Annu. Rev. Astron. Astrophys. 1988. 36: 539-598 Copyright © 1998 by Annual Reviews. All rights reserved

1. HISTORICAL INTRODUCTION

The evidence for highly collimated jets in astrophysics goes back to the early radio observations of twin lobes in extended radio galaxies, of which the prototype is Cygnus A (Jennison & Das Gupta 1953). After associating them with optical galaxies at cosmological distances, it was clear that they had gigantic dimensions (up to megaparsec scales) and astonishing powers (up to 10⁴⁷ erg s^-1) emitted as nonthermal radio continua of synchrotron type. These facts made a single ejection event from the nucleus of the parent galaxy unlikely and, in general, posed a serious energetic problem (Burbidge 1958). In fact, the stopping distance of a plasmon moving in a constant density environment is D ~ s / , where s is the plasmon scale and the density ratio environment/plasmon. Unless dense plasmons are considered, which would then require high kinetic energy for their ejection (up to 10⁶¹erg), the typical geometry should display D s, in contrast with observations. The same amount of energy delivered continuously on times 10⁷ years by supersonic outflows (with v_s << v_j so that s << D) is a less severe problem (Rees 1971, Scheuer 1974).

In addition, the short synchrotron lifetimes of relativistic electrons do not allow radio emission for more than ~ 10⁶ years unless reacceleration is introduced to the picture, and the situation is obviously much worse for higher frequencies. Again, this phenomenology could be explained more economically in terms of fluid jets continuously transferring energy and momentum from the galactic nuclei into the lobes and maintaining "in situ" particle reacceleration (Blandford & Rees 1974).

Finally, with the increase in sensitivity and angular resolution of radio telescopes, bridges of nonthermal emission were detected connecting galactic nuclei and radio lobes (see e.g. Miley et al 1975, Turland 1975, Bridle & Fomalont 1976, van Breugel & Miley 1977); a complete summary is given by Miley (1980). Although nonthermal continua did not allow Doppler measurements of velocities in these bridges, it was clear that a permanent physical link existed between nuclei and lobes characterized by a surprising collimation.

Very long baseline interferometry (VLBI) observations traced the outflow collimation down to subparsec scales and allowed measurements in several cases of superluminal proper motions (Cohen et al 1971, Moffet et al 1971, Whitney et al 1971). This fact, together with a statistically significant presence of one-sided jets in strong sources, was considered evidence that jets may, at least in some cases, be relativistic.

Eventually jets were discovered to emit also in the optical, X-, and -ray bands, and their relationship with very high-energy phenomena originating in the deep cores of active galactic nuclei (AGNs) was definitively established. In this respect, three recent observational developments must be mentioned:

1. The Hubble Space Telescope (HST) has gathered clear evidence of a closespatial connection between thermal and nonthermal radiation emissions in the central regions of some AGNs; in particular, the nonthermal emission corresponds to the initial part of the jet that appears to compress the external interstellar plasma while ploughing its way out (Capetti et al 1996).
2. The Compton Gamma-Ray Observatory (CGRO) has detected strong and highly variable -ray emission from blazars, suggesting that these objects owe their enormous brightness to relativistically Doppler-boosted radiation from jets pointing toward the Earth (Hartman et al 1992).
3. Monitoring of intraday/intranight variability of blazars supports the idea that beaming of jets with Lorentz factors as high as 10³ can explain their huge energetics and rapid time scales of variability (Witzel 1992); however, coherent radiation mechanisms could partly reduce this request.

Table 1 summarizes the characteristic physical parameters of AGNs and their jets. Figure 1 is a representative collection of sample morphologies.

Figure 1. Collimated jets from active galactic nuclei (AGNs): (a) Cygnus A in early observations; (b) VLA radio map of Cygnus A. (c) Superluminal motions in 3C 345. (d) Hubble Space Telescope (HST) map of Markarian 3 (Capetti et al 1996); and (e) geometry of Doppler beaming in one-sided jets and blazars.

For detailed analyses of existing data on jets, from radio to high frequencies, we refer to the many published reviews, first of all those that appeared in this series by De Young (1976), Miley (1980), Kellermann & Pauliny-Toth (1981), Bridle & Perley (1984). Recent HST optical data are presented by Macchetto (1996) and high-frequency data by Hartman et al (1992). The physical parameters of jets are commonly derived under the assumption they are "optically thin incoherent synchrotron sources." In particular, with an eye to the energy budget, estimates are made in the assumption of minimum energy requirement corresponding to equipartition between relativistic electrons (and protons or positrons) and magnetic fields (Burbidge 1958). For the cores, the optically thin approximation breaks down and other models are used. Further estimates can be done using polarization, depolarization, and Faraday rotation measures. Using these diagnostics, typical average physical parameters of extended radio galaxies can be obtained (Table 2).

Table 2. Radio galaxies

	Core	Jet	Hot spot	Lobe
D (size) (kiloparsecs)	10-3	2-103	5	50-103
Beq (Gauss)		10-3	10-3	10-5
ne,rel (cm-3)		10-2-10-5	10-2	10-4
Polarization (%)	2	0-60	15	0-60
Spectral index ()	0.0	0.6	0.6	0.9
vflow / c	1	10-1	10-3	10-3