Observing from the Cape of Good Hope, Herschel (1847) was the first to note the peculiar nature of NGC 5128: "A nebula consisting of two lateral portions and no doubt of a small streak of nebula along the middle of the slit or interval between them, having a star at its extremity" and " ... a very problematical object, and must be regarded as forming a genus apart, since it evidently differs from mere `double nebulae' ...". Although Hubble (1922) included the galaxy in a list of (mostly galactic) nebulae with line emission, it was mostly ignored until the advent of radio astronomy about 50 years ago (but see Paraskevoulos 1935). Shortly after World War II, Bolton (1948) announced the discovery of six discrete sources of 100 MHz radio emission in addition to the already known case of Cygnus A. Precisely a year later, using their remarkable "cliff-interferometry" technique, Bolton, Stanley & Slee (1949) identified the sources Taurus A, Virgo A and Centaurus A with the nebulous objects M1 (Crab nebula), NGC 4486 (M 87) and NGC 5128 respectively. Although they were actually convinced of the extragalactic nature of the latter two, their published discussion was noncommittal, reflecting referee objections to this notion (R.D. Ekers, private communication). More precise measurements by Mills (1952) confirmed the identification of Centaurus A with NGC 5128. Discussing the nature of optical counterparts to the newly discovered radio sources, Baade & Minkowski (1954) established NGC 5128 as an unusual extragalactic system, located in space well "beyond M 31". They described it as "... an unresolved E0 nebula with an unusually strong and wide central dark band, a combination highly anomalous for a spherical nebula" and interpreted it as the interaction or merger of an elliptical and a spiral galaxy. Nevertheless, the merger hypothesis then receded into the background until it was revived some 25 years later by the work of Graham (1979), Dufour et al. (1979), Tubbs (1980) and Malin, Quinn & Graham (1983). It is now the favoured explanation for the unusual characteristics of NGC 5128.
Other radio measurements soon established the nonthermal nature of the radio emission (Haddock, Mayer & Sloanaker 1954). Subsequent radio observations with increased sensitivity yielded ever-larger dimensions for the object until the low-frequency studies by Sheridan (1958) and Shain (1958) first established an overall size of 8° × 4° for Centaurus A - much larger than the 20' size of the optical parent galaxy NGC 5128. A multi-frequency radio study by Cooper, Price & Cole (1965) showed that at decimetre wavelengths the source consists of bright central emission, representing 20% of the total flux density and much larger lower-brightness lobes to the north and south, with 40% of the radio flux each. The advent of interferometric techniques, in particular aperture synthesis radio telescopes with their superior angular resolution, has since led to the discovery of many more similar radio galaxies (cf. Allen, Hanbury Brown & Palmer 1963; see review by Miley 1980). It has also become clear that many other elliptical galaxies have more or less pronounced dust lanes (Hawarden et al. 1981; Sadler & Gerhard 1985; van Dokkum & Franx 1995), including the other major southern radio galaxy NGC 1316 (Fornax A).
Both its striking optical appearance (Fig. 1), and its associated giant radio source have made NGC 5128 (1) into one of the most extensively studied galaxies in the southern hemisphere. However, in their comprehensive review Ebneter & Balick (1983) point out that NGC 5128 if placed at similar large distances would look very much like other radio elliptical galaxies, in spite of its apparent peculiarity. By its fortuitous proximity, Centaurus A uniquely allows detailed studies aimed at determining the nature of the galaxy and in particular the origin of the giant radio source it is hosting. A probably incomplete search of the literature published since Ebneter & Balick's (1983) review reveals over a hundred refereed papers mentioning Centaurus A in the title, whereas in the same period the number of papers referring to Centaurus A was almost 700. This review focusses on the observed properties of Centaurus A / NGC 5128 and their immediate interpretation. For a more complete review of the early work up to 1983, the reader is referred to Ebneter & Balick (1983), and for more general information to recent proceedings such as "The Second Stromlo Symposium - The Nature of Elliptical Galaxies" (Eds. N. Arnoboldi, G.S. da Costa, P. Saha, 1997 PASP Conference Series Vol. 116), "Energy Transport in Radio Galaxies and Quasars" (Eds. P.E. Hardee, A.H. Bridle, J.A. Zensus, 1996 PASP Conference Series Vol. 100) and "Extragalactic Radio Sources" (Eds. R. Ekers, C. Fanti, L. Padrielli, 1996 IAU Symposium 175).
|
Figure 1. B-band image of NGC 5128 shows the traditional, almost circular inner part of the elliptical galaxy, crossed by a complex "dark band". The dark band is the projection of a strongly warped thin disk (Section 4). Note bright star clusters at the northwestern and southeastern edges of the band, and isolated dust clouds silhouetted against the stellar background. Supernova 1986g is the brightest of the two stars seen in the dark band about 45 mm from the left and 40 mm from the bottom. (Courtesy D. Malin, Ango-Australian Observatory) |
With its great apparent brightness and galactocentric velocity of about V0 = +325 km s-1, NGC 5128 is clearly a nearby galaxy. Early estimates of its distance ranged from 2.1 Mpc (Sersic 1958) to 8.5 Mpc (Sandage & Tammann 1974) and a "compromise" distance of 5 Mpc has frequently been used in the past. Many of the earlier distance values were derived by indirect means: group membership, redshift, comparison to Virgo cluster distance (assuming the latter to be known) etc. Summaries of the various distance determinations prior to 1993 can be found in Shopbell, Bland-Hawthorn & Malin (1993) and especially de Vaucouleurs (1993).
