Bilateral symmetry in active galaxies

Physics Reports 303 (1998) 81—182
Bilateral symmetry in active galaxies
D.G. Banhatti!,"
! School of Physics, Madurai-Kamaraj University, Madurai 625 021, India1
" Astronomisches Institut Mu( nster, Wilhelm-Klemm-Stra}e 10, D-48149 Mu( nster, Germany
Received November 1997; editor: D.N. Schramm
Contents
1. Introduction 84
1.1. Active galaxies defined 84
1.2. Discovery of bilateral structure 85
1.3. Spectral classes 86
1.4. Radiation mechanisms 87
1.5. Quasars 87
1.6. Active galaxies and their nuclei 88
1.7. This review 90
2. Structures and symmetry parameters 91
2.1. Extragalactic jets 92
2.2. Straight doubles 94
2.3. Bent doubles 99
2.4. Extended optical and X-ray
structure 100
2.5. Intermediate scale structure 100
2.6. Compact structure 102
3. Models 104
3.1. Two kinds of models and some
recent developments 105
3.2. Beam models — Details 107
3.3. Numerical simulation of jets 118
3.4. Deductions from bilateral asymmetry
of straight doubles 119
3.5. Summing up 127
4. Statistics of structures 128
4.1. Extragalactic jets 129
4.2. Straight doubles 137
4.3. Bent doubles 143
4.4. Extended optical and X-ray structure 146
4.5. Intermediate scale structure 149
4.6. Compact structure 153
5. Cosmological evolution of structures 162
5.1. Background 162
5.2. Luminosity and number density evolution 163
5.3. Linear size evolution and distortion 164
5.4. The k—z correlation for superluminals 166
5.5. Summary 166
6. Concluding remarks 166
6.1. Analogy with stellar astrophysics 167
6.2. Present status of observations and
theory 168
References 169
1Present address.
0370-1573/98/$29.00 ( 1998 Elsevier Science B.V. All rights reserved
PII S 0 3 7 0 - 1 5 7 3 ( 9 7 ) 0 0 0 9 1 - 4

BILATERAL SYMMETRY IN ACTIVE
GALAXIES
D.G. BANHATTI!,"
! School of Physics, Madurai-Kamaraj University, Madurai 625 021, India
" Astronomisches Institut Mu( nster, Wilhelm-Klemm-Stra}e 10, D-48149 Mu( nster, Germany
AMSTERDAM — LAUSANNE — NEW YORK — OXFORD — SHANNON — TOKYO

Abstract
The bilateral symmetry around active galaxies was discovered in the radio continuum in the 1950s. It has since been
found to manifest itself throughout the electromagnetic spectrum in the continuum as well as lines, on scales from
subparsec to megaparsecs. Detailed observations have often led to classification schemes based on different length- and
time-scales, and types and stages of activity. Consolidation of continuum and line observations in total and polarized
intensities in various bands of the electromagnetic spectrum from c-rays to radio should lead to a more complete picture,
unifying these schemes into a physical whole. We review these observations of active galaxies and their current models,
and how well the two fit together. Our emphasis throughout is on bilateral structure in and around active gala-
xies. ( 1998 Elsevier Science B.V. All rights reserved.
PACS: 98.54.!h
D.G. Banhatti / Physics Reports 303 (1998) 81—182 83

„o Aai and Baba
The study of nebular forms, as they appear on direct photographs, leads to the
conclusion that nebulae are closely related members of a single family. They are
constructed on a fundamental pattern which varies systematically through a limited
range. The nebulae fall naturally into an ordered sequence of structural forms and are
readily reduced to a standard position in the sequence. At such a standard stage, the
relation between apparent size and brightness is just that to be expected if the same
nebula could be examined from various distances. The dispersion in the apparent
characteristics of nebulae is remarkably small.
E. Hubble (1936)
1. Introduction
Various aspects of active galaxies have been reviewed from time to time. Miley (1980) has
reviewed the structure of extended extragalactic radio sources, Kellermann and Pauliny-Toth
(1981) and Zensus and Kellermann (1994) the structure of compact radio sources, while Asse´o and
Sol (1987) and Saikia and Salter (1988) have reviewed, from somewhat different points of view, the
polarization properties of active galaxies. Wiita (1985) and a special issue of the Publications of the
Astronomical Society of the Pacific with Introduction by Jones (1986) have summarized the
observations of active galaxies and their fundamental interpretations. Trimble and Woltjer (1986)
succintly review quasars on their silver jubilee. Recent summaries of the subject include three
articles based on a series of lectures covering different aspects (Courvoisier and Mayor, 1990).
Frank et al. (1985) and Treves et al. (1988) review the basic elements of the theory of accretion.
Other reviews and updates on different aspects of active galaxies are mentioned at appropriate
places. In this review, the emphasis is, however, on the bilateral structure in active galaxies.
This introductory section gives a brief background on active galaxies and the mechanisms that
give rise to the observed structure. Further, the physical conditions in different regions of a typical
active galaxy are described — from the nucleus to the very extended structure. It is then possible to
state the focus of the review, and list the parameters specifying bilateral symmetry.
1.1. Active galaxies defined
About a decade before activity was discovered in galaxies, Hubble (1936) classified them as
elliptical, lenticular, spiral and irregular from their appearance on optical photographs, and found
that the surface brightness at a given position in the tuning fork sequence (Fig. 1) is the same and
decreases systematically from the circular form to the loose spiral. Later, the presence or absence of
peculiarities on optical photographs and in spectra led to another classification of galaxies as active
or normal. A galaxy is called active if it shows nonthermal emission associated with (and possibly at)
its nucleus and/or emission not attributable to normal stellar processes ( from star formation to
supernovae) or to a hot gaseous disk or halo. However, it is found that active nuclei reside in all the
Hubble types: LINERs (low ionization nuclear emission line regions) and Seyfert nuclei live almost
always in spirals, quasars in spirals, lenticulars and ellipticals, and radio galaxies and BL Lac
84 D.G. Banhatti / Physics Reports 303 (1998) 81—182

