Annu. Rev. Astron. Astrophys. 1980. 18: 165-218 Copyright © 1980 by Annual Reviews. All rights reserved

4.4. Jets

Long before celestial radio sources were dreamed of, an optical jet was seen in the nebula M 87 (Curtis 1918). More than fifty years later a radio counterpart of this jet was found in the associated radio source Virgo A (Hogg et al. 1969, Wilkinson 1974, F.N. Owen and P. Hardee, in preparation). See Figure 10. The radio and optical spectra of this jet and its high optical polarization (Baade 1956, Hiltner 1959, Schmidt et al. 1978) are all indicative of nonthermal emission. On the identification of 3C273, the brightest known quasar, an optical jet was also seen in the direction of the extended radio emission (Hazard et al. 1963). For several years no further examples of jets were recognized; M 87 and 3C273 were regarded as exceptional, but in fact their most exceptional property is probably their proximity.

Figure 9. The radio source associated with NGC 6251 showing the alignment of the core and jet relative to the overall source structure (from Readhead et al. 1978a and kindly supplied by Marshall Cohen).

Figure 10. The jet and steep spectrum core of Virgo A/M 87 (VLA, 5 GHz) (from F.N. Owen and P. Hardee, in preparation and kindly supplied by the authors).

The situation changed in the mid-seventies. Long narrow radio structures were found to emanate from the nuclei of the radio galaxies 3C219 (Turland 1975b) and B 0844 + 31 (van Breugel & Miley 1977) and it was realized that analogous jet-like structures were also present at the heads of several tailed galaxies (van Breugel & Miley 1977). Several additional examples of jets were then recognized and at the time of writing about twenty radio jets are known. Some of the most prominent examples are in the double sources NGC 315 (Bridle et al. 1976, Bridle et al. 1979a), 3C31 (Burch 1977b, 1979c), NGC 6251 (Waggett et al. 1977), 3C449 (Perley et al. 1979, Bystedt & Högbom 1979) B 0844 + 31 (van Breugel 1980b), and HB 13 (Masson 1979) and the tailed sources 3C 83.1 B/NGC 1265 (Miley et al. 1975, Owen et al. 1978a), 3C129 (van Breugel & Miley 1977, Owen et al. 1979, Downes 1980), and 3C 465 (van Breugel 1980c). Some of these are illustrated in Figures 11, 9, 12, 14, 13 and 3.

Figure 11. The 1-Mpc radio source associated with NGC 0315 (Westerbork 0.6 GHz) showing the jet and its overall rotational symmetry. The cross indicates the position of the associated galaxy (from Willis 1978 and kindly supplied by the author).

Figure 12. The inner region of 3C449 (see Figure 2) showing the jet (VLA, 1.4 GHz) (from Perley et al. 1979 and kindly supplied by Ed. Fomalont).

Figure 13. The head of the tailed source 3C129 showing the symmetric wiggles (from van Breugel & Miley 1977).

To date, most radio jets have been found in nearby relatively low luminosity sources. This may be partly because nearby sources can be studied in greater detail. However, in the more luminous edge-brightened doubles the distinction between the features previously called "bridges" and "jets" is often unclear. Several quasars, e.g. 1004 + 13, 1047 + 09, 1058 + 11 (Miley & Hartsuijker 1978), and 4C 32.69 (Potash & Wardle 1979) have large-scale structures reminiscent of jets. Higher resolution observations with the VLA have recently confirmed the presence of jets in 4C 32.69 (Potash & Wardle 1980) and 1004 + 13 (E. Fomalont and G.K. Miley, in preparation).

