Annu. Rev. Astron. Astrophys. 1984. 22: 319-58 Copyright © 1984 by Annual Reviews. All rights reserved

1. INTRODUCTION

Powerful extended extragalactic radio sources pose two vexing astrophysical problems [reviewed in (147) and (157)]. First, from what energy reservoir do they draw their large radio luminosities (as much as 10³⁸ W between 10 MHz and 100 GHz)? Second, how does the active center in the parent galaxy or QSO supply as much as 10⁵⁴ J in relativistic particles and fields to radio "lobes" up to several hundred kiloparsecs outside the optical object? New aperture synthesis arrays (68, 250) and new image-processing algorithms (66, 101, 202, 231) have recently allowed radio imaging at subarcsecond resolution with high sensitivity and high dynamic range; as a result, the complexity of the brighter sources has been revealed clearly for the first time. Many contain radio jets, i.e. narrow radio features between compact central "cores" and more extended "lobe" emission. This review examines the systematic properties of such jets and the clues they give to the physics of energy transfer in extragalactic sources. We do not directly consider the jet production mechanism, which is intimately related to the first problem noted above - for reviews, see (207) and (251).

1.1. Why "Jets"?

Baade & Minkowski (3) first used the term jet in an extragalactic context, describing the train of optical knots extending ~ 20" from the nucleus of M87; the knots resemble a fluid jet breaking into droplets. They suggested that "the jet was formed by ejection from the nucleus" (even though its continuous spectrum gave no clue to its velocity) and that an [O II] 3727 emission line in the nucleus whose centroid is blueshifted by several hundred kilometers per second from the systemic velocity is "emitted by a part of the material which forms the jet and is still very close to, if not still inside, the nucleus." The "optical wisp" (227) near the QSR 3C 273, which resembles the M87 knots, was also called a jet without direct evidence for outflow. Radio detection of these optical "jets" (114, 160) prompted description of narrow features in other 3C sources (168, 253) as "radio jets." In 1973, refined versions of the continuous outflow, or "beam," models for extragalactic sources proposed earlier by Morrison (162) and Rees (204) were developed (25, 147, 221). The new models (a) obviated adiabatic losses which led "explosive" models to require that the compact precursors of extended sources be far more luminous than any actually observed; and (b) they explained how the synchrotron lifetimes of electrons in bright lobe "hot spots" can be less than the light travel time to the hot spots from the parent galaxy or QSR (e.g. 113). Continuous flow models and jet data have kept close company ever since.

1.1.1   CAVEAT   The term jet connotes continuous outflow of fluid from a collimator, but there is no direct evidence for flow in any continuous extragalactic "jet." VLBI studies of proper motions of knots in some compact radio sources - reviewed in (57) - suggest outflow of jetlike features from stationary "cores," but only in 3C 345 (7) has this been tested in an external reference frame. Such proper motions cannot be monitored in truly continuous emission. Narrow kiloparsec-scale features are therefore called jets mainly because they occur where "beam" models required collimated outflow from active nuclei.

1.2. What Makes a "Narrow Feature" a "Jet"?

Terminology that so prejudges source physics should be used sparingly, so we require (as in 27) that to be termed a jet, a feature must be

1. At least four times as long as it is wide,

2. Separable at high resolution from other extended structures (if any), either by brightness contrast or spatially (e.g. it should be a narrow ridge running through more diffuse emission, or a narrow feature in the inner part of a source entering more extended emission in the outer part),

3. Aligned with the compact radio core where it is closest to it.

1.2.1   EXAMPLES   Figures 1 to 5 show examples of jets of various powers and sizes. They also illustrate some ambiguities - with less sensitivity, the NGC 6251 jet (Figure 2) breaks into discrete knots, not all of which are elongated along it. We call a train of knots a jet, however, only if it has more than two knots or if some knots are elongated along it (e.g. Figures 3 and 5). (We prefer to exclude some blobby jets temporarily than to apply the prejudicial term jet too liberally.) The elongated outer lobes of some edge-darkened sources, e.g. 3C 31 (245), may equally plausibly be termed broad jets (87), so dividing them into "jet" and "lobe" segments by morphology alone may be subjective. We ask that they contain a "spine" of bright emission meeting criterion (2.) before we call them jets.

Figure 1. VLA maps of the jets in the weak radio galaxy M84 at 4.9 GHz, with the right panel showing detail of the central region. The peak on these maps is the radio core. Note the one-sided bright base of the northern jet and the faint cocoon of emission flaring from both jets beyond 5" from the core (data of R.A. Laing and A.H. Bridle, in preparation).

Figure 2. The structure of the jet and counter jet in NGC 6251 over a wide range of angular scales. Note the knotty substructure of the jet, and the large-scale "wiggle" (middle panel). The nuclear "core-jet" (bottom panel) and the mean position angle of the larger-scale jet are misaligned by 4.5 ± 1° [data from (279) - top panel; R.A. Perley and A. H. Bridle (in preparation) - second and third panels; (183) - fourth panel; and (56) - bottom panel].

Figure 3. VLA maps of two one-sided jets in extended QSRs at 4.9 GHz, adapted from (173). Note the knotty structure of the jet in the left panel and the "gap" and the "wiggle" in the jet in the right panel. The peak on each map is the radio core.

Figure 4. The structure of the core and jet in 3C 120 over a wide range of angular scales. The three brightest features in the lower-left panel exhibit superluminal expansion (268). Note the continuity between the different scales and the large misalignment between the smallest and largest structures (left panels). Montage provided by Drs. R.C. Walker and J.M. Benson.

Figure 5. MERLIN/EVN map of the one-sided jet in the strongly core-dominated QSR 3C 418, provided by Dr. T.B. Muxlow. This is the most powerful core known to be associated with a jet. Note the sharp curvature of the jet near the core and its knotty Structure. The tic marks are 0.36" apart.