Attempts to use the occurrence of the SNIa supernova 1986g in NGC 5128 to derive its distance (Frogel et al. 1987; Phillips et al. 1987; Ruiz-Lapente et al. 1992) were frustrated by uncertainties in the internal reddening of NGC 5128 and in the absolute magnitudes of the supernova (see Sect. 3.3). It appears more useful to assume a distance to determine the supernova properties than the other way around. Other methods have been more succesful, consistently yielding a relatively nearby distance. The accuracies of these techniques were reviewed by Jacoby et al. (1992). Tonry & Schechter (1990) interpret the globular cluster counts by Harris et al. (1984a, 1986) to imply a distance modulus (m - M)0 = 27.53 ± 0.25 mag. Planetary nebula counts by Hui et al. (1993a) yield (m - M)0 = 27.73 ± 0.14. The lower value (m - M)0 = 27.48 ± 0.06 obtained by Tonry & Schechter (1990) from the surface brightness fluctuation of globular clusters must be revised to (m - M)0 = 27.71 ± 0.10 (Tonry 1991). Finally, Soria et al. (1996) estimate (m - M)0 = 27.72 ± 0.20 from halo red giant branch stars observed with HST/WFPC2. As there is no a priori preference for any of these methods, and as the weighted and unweighted averages of the various derived distance moduli are practically identical, the best value is thus (m - M)0 = 27.67 ± 0.10, corresponding to a distance of D = 3.4 ± 0.15 Mpc. Note that at this distance, 1' on the sky corresponds almost precisely to 1 kpc. (2)
The systemic velocity of NGC 5128 is most frequently derived from measurements of components of the disk (Sect. 4). An overview of estimates up to 1984 can be found in Hesser et al. (1984); Table 1 lists all estimates a. later than 1975; b. having a quoted accuracy of 10 km s-1 or better; and c. not based on HI or molecular line absorption (see Sect. 7). The estimates range from 536 to 551 km s-1, and the mean is VHel = 543 ± 2 km s-1.
| Systemic Velocity | Obtained from: | Reference |
| VHel (km s-1) | ||
| 548 ± 5 | nebular emission lines | Graham (1979) |
| 551.4 | H emission
| Whiteoak & Gardner (1979) |
| 541 ± 8 | nebular emission lines | Rodgers & Harding (1980) |
| 545 ± 5 | H emission
| Marcelin et al. (1982) |
| 538 ± 10 | stellar absorption lines | Wilkinson et al. (1986) |
| 536 ± 5 | H emission
| Bland et al. (1987) |
| 542 ± 7 | HI emission | van Gorkom et al (1992) |
| 541 ± 5 | CO modelling | Quillen et al. (1992) |
| 541 ± 7 | PN emission lines | Hui et al. (1995) |
Based on position and radial velocity, NGC 5128 is part of a group
of 25 galaxies extending over about 25° on the sky
(de Vaucouleurs 1975;
Webster et al. 1979;
Hesser et al. 1984;
Côté 1995,
1997).
In addition to NGC 5128, major members of the group are NGC 4945 and NGC 5236 (M 83) as well as the lesser NGC 5102, NGC 5253 and some twenty dwarf
galaxies (Fig. 2). NGC 5128 is the only massive
elliptical in the group.
The group galactocentric velocity is very close to the
individual velocities V0 = 320-330
km s-1 of the major members
NGC 4945, NGC 5128 and NGC 5236
(Hesser et al. 1984).
The members have an average
projected radial distance to the centre of mass of 0.72 Mpc. The group has
a crossing time of 5 gigayear and is probably still collapsing
(Côté 1995,
1997).
Assuming a distance of 3.4 Mpc, the group mass estimate by
Hesser et al. (1984)
becomes 5-17 × 1012 M
. It is quite remarkable
that the other two major members, NGC 4945 and NGC 5236, as well as
NGC 5253, exhibit signs of unusually vigorous star
formation, while
NGC 4945 also has a
nuclear outflow. In addition, the dwarfs in the group have relatively high
central surface brightnesses and exhibit clear star formation activity
(Côté 1995).
|
Figure 2. The Centaurus group of galaxies, with its brighter members identified. From Côté et al. (1997). |
NGC 5128, itself a strong X-ray source (Sects. 2.4; 5.5), is surrounded by several X-ray point sources of a nature still unknown, although the majority of them seems to be associated with the galaxy (Arp 1994; Döbereiner et al. 1996).
1 Useful photometric UBVR(I) sequences for NGC 5128 have been published by Graham (1980), McElroy & Humphreys (1982) and Zickgraf et al. (1990), reaching down to visual magnitudes of 16.6, 18.2 and 22.7 magnitudes respectively. Astrometric coordinates of field stars near the centre of NGC 5128 can be found in Griffin (1963). Very accurate field star positions and magnitudes may be obtained from http://astro.estec.esa.nl/SA-general/Projects/Hipparcos. Back.
2 In the following all specific values quoted are reduced to D = 3.4 Mpc so that they may be different from the published values as given in the associated reference. Back.