4.4.1 ORIGIN Before the widespread existence of jets was suspected, a considerable body of evidence had accumulated that extended radio sources are powered quasi-continuously from the nuclei of their parent galaxies (Section 1.4). Although the existence of visible jets was not explicitly predicted by models of the energy transport, it seemed logical to associate these narrow structures that emanate from the radio cores and penetrate for several hundred kiloparsecs into the radio lobes with the umbilical cords that carry the energy required to nurture the extended emission. There are several byproducts of the energy-pumping process that could produce the observed radiation. Possibilities are collisions between freshly ejected plasmons and older "relics" (Christansen et al. 1977), or shocks within a relativistic beam (Rees 1978a, Blandford & Königl 1979). Dissipation at the edges of a beam has been suggested as a possible alternative (Blandford & Rees 1978), but since there is no indication that the sides of jets are enhanced (Perley et al. 1979), this is unlikely to be the dominant process for jets in low luminosity sources.

Within the context of these various mechanisms one can list three properties of the energy transport process that may exert a dominating influence on the observed luminosities and other properties of the jets.

1. The angle that the jet is inclined to the line of sight. If the jets represent outflow at relativistic velocities one might observe a dependence of intensity on aspect (cf Rees 1978a, Blandford & Königl 1979).

2. The amount of energy being conducted to the lobes. The highly energetic processes required to generate radiation in the jets may vary sporadically as a result of unsteady nuclear activity (Rees 1978a).

3. The efficiency of the energy transport. Inefficiency may be related to the degree to which energy transport is relativistic (Rees 1978a) or to the amount of interstellar matter entrained in the beam (Blandford & Königl 1979).

There is as yet an insufficiently large sample to investigate these questions properly, but simple interpretation of some of the individual morphological properties of jets provides several clues as to their nature and to the processes involved. Since, at the time of writing, sources containing radio jets are being studied intensively by the VLA, the following discussion is inevitably very preliminary.

4.4.2 SOME PROPERTIES The few radio jets that have been studied in detail show considerable diversity. For example, the lengths of the jets range from ~ 2 kpc in Virgo A to ~ 260 kpc in NGC 315 and their spectral indices from ~ -0.5 to ~ -0.8. Three properties of jets deserve special mention.

First, in several cases (e.g. 3C449, 3C83.lB/NGC 1265, NGC 315, 3C219) there appear to be "gaps" of a few kpc between the nuclear core and the location where the jets are seen to begin (Perley et al. 1979). This may be caused by large variations in the power of the nuclear machine, or to more efficient and hence less visible energy transport near the nucleus due to a more highly collimated and less randomized beam of relativistic electrons.

Second, several jets are observed to have wiggles on a scale of about 30 kpe. Examples are B0844+31 and 3C129 (van Breugel & Miley 1977, Owen et al. 1979), 3C 31 (Burch 1977a, b, 1979a,c, Fomalont et al. 1979), 3C449 (Perley et al. 1979, Bystedt & Högbom 1979), HB 13 (Masson 1979), and 3C83.1B/NGC 1265 (Owen et al. 1978a). Figure 13 shows the symmetric wiggles clearly apparent in the head of the tailed source 3C129. The wiggles have been attributed either to precessional motion of the central nuclear machine during the source lifetime (van Breugel & Miley 1977) or orbital motion of the parent galaxy (Blandford & Icke 1978, Bystedt & Högbom 1979).

Third, an intriguing aspect of jets that have been surveyed is their asymmetry. Except for 3C449 (Perley et al. 1979) and 3C83.1B/NGC 1265 (Owen et al. 1978a), the flux ratio between opposing jets is much larger than the flux ratio between the corresponding extended lobes (e.g. Willis et al. 1978, Cohen & Readhead 1979). Several possible explanations have been proposed to explain how asymmetric jets can power symmetric radio sources. Because of the existence of some two-sided jets the possibility that jets are accretion wakes formed by the motion of their parent galaxies through an intergalactic medium (Yabushita 1979) can be rejected. Another possibility is that the brightness of the jets (reflecting the efficiency of energy transport) may be strongly influenced by a nonuniform galactic environment, but it is not clear why the properties of the environment should differ on opposite sides of the nucleus. Third, opposing jets may have similar intrinsic intensities with the receding one apparently weakened by the Doppler effect (Rees 1978a, Blandford & Königl 1979). Typical flow velocities of ~ 0.3c would then be implied. A powerful argument against this view is the observation that some one-sided jets bend without undergoing significant intensity changes (E. van Groningen, C. Norman, and G.K. Miley, in preparation). Fourth, the nuclear engine may flip, supplying energy to the low lobes alternately (Willis et al. 1978). Again it is not clear how this could occur. A fifth possibility is that the beam shines through a clumpy nucleus and that these clumps occasionally block the beam.

