway.

The early 1950s saw progress in radio surveys, position determinations, and optical identifications. A class of sources fairly uniformly distributed over the sky was shown by the survey by Ryle, Smith, and Elsmore (1950) based on observations with the Cambridge interferometer. Smith (1951) obtained accurate positions of four discrete sources, Tau A, Vir A, Cyg A, and Cas A.

Smith's positions enabled Baade and Minkowski (1954) to make optical identifications of Cas A and Cyg A in 1951 and 1952. At the position of Cyg A, they found an object with a distorted morphology, which they proposed was two galaxies in collision. Baade and Minkowski found emission lines of [Ne V], [O II], [Ne III], [O III], [O I], [N II], and Halpha, with widths of about 400 km s-1. The redshift of 16,830 km s-1 implied a large distance, 31 Mpc, for the assumed Hubble constant of H0 = 540 km s-1 Mpc-1. The large distance of Cyg A implied an enormous luminosity, 8 × 1042 erg s-1 in the radio, larger than the optical luminosity of 6 × 1042 erg s-1. (Of course, these values are larger for a modern value of H0.)

This period also saw progress in the measurement of the structure of radio sources. Hanbury Brown, Jennison, and Das Gupta (1952) reported results from the new intensity interferometer developed at Jodrell Bank, including a demonstration that Cyg A was elongated, with dimensions roughly 2 arcmin by 0.5 arcmin. Interferometer measurements of Cyg A by Jennison and Das Gupta (1952) showed two equal components separated by 1.5 arcmin that straddled the optical image, a puzzling morphology that proved to be common for extragalactic radio sources.

Radio sources were categorized as `Class I' sources, associated with the plane of the Milky Way, and `Class II' sources, isotropically distributed and possibly mostly extragalactic (e.g., Hanbury Brown 1959). Some of the latter had very small angular sizes, encouraging the view that many were "radio stars" in our Galaxy. Morris, Palmer, and Thompson (1957) published upper limits of 12 arcsec on the size of 3 class II sources, implying brightness temperatures in excess of 2 × 107 K. They suggested that these were extragalactic sources of the Cyg A type.

Theoretically, Whipple and Greenstein (1937) attempted to explain the Galactic radio background measured by Jansky in terms of thermal emission by interstellar dust, but the expected dust temperatures were far too low to give the observed radio brightness. Reber (1940a) considered free-free emission by ionized gas in the interstellar medium. This process was considered more accurately by Henyey and Keenan (1940) and Townes (1947), who realized that Jansky's brightness temperature of ~ 105 K could not be reconciled with thermal emission from interstellar gas believed to have a temperature ~ 10, 000 K. Alfvén and Herlofson (1950) proposed that "radio stars" involve cosmic ray electrons in a magnetic field emitting by the synchrotron process. This quickly led Kiepenheuer (1950) to explain the Galactic radio background in terms of synchrotron emission by cosmic rays in the general Galactic magnetic field. He showed order-of-magnitude agreement between the observed and predicted intensities, supported by a more careful calculation by Ginzburg (1951). The synchrotron explanation became accepted for extragalactic discrete sources by the end of the 1950's. The theory indicated enormous energies, up to ~ 1060 ergs for the "double lobed" radio galaxies (Burbidge 1959). The confinement of the plasma in these lobes would later be attributed to ram pressure as the material tried to expand into the intergalactic medium (De Young and Axford 1967). A mechanism for production of bipolar flows to power the lobes was given by the "twin exhaust model" of Blandford and Rees (1974).

The third Cambridge (3C) survey at 159 MHz (Edge et al. 1959) was followed by the revised 3C survey at 178 MHz (Bennett 1962). Care was taken to to minimize the confusion problems of earlier surveys, and many radio sources came to be known by their 3C numbers. These and the surveys that soon followed provided many accurate radio positions as the search for optical identifications accelerated. (AGN were also discovered in optical searches based on morphological "compactness" [Zwicky 1964] and strong ultraviolet continuum [Markarian 1967] and later infrared and X-ray surveys.) Source counts as a function of flux density ("log N - log S") showed a steeper increase in numbers with decreasing flux density than expected for a homogeneous, nonevolving universe with Euclidean geometry (e.g., Mills, Slee, and Hill 1958; Scott and Ryle 1961). This was used to argue against the "steady state" cosmology (Ryle and Clark 1961), although some disputed such a conclusion (e.g., Hoyle and Narlikar 1961).

Next Contents Previous