Fig. 1. Tuning fork sequence of Hubble galaxy types (from Hubble 1936, 1958, with permission from Yale University
Press).
objects almost always in ellipticals and irregulars. Osterbrock, who has been active for decades in
this area of research, observing as well as theorizing, has reviewed active galactic nuclei at many
phases of the activity (e.g., Osterbrock, 1984, 1989, 1991). Interestingly, there seems little agreement
among experts on what is a peculiar galaxy (Naim and Lahav, 1997), although an active galaxy is
a well-defined object, as per the definition above.
1.2. Discovery of bilateral structure
After the birth of radio astronomy in the 1930s with Jansky’s detection of nonthermal radio
continuum at j 14.6m (20.5MHz) from the centre of our Milky Way Galaxy, an Sbc spiral, many
discrete radio sources were found. Among the 11 brightest of these, identified with already familiar
objects in the sky (other than the Sun and Jupiter), five are galaxies (Cyg A, Vir A (M 87), Cen A
(NGC 5128), Andromeda Nebula (M 31) and Sgr A (Galactic Centre)), four are supernova
remnants (including Tau A (Crab Nebula)) and two H` (i.e., H II) regions. In the 1950s, Hanbury
Brown, Jennison and DasGupta at Jodrell Bank found Cyg A to be elongated along position angle
(PA) 113°. Combining their measurements of its structure at Cambridge and other similar
observations at Sydney near j&0.3m (&0.1GHz), they showed Cyg A to be a double radio
source of size &1.5@ with two components of size &1@]0.5@. These components straddle with little
overlap an optical galaxy of maximum size 0.5@ in PA 150°, of disturbed appearance and with a high
excitation emission spectrum (see review by Sullivan, 1982). This was the first detection of
a bilaterally symmetric structure associated with an active galaxy (Jennison and DasGupta, 1953).
Today it is known that such a structure is ubiquitious among many active galaxies. Various
components of this structure can be conveniently labelled core, jets, knots, hotspots, lobes, bridges,
filaments, etc., as observed mainly in the radio band, and broad-line region (BLR), intermediate
fluorescence zone (IFZ), nuclear narrow-line region (NNLR) and extended narrow-line region
(ENLR) as observed mainly through optical spectra. The IFZ is also called NFR, for narrow-line
fluorescence region, while ENLR is sometimes called EELR, for extended emission line region.
Although the two arms of a two-armed spiral galaxy also form a bilaterally symmetric structure,
it is not of interest to us as the emission from the stars and gas which make up the two arms in the
D.G. Banhatti / Physics Reports 303 (1998) 81—182 85

disk is considered normal and does not brand the galaxy active. We are, however, interested
in some of these spirals with Seyfert nuclei, especially when oriented edge-on, as they show
bilateral structure nearly perpendicular to the disk, in addition to the disk spiral (e.g., Colbert et al.,
1996a,b).
1.3. Spectral classes
Early radio surveys were done at frequencies less than about a GHz (jZ30 cm) and hence
predominantly showed up extended (Z10A to a few degrees) radio sources with steep radio
continuum spectra (i.e., a slope of about !0.5 or steeper on a log—log plot of flux density versus
frequency). High frequency surveys mainly found compact ([1A) flat spectrum radio sources,
and also some steep spectrum ones of sizes of a few arcsec. Most extended and compact
steep spectrum radio sources have a double structure with a high degree of bilateral symmetry.
Extended radio sources also have a flat spectrum core coincident with the optical galaxy,
quasar or BL Lac object, and may have a small scale core jet structure when observed at
finer resolutions of milliarcsec. This jet often continues to tens of arcsec, sometimes with a
continuous change of direction, with no jet or at most a weak one on the other side, called the
counterjet side. However, most compact flat spectrum sources show a highly asymmetric core-jet
structure at all resolutions. The early emphasis on extended steep spectrum and compact flat
spectrum radio sources resulted in the neglect of compact steep spectrum radio sources, which were
later found to have two types of structure: a compact flat spectrum core or core-jet plus a diffuse
steep spectrum component on the jet side, and nearly equal compact components of steep radio
spectra with a possible weak compact flat spectrum component coincident with the AGN. This
last subclass as yet has only a handful of members (e.g., Wilkinson et al., 1994). See Table 1 for a
summary.
Table 1
Spectral classes of radio sources
Spectral class Best observation Angular size Structure Remarks
frequency (GHz)
Extended (1 10A to a few degrees Double/triple with Includes straight and
steep spectrum varying degrees of
bilateral symmetry
bent doubles, edge-brightened
as well as edge-darkened
structures
Compact &1 1A—10A Core-jet#diffuse Core-jet may show
steep spectrum OR superluminal proper motion.
Compact double/triple Compact doubles are edge-
brightened
Compact ’1 (1A Asymmetric Often shows apparent
flat spectrum (often one-sided)
core-jet
superluminal motion
86 D.G. Banhatti / Physics Reports 303 (1998) 81—182

