Annu. Rev. Astron. Astrophys. 1977. 15: 69-95 Copyright © 1977 by Annual Reviews. All rights reserved

3. SURVEYS AND MORPHOLOGY

In 1943, Seyfert studied six galaxies that had been noted to have unusual emission lines in the collection of Mt. Wilson spectra accumulated for redshift studies. The sample did not increase significantly for 25 years, until the Markarian and Zwicky surveys. Beginning in 1967, Markarian has published seven lists containing about 700 galaxies (with two more lists in press as of this writing) distinguished by their unusually strong ultraviolet continua (Markarian 1967, 1969a, 1969b, Markarian & Lipovetsky 1971, 1972, 1973, 1974). (Improved positions for some of these are given by Peterson 1973.) These objects have a space density of about one per ten square degrees and were found on low-dispersion, objective prism survey plates obtained with the 1-m Schmidt at Byurakan Observatory. Subsequent studies showed that the majority of the Markarian galaxies have emission-line spectra (see references in Weedman 1977). The Markarian survey reaches to 17 mag but seems to be complete only to about 16 mag (Huchra & Sargent 1973). Approximately 10% of the Markarian galaxies are Seyferts, and the subsequent spectroscopic discovery observations of the Seyferts thus found are referenced in Table 1. In contrast to Markarian's spectroscopic survey techniques, Zwicky compiled lists of compact galaxies based on their morphological appearance on the Sky Survey plates. He selected objects that were almost stellar, having soft edges or faint associated nebulosity. These lists were never published, but spectroscopic surveys primarily by Sargent (1970a) of a number of Zwicky's compact galaxies revealed several Seyferts.

Other surveys analogous to these are now underway and have already provided some Seyfert galaxies. A southern hemisphere objective prism survey is being conducted with the 61-cm Curtis Schmidt telescope at Cerro Tololo (Smith 1975, 1976, Smith et al. 1976). The Tololo survey has higher resolution than the Markarian survey and selects objects because they have emission lines in their spectra. An important variety of emission-line galaxies have been found this way; observations so far available indicate that similar sorts of galaxies are found in the Tololo lists as in the Markarian lists. The Tololo survey as instituted and carried out by M. G. Smith covered approximately 3000 square degrees and resulted in the discovery of about 500 emission-line galaxies, whose publication is beginning (Smith et al. 1976). Another portion of the southern sky (-20° 8°) is being surveyed in the same way with the Curtis Schmidt by astronomers from the University of Michigan, which owns this telescope. Several hundred emission-line galaxies have been found in this region, and they are accessible from the north (G. MacAlpine, personal communication). Large-scale follow-up studies of all these Tololo galaxies, expected to yield numerous new Seyferts, are underway.

Morphologically oriented surveys have also been conducted recently. A list of 591 galaxies of high surface brightness has been prepared by Arakelian (1975). He estimated the surface brightnesses by comparing the magnitude measures in the Zwicky catalogs of galaxies (not the Zwicky compact galaxies) to the diameter measures in the Vorontsov-Velyaminov morphological catalogs of galaxies. Six new Seyferts have already been found in Arakelian's lists (Table 1). A spectroscopic survey of southern galaxies classified as having bright nuclei was carried out by Martin (1976), who found one new Seyfert in this way.

It is important, of course, to study the galaxies with bright nuclei that are not called Seyferts in hopes of finding just when and how the Seyfert activity turns on. We do not know, for example, whether there are galaxies with narrow emission lines that have nonthermal continuous spectra like the Seyferts. Certainly there are bright nucleus galaxies like NGC 4385 (Mkn 52) and NGC 7714 (Mkn 538) which have very strong but narrow emission lines with intensity ratios like those in conventional HII regions. The UBV colors correspond to collections of hot stars (Weedman 1973, Huchra 1976), but some of these galaxies have infrared luminosities comparable to the Seyferts (Rieke & Low 1972). Another intriguing result is that interacting galaxies seem to have strong nuclear emission lines unexpectedly often. Representative examples are NGC 2992 (Osmer et al. 1974b) and the aforementioned NGC 7714 (Burbidge 1968). Adams (1977) also noted a probable surplus of interacting systems among the Seyferts themselves. The large samples of galaxies from the Markarian and Tololo surveys may allow us to determine how such galaxies relate to the Seyfert problem.

