A Survey of Kiloparsec-Scale Radio Outflows in Radio-Quiet

Preparing to load PDF file. please wait...

0 of 0
100%
A Survey of Kiloparsec-Scale Radio Outflows in Radio-Quiet

Transcript Of A Survey of Kiloparsec-Scale Radio Outflows in Radio-Quiet

View metadata, citation and similar papers at core.ac.uk

brought to you by CORE
provided by RIT Scholar Works

Rochester Institute of Technology
RIT Scholar Works
Articles
6-20-2006
A Survey of Kiloparsec-Scale Radio Outflows in Radio-Quiet Active Galactic Nuclei
Jack F. Gallimore
Bucknell University
David Axon
Rochester Institute of Technology
Christopher P. O'Dea
Rochester Institute of Technology
Stefi A. Baum
Rochester Institute of Technology
Alan Pedlar
University of Manchester

Follow this and additional works at: http://scholarworks.rit.edu/article
Recommended Citation
Jack F. Gallimore et al 2006 AJ 132 546 https://doi.org/10.1086/504593
This Article is brought to you for free and open access by RIT Scholar Works. It has been accepted for inclusion in Articles by an authorized administrator of RIT Scholar Works. For more information, please contact [email protected]

A Survey of Kiloparsec-Scale Radio Outflows in Radio-Quiet Active Galactic Nuclei
Jack F. Gallimore, Department of Physics & Astrronomy, Bucknell University, Lewisburg, PA 17837 USA, [email protected]
David J. Axon and Christopher P. O’Dea, Department of Physics, Rochester Institute of Technology, 84 Lomb Memorial Drive, Rochester, NY 14623
Stefi A. Baum, Chester F. Carlson Center for Imaging Science, Rochester Institute of Technology, 54 Lomb Memorial Drive, Rochester, NY 14623
Alan Pedlar, The University of Manchester, Jodrell Bank Observatory, Macclesfield, Cheshire SK11 9DL
ABSTRACT
Seyfert galaxies commonly host compact jets spanning 10—100 pc scales, but larger structures are resolved out in long baseline, aperture synthesis surveys. Previous, targeted studies showed that kiloparsec-scale radio structures (KSRs) may be a common feature of Seyfert and LINER galaxies, and the origin of KSRs may be starburst or AGN. We report a new Very Large Array (VLA) survey of a complete sample of Seyfert and LINER galaxies. Out of all of the surveyed radio-quiet sources, we find that 44% (19 / 43) show extended radio structures at least 1 kpc in total extent that do not match the morphology of the disk or its associated star-forming regions. The detection rate is a lower limit owing to the combined effects of projection and resolution. The infrared colors of the KSR host galaxies are unremarkable compared to other Seyferts, and the large-scale outflows orient randomly with respect to the host galaxy axes. The KSR Seyferts instead stand out by deviating significantly from the far-infrared – radio correlation for starforming galaxies, with tendency towards radio excess, and they are more likely to have a relatively luminous, compact radio source in the nucleus; these results argue that KSRs are powered by the AGN rather than starburst. The high detection rate indicates that Seyferts generate radio outflows over a significant fraction of their lifetime, which is much longer than the dynamical timescale of an AGN-powered jet but comparable instead to the buoyancy timescale. The likely explanation is that the KSRs originate from jet plasma that has been decelerated by interaction with the nuclear ISM. Based on a simple ram pressure argument, the kinetic power of the jet on kiloparsec scales is about three orders of magnitude weaker than the power of the jet on 10—100 pc scales. This result is consistent with the interaction model, in which case virtually all of the jet power must be lost to the ISM within the inner kiloparsec.
Keywords: galaxies: nuclei; galaxies: active; galaxies: Seyfert; galaxies: jets
p. 1