1.4. Radiation mechanisms
Early radio observations of the Crab Nebula (Tau A) showed a continuum of steep spectrum
and significant polarization for the first time in a celestial source. This was explained as
due to synchrotron and associated radiation processes, i.e., radiation from relativistic charged
particles gyrating in a magnetic field. Extragalactic radio sources have similar spectrum
and polarization and their radio emission is also thought to be synchrotron radiation from
relativistic electrons with a power law energy spectrum of slope about !2.7 (i.e., the number N
E
of electrons of energy E being N
E
JE~2.7), around energies of a few hundred GeV gyrating
in magnetic fields of a few to hundreds of lG, modified by synchrotron self absorption,
adiabatic expansion, inverse Compton and induced Compton scattering, diamagnetic effects, etc.
(Alfve´n and Herlofson, 1950 (reproduced in Sullivan, 1982); Shklovskii, 1963; Singal, 1986a,
Singal, 1986b; Blandford, 1990; Lieu and Axford, 1993). This nonthermal continuum has
been found to extend from radio to X-rays (e.g., Woltjer, 1990; Courvoisier and Blecha, 1994).
Repeated acceleration through shocks naturally gives rise to the required power law energy
spectrum, although there is still doubt if that is the correct form to model (e.g., Rudnick and
Katz-Stone, 1996, and references therein; Salas et al., 1995; Meisenheimer et al., 1996). The
radiation mechanisms for the nonthermal continuum seen in active galaxies have thus yet to be
firmly established, and there are attempts to find alternatives (e.g., Krishan and Wiita, 1990; Yoon
and Ziebell, 1996).
1.5. Quasars
3C 273 was the first radio source to be identified with an optical object of stellar appearance. It is
also unique in having an optical jet, similar to the giant elliptical M 87 (Vir A). (Continuum optical
emission has been detected to date in over a dozen radio jets.) Many other radio sources were later
identified with stellar objects. These quasi-stellar radio sources (quasars) are the most powerful
type of active galaxies and have highly redshifted emission spectra.
Those of redshifts Z2 have an ultraviolet excess compared to stars. Similar quasi-stellar objects
(QSOs) were found on the basis of UV excess without any (or at most faint) radio emission, and
soon overwhelmed in number the quasi-stellar radio sources. Today all these are called quasars,
and are known to be luminous active nuclei of distant galaxies (upto redshifts of 5), occasionally
radio-loud, but mostly very faint in the radio. (Ve´ron-Cetty and Ve´ron, 1996, list some 8600
quasars, &20% being radio loud above 2.4mJy at 1.4GHz (Bischoff and Becker, 1997).) Many are
X-ray sources, and some have been discovered in the X-ray band, especially with the space-borne
Einstein X-ray observatory. The InfraRed Astronomical Satellite (IRAS) has also discovered
quasars and measured known ones in the infrared band. (See Trimble and Woltjer (1986) for
a review on quasars.) Quasars having extended radio structure are difficult to distinguish from
radio galaxies on the basis of radio properties alone, though more of them have asymmetric
radio structures and brighter compact flat spectrum cores than do radio galaxies, and they are also
more radio luminous on average. Further, quasars with extended radio structures have a smaller
mean physical size than comparable radio galaxies. These differences and similarities among the
radio loud quasars and radio galaxies have prompted models unifying these two classes of AGNi.
In these models, quasars have their radio axes closer to the observer’s line of sight than do radio
D.G. Banhatti / Physics Reports 303 (1998) 81—182 87