From the number of Seyfert galaxies among bright Shapley-Ames galaxies, it was concluded that 1% of spiral galaxies are Seyferts (de Vaucouleurs & de Vaucouleurs 1968). This estimate remains valid for the larger sample now available, because Huchra & Sargent (1973) concluded that 5% of all galaxies were Markarian galaxies, of which 10% are Seyferts. Consequently, about 1 galaxy in 200 (of all types) is a Seyfert. Why? Do 1% of spiral galaxies remain Seyferts forever, or does each galaxy spend 1% of its life in the Seyfert phase? The current research relevant to this problem is into the morphological appearance of Seyfert galaxies to determine whether the presence of a Seyfert nucleus is correlated with any other property of a galaxy.

The most striking empirical result is a deficiency of ellipticals among the Seyferts. There are ten galaxies now known to be Seyferts that are bright enough to be Shapley-Ames galaxies. Van den Bergh (1975b) has emphasized the fact that all ten are spirals. A comprehensive morphological study of 80 Seyfert galaxies was recently completed by Adams (1977). He finds that Seyferts occur in a broad distribution of ordinary and barred spirals but that there are very few ellipticals. Counting those galaxies with amorphous but unclassifiable envelopes, Adams finds a maximum allowable limit of 10% for the fraction of Seyferts that could be ellipticals, but the more reasonable limit is 5%. By contrast, 33% of the galaxies in Humason et al. (1956) are ellipticals. Because many of the Seyfert galaxies have such small angular sizes as to be unclassifiable, it is possible that none are ellipticals. The only Seyfert that can be called a giant elliptical because of its dominant appearance in a cluster is NGC 1275. It has many other properties that are unique among Seyferts, so much so that it is clearly an anomalous object. (Had it not been included in Seyfert's original study, it probably would not have been grouped with the other galaxies called Seyferts today, but more likely would be considered a peculiar radio galaxy.) Adams and van den Bergh discuss the implications of this elliptical deficiency for our understanding of QSOs, as the morphology of Seyferts is in dramatic contrast to that of the radio galaxies.

Previous studies had pointed out the peculiar appearance of certain classical Seyferts, especially the presence of faint annular structures (Burbidge et al. 1963, Hodge 1968). Adams adds some examples of these as well. Such effects could be important for demonstrating interactions between the disk and the nucleus, although they would not show whether the nuclear activity triggered changes in the disk or vice-versa. A possible correlation between the spectroscopic properties of the nuclei and the appearance of the galactic disk was suggested by Khachikian & Weedman (1971). To some degree, Adams confirms this in the sense that no Seyferts of spectroscopic class 2 (see below) have spiral structure as conspicuous as that in some of class 1. The only suggested explanation so far for the various peculiarities referenced is that mass outflow from the nuclei of the Seyferts is somehow affecting the galactic disks.

The above-mentioned morphological studies, which demonstrate the spiral nature of many Seyfert galaxies, are of special importance regarding the redshift controversy. Continuity arguments demonstrating a luminosity overlap between Seyferts and QSOs provide empirical evidence for cosmological redshifts of QSOs, if the Seyfert redshifts are cosmological. It is therefore imperative to show that some "Seyfert galaxies" whose cosmological luminosities would compare to QSOs really are distant galaxies. The presence of spiral arms is an acceptable morphological indication that a nebulous object really is a galaxy. For more amorphous envelopes, it has been suggested that absorption lines must be detected in the envelopes before they can be considered as galaxies made of stars (Burbidge 1973). Such absorption lines are difficult to detect. No such lines are found even in the outer disk of the spiral Seyfert NGC 4151 (Simkin 1975). However, extensive photometric studies of the disks of the larger Seyferts show that their colors correspond to those expected for galaxies of stars (Penston et al. 1974, Dibay & Lyutyi 1971, Walker et al. 1974, Zasov & Lyutyi 1973). For higher redshift Seyferts, a morphological demonstration that real galaxies are present is difficult. Adams (1977) finds that only 21% of all Seyferts are unresolved or have amorphous main bodies, but that this fraction is 80% for those with redshifts above 20,000 km sec^-1. This is as expected, of course, if the Seyferts are really galaxies whose angular size decreases in proportion to their distance.