1 INTRODUCTION
As a class, Seyfert galaxies show all of the characteristics of more luminous active galactic nuclei (AGNs), including broad optical line emission, variability, and compact nuclear radio emission, which includes contributions both from star formation and the AGN (Wilson & Willis 1980; Condon et al. 1982). High angular resolution studies, which resolve out the emission related to star-formation, found that the compact, nuclear radio structures can be complex, comprising multiple, aligned components, linear or Sshaped features, lobes, and loops. The general consensus is that this high brightness temperature emission arises from low-power radio jets and outflows, analogous to the larger jets found in radio galaxies, but perhaps distorted or stunted by interaction with the surrounding ISM of the host spiral galaxy (Ulvestad, Wilson, & Sramek 1981; Booler, Pedlar, & Davies 1982; Wilson & Ulvestad 1982b; Neff & de Bruyn 1983; Pedlar et al. 1983; Wilson & Ulvestad 1983; Pedlar, Unger, & Dyson 1985; Wilson & Ulvestad 1987; Gallimore, Baum, & O'Dea 1996a; Gallimore et al. 1999; Whittle & Wilson 2004).
Radio imaging surveys of statistically well-defined samples of Seyfert galaxies have mainly employed long interferometric baselines (i.e., VLA A-array, or MERLIN) that filter out low surface brightness emission distributed over larger angular scales (Ulvestad & Wilson 1984b, 1984a; Ulvestad & Wilson 1989; Kukula et al. 1995; Thean et al. 2000; Thean et al. 2001b). These surveys may have therefore artificially suppressed the detection of larger scale outflows. For instance, while the typical jet length scale in these long-baseline surveys is ~ 10 pc – a few hundred pc, early, targeted VLA observations, obtained with shorter baselines than the later surveys, found larger scale radio outflows, exceeding 1 kpc in full extent (Wilson & Willis 1980; Hummel, van Gorkom, & Kotanyi 1983). One of the more spectacular examples is Mrk 6, which displays a ~ 600 pc nuclear jet roughly aligned with 10 kpc-scale lobes (Baum et al. 1993; Kukula et al. 1996). In fact, such large scale outflows may be common: Baum et al. (1993) used the Westerbork synthesis telescope (~ 5″ resolution at the ν = 5 GHz) to observe a sample of 13 radiobright Seyfert galaxies, and, remarkably, 12 of these sources showed extended radio lobes extending up to several kpc from the nucleus.
The questions put forward by these results are whether such kiloparsec-scale radio structures (KSRs) are truly are so common and whether they are related to the AGN or a starburst. For example, the KSR might be an extension of the nuclear radio jet or perhaps trace relic jet material connected to a past episode of nuclear activity. On the other hand, KSRs are found in starburst galaxies, such as M82 (Seaquist & Odegard 1991) and NGC 253 (Carilli et al. 1992); such starburst KSRs probably originate from a superwind generated by the cumulative effects of stellar winds and supernovae (Chevalier & Clegg 1985; Heckman, Armus, & Miley 1987; McCarthy, van Breugel, & Heckman 1987; Heckman, Armus, & Miley 1990). Superwinds escape along the rotation axis of the host galaxy, and associated KSRs therefore appear oriented (in projection, at least) along the minor axis of the host galaxy. Seven of the 12 Seyfert KSR sources in the Baum et al. (1993) survey align within 30° of the galaxy minor axis. The ratios of the KSR radio flux density to the IRAS 60µm flux density furthermore fall with the narrow range of values
p. 2

found for star-forming spiral galaxies. The authors argued in favor of a starburst origin for those particular Seyfert KSRs.
Colbert et al. (1996a) used the VLA in C-configuration (~ 5″ angular resolution at ν = 5 GHz) to study a sample of ten edge-on Seyfert galaxies and four star-forming (nonSeyfert) galaxies reasonably matching the Seyfert sample in morphology, redshift, size, and inclination. The main result was, again, that KSRs are common among Seyfert galaxies, even more common than in starburst galaxies at the sensitivity of the survey. There are also some clear morphological differences between the Seyfert and starburst radio sources: Seyfert KSRs are commonly show confined structures resembling the radio lobes of radio galaxies, whereas the starburst radio structures appear more spherically symmetric (i.e., wind-like). Moreover, they found that the Seyfert KSRs do not necessarily align with the minor axis of the host galaxy, but the starburst KSRs always align with the minor axis. Colbert et al. concluded that the Seyfert KSRs probably result from directed outflows that arise from the nuclear jet but that may have been diverted or otherwise affected by a starburst-driven wind or interactions with ISM. The scale of these outflows also raises the possibility that they may be relics of past activity (e.g., Sanders 1984).
Towards improving our understanding of Seyfert radio structures, we present a VLA1 search for KSRs in a statistically well-defined sample of optically-selected and infraredselected Seyfert galaxies. In contrast to the sample of Colbert et al. (1996a), the present sample was not limited by host-galaxy inclination and is larger: 43 active galaxies in the present sample vs. 10 in Colbert et al. We further employed the VLA in D-configuration for improved surface brightness sensitivity. The primary goals of the present study are to assess the commonness of Seyfert KSRs and to search for clues as to their origin. This paper is organized as follows. The observations and data handling are discussed in Section 2. The primary results, including how KSRs are identified, the detection fraction, and details on individual objects, are presented in Section 3. Section 4 provides analysis of the detection statistics, including a comparison of KSR and non-KSR Seyferts. We discuss the implications of these statistics and relevant timescales in Section 5 and summarize the main results in Section 6.
For the purpose of assessing the linear scales of the radio sources, we adopt H0 = 70 km s−1 Mpc−1. The exceptions are NGC 4388, NGC 4501, and NGC 4579, which belong to the Virgo cluster (Binggeli, Sandage, & Tammann 1985), and we assume a distance of 16 Mpc to these galaxies (Graham et al. 1999).
1 The Very Large Array is operated by the National Radio Astronomy Observatory, which is a facility of the National Science Foundation operated under cooperative agreement by Associated Universities, Inc.
p. 3