galaxies. (See Hutchings, 1987; Price et al., 1993; Hutchings et al., 1994; Neff et al., 1995; Hes et al.,
1996; for studies of matched samples of radio galaxies and radio quasar galaxies.)
1.6. Active galaxies and their nuclei
1.6.1. Active galactic nuclei (AGNi)
The optical spectra of normal galaxies show stellar light with absorption lines, while active
galaxies show emission lines with an underlying nonthermal continuum which overwhelms the
starlight. The weakest of these are LINER nuclei, then come Seyfert 2’s, Seyfert 1’s and quasars. BL
Lac objects are brighter than Seyfert 1’s, are much more variable on time-scales from years down to
days and even hours, and show a featureless spectrum most of the time. The comparatively weak
emission lines shine through the brighter continuum only occasionally.
The variability of the nonthermal continuum of active nuclei from radio to X-ray on time-
scales of years to days indicates that the powerhouse is only light-days in extent. When variability
on the scale of hours is seen, one infers structure on the scale of light-hours (& tens of
AU). Variability is also seen in emission lines, which can be considered to be lit up by
the underlying continuum. In general, the line emission from an active galaxy comes from
four regions characterized by progressively different values of the four physical parameters:
the linewidth to zero intensity W, the electron density n, the filling factor f and the ionization
parameter IP (i.e., the number of photons per ionizable nucleon). All the regions are inferred to
have the same electron temperature„&104 K. Table 2 gives the values of these parameters for the
four regions.
Table 2
Physical conditions in an AGN
Zone dr r … n „ f IP
BLR or BELR subpc a few 0.1 pc 5—30 Z109 104 &0.01 &0.01
NFR or IFZ a few pc a few pc 1—5 106—109 104 (0.01 &0.01
NNLR or
NLR or NELR
&kpc a few 10pc to
&kpc
0.2—1 102—106 104 ;0.01 &0.001 for a
LINER;
&0.01—0.1 for a
Sy 2
ENLR or EELR &10 kpc to
a few 100kpc
10 s to 100 s
of kpc
0.2—1 0.1—10 104 &0.01 2
Notes: Zone"Emission line zone; dr"size of zone; r"distance of zone from AGN engine; …"line width at zero
intensity (in 103km s~1); n"electron density (in cm~3); „"electron temperature (in K); f"filling factor; IP"ioniz-
ation parameter; BLR or BELR"broad (emission) line region; NFR"nuclear fluorescence region; IFZ"intermediate
fluorescence zone; NNLR or NLR or NELR"(nuclear) narrow (emission) line region; ENLR"extended narrow line
region; EELR"extended emission line region.
88 D.G. Banhatti / Physics Reports 303 (1998) 81—182

The continuum emission from infrared to X-rays underlies and excites the line emission in the
BLR, the NFR, the NNLR and the ENLR, while the radio continuum from relativistic plasma in
a magnetic field is superimposed on all these scales, in varying degrees depending on the type of
active galaxy, its orientation and motion, and the stage in its activity.
1.6.2. Models
In the most popular class of models, viz., beam models (e.g., Begelman et al., 1984), proposed to
broadly account for these observations, the powerhouse or central engine is thought to be
a supermassive black hole accretion disk system at the nucleus, which emits highly relativistic
supersonic plasma beams along the rotation axis of an optically and geometrically thick torus.
These beams are observed as jets, mainly in the radio continuum, while the surface of the accretion
disk emits the infrared to X-ray continuum radiation. In addition, the disk may emit a bipolar
wind, wider and slower than the jets and forming a sheath around them. In the beam models, the
type of structure seen depends on (1) the accretion rate relative to the Eddington limit for the
particular black hole mass (&108 M
_
; equivalent density &1023.5 cm~3, i.e., &1 g cm~3), (2) the
type of galaxy in which the system resides, (3) companion galaxies, (4) cluster membership, (5) the
stage of activity of the galactic nucleus, (6) the relative orientations of the different rotation
axes/disk planes and our line of sight to the nucleus, and (7) the motion of the parent galaxy relative
to the ambient medium.
The other class of models (viz., slingshot models (e.g., Saslaw et al., 1974, Valtonen, 1984)) is yet
to be explored in as much detail. In slingshot models, massive compact objects, possibly black holes
with accretion disks, are ejected violently as slingshots of close binaries and singles from the nuclear
region of a galaxy, and the observed structure is due to the emission from their trails, as well as
from the vicinity of the compact objects themselves.
It may well turn out that the successful model has ingredients from both the slingshot and beam
theories. Our discussion in this review, even when not directly addressed to matters concerning
models, will assume the beam models as the backdrop, as does much of literature on AGs to date.
At appropriate places, we will also mention interpretations of observations within the slingshot
models.
1.6.3. Morphology and bilateral symmetry
The radio jets originate on the sub-pc scale and may continue to shine to Mpc scales for the most
powerful extended radio sources associated with giant elliptical galaxies. Precession of the central
engine can induce inversion symmetry, while motion of the galaxy relative to the intergalactic
medium produces mirror symmetry in the bilateral structure, most often broken due to projection
on the sky plane. Inversion and mirror symmetries can also be the result of backflows of jet
material from the brighter regions called hotspots toward the core in the form of extended low
brightness lobes and bridges. The less powerful jets form smaller bilateral radio sources in the
interstellar media of Seyfert galaxies in their disks or perpendicular to them. These jets may be seen
along with the obscuring torus, observed in atomic and molecular radio emission lines, including
maser lines, in the NFR. The torus appears nearly edge-on in Seyfert 2s and nearly face-on in
Seyfert 1s.
The sub-pc jets, mapped by Very Long Baseline Interferometry (VLBI) with milliarcsec resolu-
tions, are most often one-sided, and the counterjet, even if present, is very weak. The hotspots in
D.G. Banhatti / Physics Reports 303 (1998) 81—182 89