4.4.3 OPTICAL EMISSION Optical emission from jets can provide unique information about the physics involved. Since synchrotron decay times for electrons emitting in the optical are typically ~ 10 yr, optical synchrotron features precisely locate regions where the electrons are accelerated (cf Section 3.1.4). An analogy with the optical/radio jet in M 87 prompted a deep search for optical counterparts of four other radio jets (Butcher et al. 1980). In two cases (3C66 and 3C31) optical jets were found with similar ratios of optical to radio emission as in M 87. These measurements suggest that a continuous nonthermal spectrum extending from ~ 10⁹ Hz to ~ 10¹⁵ Hz is a fairly common property of jets in radio galaxies and that acceleration of the relativistic particles occurs along the jet. The recent detection of X-ray emission from the M 87 jet (P. Gorenstein, E. Schreier, and E. Feigelson, private communication) shows that some jets emit over eight decades of frequency.

M 87 is the only nonthermal jet whose optical properties have been studied in some detail (e.g. de Vaucouleurs et al. 1968). Optical polarization measurements of the M87 jet (Schmidt et al. 1978) show a fairly ordered magnetic field that changes direction from knot to knot. The changes in the optical polarization angle along the jet are mimicked closely by the 6-cm radio polarization angle. This suggests that the electrons that produce the optical and radio emission originate in roughly the same regions of the magnetic field and that there is little Faraday rotation within the knots. However, the optical and 6-cm radio polarization angles are offset from each other by ~ 75° indicating considerable foreground rotation. The foreground rotation may be produced within the 2-kpc steep-spectrum core of Virgo A (Turland 1975a, Forster 1980) or perhaps within one of the 40-kpc extended radio lobes (B_me ~ 4 × 10^-6 G; Andernach et al. 1979) which coexists with the X-ray halo (n_t ~ - 10^-2 cm^-3; Catura et al. 1972).

Optical jets with significant thermal emission have been reported in the giant radio galaxies Centaurus A (Blanco et al. 1975, Dufour & van den Bergh 1978) and DA 240 (Burbidge et al. 1975, 1978). Although an X-ray jet has been seen in Centaurus A (Schreier et al. 1980), a radio jet has been seen in neither of these galaxies. The absence of an observed radio jet in Centaurus A may merely reflect the fact that this southern source is inaccessible to the most sensitive radio telescopes with sufficient resolution. Other extended radio sources in which jets may have been seen are the quasar PKS 0837-12, which has a visible extension along its radio axis (Wehinger & Wyckoff 1978), and some of the 18 candidates noted by Ghigo (1978).

4.4.4 CONE ANGLES AND MAGNETIC FIELDS/FREEDOM FOR JETS? Are the jets in pressure equilibrium with their surroundings? Model builders prefer jets to be free rather than confined. In a free jet the thermal electron density exceeds that of the surroundings, and Kelvin-Helmholtz instabilities, which may destroy a confined jet (Blandford & Pringle 1976, Turland & Scheuer 1976), would probably be unimportant.

For a free jet the cone angle subtended at the nucleus should be constant along its length. In the giant source NGC 6251 (Readhead et al. 1978a) this angle is indeed observed to be constant over a range of more than 10⁵ in length.