powerful (’1024.5WHz~1 at 1400 MHz, with H
0
"75 km s~1Mpc~1, q
0
"1/2) straight doubles
are situated at their outer edges, earning them the epithet edge-brightened. For these straight
edge-brightened powerful doubles, the jet side has lower depolarization compared to the counterjet
(or weaker jet) side, consistent with the jet side being nearer and hence having less Faraday depth
than the counterjet side. Most low luminosity doubles ( 1024.5WHz~1) have the brightest
regions nearest the nucleus (edge-darkened), and may have distorted structures due to bending
of jets.
This dichotomy into edge-brightened and edge-darkened structures at a fairly well-defined radio
luminosity for double radio sources (Fanaroff and Riley, 1974), which was later found to be
significant in more than one way, forms a cornerstone in the phenomenology of extragalactic
extended radio sources, and has been fruitfully applied beyond its original intent. Owen (1993)
summarizes studies of AGs across this radio luminosity break with respect to other properties like
optical luminosity and radio linear size. The details are given by Ledlow and Owen (1996) and
references therein. Baum et al. (1995) (and references therein) complement these studies by
considering line emission as well. De Young (1993) and Bicknell (1995) have modelled, from
somewhat different points of view, the division in the radio vs optical luminosity plane. Bowman
et al. (1996) simulate, based on Komissarov’s (1994) work, hot and cold jets to form edge darkened
and edge brightened doubles, respectively. Meier et al. (1997) find, through time dependent
magnetohydrodynamic simulations, that magnetic forces dominate over gravity in edge brightened
structures which are fed by highly relativistic jets, while edge darkened structures are fed by mildly
relativistic jets dominated by gravity.
The ENLR also has a bilateral structure, at the distance of tens to hundreds of kpc, though
broader and less sharply defined than the radio morphology described above.
1.6.4. Symmetry parameters
It is then evident that extragalactic radio sources have varying degrees of bilateral symmetry in
their structure. The bilateral asymmetry can be due to orientation of the jet axis close to the line of
sight and the consequent Doppler effect and different light travel times to the observer from
different points on the axis, plus effects due to intrinsic asymmetries in the environment. In the
slingshot models, the ejection process is, in general, asymmetric and the ejected masses are not
always equal; further asymmetries are introduced by the possibly asymmetric environment. The
bilateral (a)symmetry of and around AGs with a more or less linear structure can be usefully
quantified by (i) the ratios of armlengths, hotspot strengths and total component strengths on the
two sides, (ii) the relative strengths of the radio core, the hotspots, and the total lobe emission,
(iii) the misalignment of the two arms from a straight line, (iv) the misalignment between the axes of
the bilateral radio structure on nuclear and larger scales, (v) the collimation of the observed
outlying structure as quantified by the angle it makes at the core and (vi) the angles giving the
orientations of the optical image, the dust lane, the rotation axis and the optical and radio
polarization axes.
1.7. This review
This brief introduction tells that a physical attribute common to all types of active galaxies is
their bilateral structure and the related asymmetry, specified by the symmetry parameters. When
90 D.G. Banhatti / Physics Reports 303 (1998) 81—182

studies on interpretation of the observed bilateral structure of extragalactic sources were first taken
up, the data available were sparse. These have grown in volume and quality today, but are scattered
in the literature. The main purpose of the review is to bring together all aspects of study on this
topic and give a coherent summary. As stated above, beam models relate the bilateral structure of
an active galaxy to its nuclear activity, and seem to account for the broad observed features.
However, there are two unsatisfactory aspects which we highlight. Firstly, bilateral structure in
certain classes of sources, e.g., an active galaxy with a large linear size and yet highly active nucleus,
are difficult to fit in with beam models. Secondly, the eminently possible task of accounting for the
finer details of bilateral structure in those cases where beam models are found adequate has not
been taken up in a holistic manner. The less popular slingshot models are yet to be explored in very
much detail. The emerging picture is then that, as of today, the observations are still a step ahead of
theory.
The layout of the review is as follows. In Section 2 are described the different types of stucture
associated with active galaxies, mainly in the radio continuum band, and some symmetry para-
meters and properties are defined which systematize these structures qualitatively as well as
quantitatively. The section also includes some detail on jets, straight doubles, bent doubles,
extended optical (and X-ray) structure, intermediate scale structure and compact (radio) structure.
Section 3 presents and critically examines the models which have been constructed to account for
all these structures as manifestations of the different stages in the activity of variously oriented
active nuclei of several types residing in different kinds of parent galaxies, the galaxies possibly
moving relative to the ambient medium. Section 4 then returns to the relative incidence of these
structures and the statistics of the symmetry parameters and interrelations between them in the
light of these models. Finally, Section 5 briefly considers changes in the structure of active galaxies
of higher and higher redshifts and presents some relevant aspects of the cosmological evolution of
active galaxies. The review closes with some analogies in the final Section 6.
2. Structures and symmetry parameters
As mentioned earlier, bilateral structures in and around active galaxies, as seen down to subpc
resolution in radio continuum maps, may be conveniently labelled cores, jets, knots, hotspots,
lobes, bridges, filaments, etc., forming straight and bent doubles.
The subpc scale jets start near the core and, as seen by optical (i.e., infrared to ultraviolet)
spectrophotopolarimetry, pass successively through (i) the subpc-sized BLR situated at upto a pc
from the core, (ii) the intermediate nuclear fluorescence region (NFR) of a few pc size at a few pc
from the core and (iii) the kpc-sized NNLR at tens of pc (see Table 2). From subpc to kpc, the jets
are more often highly asymmetric (almost one-sided) and may show sharp bends. As they pass
through the ENLR tens to hundreds of kpc in size, as determined by long-slit optical spectroscopy,
they may bend gently, and are more symmetric unless they are one-sided. Finally, the jets either
terminate in bright, possibly multiple, hotspots or diffuse away in less and less bright knots forming
low-brightness lobes and sometimes filaments within lobes. In most cases, the hotspots are
accompanied by low-brightness bridges extending toward the core, formed due to the jet-material
flowing around and back after passing through the hotspots, identified with jet — ambient medium
working surface.
D.G. Banhatti / Physics Reports 303 (1998) 81—182 91

Some active galaxies show bilateral radio structure within the interstellar medium, like the
ellipticals showing kpc-sized doubles or asymmetric one-sided core-jets, and the Seyferts (which are
almost always spirals) showing fairly symmetric doubles. The innermost subpc to pc scale core-jet
structure, as determined in the radio continuum by VLBI, may show knots moving away on
timescales of months to years from the cores in quasars and radio galaxies, ostensibly at super-
luminal speeds. Optical spectrometry also shows absorption lines (narrower than in NNLR)
indicating circumgalactic moving clouds observed against the predominantly ultraviolet con-
tinuum radiation from the core. The different emission lines are also superposed on this ionizing
continuum. Most recently, Miyoshi et al. (1995) used VLBI to observe water maser lines from the
inner part (tenth of a parsec) of the active edge-on spiral galaxy (classified as a Seyfert 2) NGC 4258
(,M 106) (see also Greenhill et al., 1995a,b), while Wilkes et al. (1995) detected the hidden nucleus
as a compact polarized continuum source (F
l
Jl~1.1) with broad (&103 kms~1) emission lines,
strongly linearly polarized at PA 85°$2°, coincident with the maser disk plane. Conway and
Blanco (1995) detect neutral hydrogen absorption toward the nucleus of the edge brightened
double Cyg A, perhaps the most extensively studied radio galaxy (e.g., Carilli and Barthel, 1996),
indicative of an atomic obscuring torus around the Cyg A central engine. Gallimore et al. (1997) get
a direct radio continuum image of the ionized obscuring disk surrounding the AGN in the Seyfert
2 NGC 1068 using sensitive 8.4 GHz milliarcsec resolution VLBI observations.
These bilateral structures on different scales are described below using qualitative and quantitat-
ive bilateral symmetry parameters, as well as some other related ones.
2.1. Extragalactic jets
Extragalactic jets (also called galaxian or galactic jets in contrast to stellar jets) extend from
active galactic nuclei (AGNi) to distances of pc to Mpc. The first jet was detected in the optical
continuum in the giant elliptical M 87 (,NGC 4486,Virgo A,3C 274) about eight decades
ago (Curtis, 1918). This and other bright jets have since been extensively mapped in the continuum
from radio to X-rays. The 6th extragalactic jet seen in the optical is in the AG 3C 120, similar to
that in PKS 0521!36 (Hjorth et al., 1995). Over a dozen extragalactic jets have been detected
optically to date. The spectrum and polarization clearly show the emission mechanism to be
synchrotron radiation (although detailed multifrequency fine resolution observations of a few
objects have recently led to a reexamination of this question, and the issue is not yet satisfactorily
settled (e.g., Rudnick and Katz-Stone, 1996, and references therein; Salas et al., 1995; Meisenheimer
et al., 1996)). Most other extragalactic jets have been mapped mainly in the radio continuum and
number about five hundred.
2.1.1. Structure of jet/counterjet
A jet may be described by the run along its length of various properties like width, intensity,
polarization, etc. and other derived ones like minimum pressure, equipartition magnetic field, etc.
As deduced from polarization observations, the magnetic field is along the jet near the core, and
becomes normal to it further on, with a small transition region where it is parallel to the jet outside,
and normal to it near the axis. This transition region is farther away from the core the more
luminous the core or the source containing the jet is. Geometrically, opening angle, distance
between knots, wavelength of wiggles, angles of bend, etc. describe the jet. Jets most often occur in
92 D.G. Banhatti / Physics Reports 303 (1998) 81—182