Also, simple arguments (Blandford & Rees 1978) predict that as a freely expanding jet widens (radius r increases) the parallel component of magnetic field should drop as r^-2, whereas the perpendicular component should vary only as r^-1. This is consistent with the magnetic field configuration observed in both 3C31 and NGC 315 (Fomalont et al. 1980) where parallel fields close to the nucleus change to predominantly perpendicular fields after about 2 kpc. Moreover, in 3C449 (Perley et al. 1979) and NGC 6251 (Readhead et al. 1978b) the magnetic field calculated from the minimum energy condition (Section 3.1.1) B_me scales roughly as r^-1, a further pointer that these jets are free, as well as weak evidence that the conditions within the jet are close to equipartition. On the assumption that the jet in 3C 449 is free and in equipartition together with several of the arguments of Section 3.1.2, Perley et al. (1979) find the velocity of electron flow along the jet to be ~ 1200 km s^-1.

However, there are a number of indications that this picture of a free jet may be oversimplified and not universally applicable.

First, the cone angles of some jets are not constant. For both NGC 315 (Bridle et al. 1979a) and 3C449 (Perley et al. 1979), the jets shown in Figures 11 and 12, the cone angle initially decreases with distance, most of this "collimation" occurring within 30 kpc of the nucleus. This suggests that the beam can be influenced, perhaps even focused, by a circumgalactic medium. A further indication that jets may be affected by the galaxies through which they pass comes from 3C 83.1B/NGC 1265, the best-studied tailed radio galaxy (Miley et al. 1975, Owen et al. 1978a). Its jet, seen in Figure 14, has been observed to flare out about 10 kpc from the nucleus and this abrupt widening has been cited as evidence for the presence of a hot (~ 10⁷ K) dense (~ 10^-2 cm^-3) interstellar medium in the central part of the elliptical galaxy (Owen et al. 1978a, Jones & Owen 1979).

Figure 14. The head of the tailed source 3C83.1B/NGC 1265 (VLA, 5 GHz) (from Owen et al. 1978a). See also Figure 8.

Second, there are differences in the magnetic field configurations observed in various jets. Although in most jets in low luminosity radio galaxies the held is observed to be predominantly perpendicular to the jets, this may not be the case for high luminosity sources. In 3C219, the only high luminosity source whose jet has been studied, the polarization distribution indicates a magnetic field aligned along the jet (Burch 1979b). Similarly, the bridges in most high luminosity sources have predominantly parallel magnetic fields (e.g. Haves & Conway 1975, Miley 1976). It is unclear whether there is a distinction between a "bridge" in an edge-brightened double source and a "jet." The question of the magnetic field structure in jets is therefore complex. The interpretation is especially difficult because polarization position angles give merely the projected magnetic field directions. Fomalont et al. (1980) suggest that the three-dimensional field configuration in 3C31 and NGC 315 may be helical.

Third, the surface brightness of the jet in 3C31 decreases roughly as if compared with r^-3 predicted by simple synchrotron theory (Valtonen 1979, Fomalont et al. 1979, Burch 1979c). Thus, to compensate for adiabatic losses there is likely to be an additional source of energy available to accelerate more particles or to amplify the field. As we have seen, even more stringent evidence for localized acceleration in jets is provided by the detection of optical emission.

Future searches for correlations between the radio and optical properties of jets and other radio-source parameters may help to elucidate the energy transport mechanism further.

The breakdown of a radio source into its elementary component parts has been useful in discussing the various processes involved, but it has inevitably been oversimplified. We have already remarked on the often arbitrary distinction between a "jet" and a "bridge." It is also not always possible to clearly distinguish between a "core" and a "jet." When the cores are observed with sufficient resolution jet-like structures are observed within them, e.g. the steep-spectrum core of Virgo A/M 87 (Figure 10) or the flat-spectrum core of NGC 6251 (Figure 9). The flat-spectrum cores, steep-spectrum cores, and jets are almost certainly all manifestations of different regimes in the energy transport process.