pairs. The counterjet in the diametrically opposite direction varies from being as intense as the jet
to being a few hundred times weaker, the limitation to its detection coming from the dynamic range
of the map. This jet/counterjet ratio (or comparison) of intensities (or other quantities) at corres-
ponding points on the two sides may vary with distance from the core. Thus, jets are often almost
one-sided on the pc scale, becoming more symmetrical further on.
2.1.2. Recent observations of jets
For multiwaveband results on three of the best observed almost one-sided jets, see Perez-
Fournon et al. (1988) and Schlo¨telburg et al. (1988) and references therein for the giant elliptical
M 87 (,NGC 4486,3C 274) at the centre of the Virgo cluster of galaxies, Fraix-Burnet and
Nieto (1988) and references therein for the quasar 3C 273,1226#023(1950.0) and Fraix-Burnet
et al. (1989) and references therein and Hardcastle et al. (1996) and references therein (who map the
counterjet also) for the elliptical 3C 66B. See also Macchetto (1992) for Hubble Space Telescope
(HST) observations of these and other jets and Biretta and Meisenheimer (1993) and Conway et al.
(1993) and references therein for reviews on Virgo A (,M 87).
The jet direction in M 87 changes by 14°$4° from small scale (&4 milliarcsec or 0.3 pc) to large
(&25milliarcsec or 2 pc) (Junor and Biretta, 1995). Biretta et al. (1995) use 15 GHz VLA
observations of the M 87 jet from 1982 to 1993 to determine outward proper motions of &0.5c in
5 of the brightest knots, while small features in the knot nearest the nucleus show superluminal
motion of (2.5$0.3)c. All these are consistent with a kinematic model in which the jet is at 43° to
the line of sight and the bulk flow Lorentz factor is Z3 over the first kiloparsec, followed by
deceleration in the knots. Despringre et al. (1996) report on j3.4mm (89 GHz) few arcsec resolution
interferometry of the jet and nucleus of M 87.
Ro¨ser et al. (1996) (and references therein) present a study of the 3C 273 jet, combining optical
and radio structures and polarizations, and show that a new optically quiet component of steep
radio spectrum may be identified with the backflow, in numerical simulations, of low density, high
Mach number jets, where Doppler beaming is absent.
Falcke et al. (1996) present HST images of the helical two sided optical jet in the NLR of the
Seyfert 2 ESO 428!G14 (,0714!2914,MCG-05-18-002), wrapping around the radio jet on
one side. For representative studies of jets mapped mainly in the radio continuum, see Perley et al.
(1984a) for 1919#405(1950.0),Cyg A, Bridle et al. (1986) and Perley et al. (1993) for 3C
219,0917#515(1950.0), Kronberg (1986), Leahy and Perley (1991) and Akujor et al. (1994) for
3C 303,1441#522(1950.0) and Pedlar et al. (1983, 1990) for NGC 1275,Per A,3C 84. Carilli
and Barthel (1996) review Cyg A in the context of the beam model. Knopp and Chambers (1997)
discover a bipolar emission line nebula through deep imaging of the z"2.5 radio source 4C 23.56,
inferring a dusty galaxy illuminated by a biconical AGN beam nearly in the sky plane.
2.1.3. Jet speeds
Speeds of extragalactic jets cannot be measured directly since they emit only nonthermal
continuum, with no emission lines. But speeds may be inferred from various arguments (see
Section 4.1.8). On the pc scale the jets are probably highly relativistic (Lorentz factors 5—10),
slowing down to tenths of the speed of light c on kpc to Mpc scales for jets associated with
edge-brightened powerful double radio sources (Meisenheimer et al., 1989) and hundredths (or less)
times c for those associated with edge-darkened less powerful doubles.
D.G. Banhatti / Physics Reports 303 (1998) 81—182 93

2.1.4. Jets on smaller scales
Astrophysical jets are also seen on scales smaller than galaxies. The radio lobe over the Galactic
centre (Sofue and Handa, 1984) and a similar lobe breaking out from the northern part of the Crab
nebula (e.g., Gull and Fesen, 1982) are rather extreme examples, more plumes than jets. Sofue (1986)
plots the expansion rate (,jet width/length from core) of the Galactic Centre Lobe (GCL), a few
Seyferts, and radio galaxies and quasars against the logarithm of their core power to find the
expansion rate decreasing from &1 to near zero, where it saturates at a small value &0.03. The
corresponding opening angles are (45° for the GCL to a few degrees for the most powerful
quasars. This (cor)relation holds over about 8 decades of the core power.
Some young stars (called T Tauri stars) have bipolar flows of rather wide opening angles
enclosing narrow bilateral jets moving at upto &102kms~1 and exciting Herbig—Haro objects
(which are the sites of interaction between the stellar wind and the circumstellar/interstellar
medium) parsecs away on two diametrically opposite sides from the young star (Lada, 1985;
Bu¨hrke et al., 1988, Poetzel et al., 1989 and references therein). Symbiotic binary stars (i.e., close
binaries with mass transfer) also show split spectral lines which can be interpreted as bipolar
flows/jets (e.g., V 1329 Cygni: Wallerstein et al. 1989), and bilaterally symmetric radio lobes as in
the case of R Aquarii (Dougherty et al., 1995). Cyg X-3 (Strom et al., 1989) and SS 433 (Kawai et al.,
1989, and references therein) are X-ray binaries with bilateral radio structure.
These two manifestations of astrophysical jets on the stellar scale are important as analogies in
the study of extragalactic jets since many fluid flow phenomena may be common. Moreover, their
central engines are more or less explicitly known astrophysical systems, and, being nearby, the jets
in these systems can be studied in great detail even with relatively coarse resolution.
A third X-ray binary, Sco X-1, was believed to have bilateral radio structure until recently
(Geldzahler and Fomalont, 1986). However, a careful reanalysis of the data, taking the radio core
variability into account, has shown the radio “lobes” to be unrelated background sources
(Fomalont and Geldzahler, 1991), although there are hints of periodic variability in the radio
emission from the core itself (Bradshaw et al., 1997).
Recently, two stellar twin radio jets associated with the two c-ray sources GRS
1915#105(1950.0) and GRO J1655!40(2000.0) (,X-ray nova Scorpii) were found to exhibit
superluminal motion, earlier seen only in distant AGNi and quasars (Mirabel and Rodrı´guez, 1994,
Tingay et al., 1995).
2.2. Straight doubles
The radio continuum emission of straight doubles is limited to a band on the sky. Apart from
two steep spectrum radio components on two sides of the radio galaxy or quasar, there is also a flat
spectrum radio core at the parent object. So the doubles are sometimes called triples. Though
quasars as a class are more powerful and asymmetric on average, there is no essential difference in
the radio structure of radio galaxies and quasars considered individually.
2.2.1. Edge darkening and turning sequences
The extended (i.e., tens of kpc and longer) radio structures of straight doubles may be conve-
niently arranged in two sequences (Miley, 1980): the edge-darkening sequence and the turning
sequence. Cyg A, the first double radio source and one of the most extensively studied, is an
94 D.G. Banhatti / Physics Reports 303 (1998) 81—182

Fig. 2. The edge-brightened extended powerful double radio source 1331!099(1950.0) (courtesy Chris Salter and
Ashok Singal).
edge-brightened double (Miley and Wade, 1971; Perley et al., 1984a). Total intensity map of
a recently discovered similar double 1331!099(1950.0) (A.K. Singal and C.J. Salter, private
communication) is shown in Fig. 2. The two components have bright double hotspots at their outer
edges. Straight lines from the brightest points on the two sides to the core are off exact alignment by
3°. This is a general property of straight doubles, the misalignment being a few degrees on average
(Macklin, 1981). The median bending angle is & 10° (e.g., Valtonen et al., 1994). The radio
luminosity decreases systematically from edge-brightened to edge-darkened straight doubles
(Fanaroff and Riley, 1974; Jones, 1990). Straight doubles in the turning sequence have S- and
Z-shapes, ranging from a gentle S to an abrupt Z. They are all of the edge-brightened type. For
example, 3C 88 (,0325#023), 3C 105 (,0404#035) and 4C 00.56 (,1514#004) are
S-shaped, while 1602#093 is Z-shaped (Alok Patnaik and Dilip Banhatti, unpublished). The
overall cone angle of the S- or Z-shape can be measured.
2.2.2. Symmetry parameters of straight doubles
For straight doubles, arm ratios, flux density ratios, brightness ratios and misalignment angles
have been used to specify their degree of symmetry (Ingham and Morrison, 1975; Banhatti, 1979,
1980; Swarup and Banhatti, 1979, 1981; Macklin, 1981; Saikia, 1984; Teerikorpi, 1984; Zie7 ba and
Chyz5 y, 1991; Best et al., 1995; Scheuer, 1995). Another parameter is the fraction of core emission at
a standard emitted frequency (Kapahi and Saikia, 1982a,b; Orr and Browne, 1982). It is possible to
form moments of the integrated brightness distribution along the major axis of the double. The
moments give different measures of the structure. The first moment gives the centroid, the second
moment the compactness, the third one skewness (indicating if the structure is edge-darkened or
edge-brightened, for example), etc. Ekers (1982) hints at a beginning in the use of moments and
D.G. Banhatti / Physics Reports 303 (1998) 81—182 95