2 OBSERVATIONS AND DATA REDUCTION
2.1 Sample Selection
43 Seyfert galaxies were selected from the CfA (Huchra & Burg 1992) and extended 12 µm (Rush, Malkan, & Spinoglio 1993) surveys of nearby AGNs. Four of these galaxies also have been classified as LINERs (galaxies with a “low ionization nuclear emissionline region,” after Heckman 1980) according to the NASA/IPAC Extragalactic Database (NED); see Table 1. For the purposes of comparing type 1 and type 2 Seyferts (Section 4.5), we adopted the Seyfert classification provided by NED.
The CfA and 12 µm surveys were chosen because of the existing database of high resolution VLA, MERLIN, and VLBA observations already available for comparison (Kukula et al. 1995; Murray et al. 1999; Thean et al. 2000; Thean et al. 2001b). At cz ~ 5000 km s−1, the present observations provide a linear resolution of ~ 5 kpc. For comparison, the KSR of Mrk 6 extends ~ 7 kpc to either side of the nuclear (100 pcscale) radio source. We therefore restricted our observations to the redshift limited by the Seyfert galaxy NGC 5548 (cz ≤ 5149 km s−1) to ensure that we could potentially resolve Mrk 6-like KSRs from bright, nuclear radio sources in these objects. The sample was further restricted to δ > −20° for access from the VLA site.
The sample comprises 20 type 1 or intermediate type 1 (1.n) Seyferts and 23 type 2 Seyferts. There may be a bias against distant, low luminosity type 2s in this joint CfA12µm sample, however, since they are more difficult to identify by optical spectroscopy, and, owing perhaps to anisotropic emission, they be weaker mid-infrared sources compared to comparably luminous Seyfert 1s (Heckman 1995; Kukula et al. 1995; Maiolino et al. 1995; Thean et al. 2001a). The mid-infrared selection effect has been under debate, however (Spinoglio et al. 1995; Fadda et al. 1998; Nenkova, Ivezic, & Elitzur 2002), but Thean et al. (2001a) demonstrated that there is no evidence for selection against low luminosity Seyfert 2s relative to Seyfert 1s particularly in the 12µm sample.
To check for possible luminosity bias in the joint CfA-12µm sample, we performed the Kolmogorov-Smirnov (KS) test to compare the redshift distributions of type 1s vs. type 2s. If distant Seyfert 2s are missing relative to comparably luminous Seyfert 1s, the distributions should be statistically distinguishable with Seyfert 2s sampling a smaller range of redshift. For the purposes of the test we grouped the intermediate type 1s with the type 1 sample. The KS test found no significant difference in the redshift distribution of the type 1s and type 2s; the probability that they were drawn from the same parent distribution of redshifts is P = 0.91.
We also compared the distribution of 12 µm luminosities (L12) calculated from measurements in the IRAS Faint Source Catalog (Moshir et al. 1992). Spinoglio and
p. 4

collaborators have argued that L12 scales almost linearly with the bolometric luminosity of the host galaxy, including the contribution of the AGN (Spinoglio & Malkan 1989; Spinoglio et al. 1995). Again, the KS test found no significant difference in the distribution of log L12 with P = 0.99. As a side result, the fact that there is no significant difference of either the redshift distribution or L12 distribution argues that mid-infrared emission is not significantly enhanced in type 1s compared to the type 2s in the sample. We conclude that the joint CfA-12µm sample affords a fair comparison of type 1s and type 2s insofar as there is no strong bias resulting from luminosity selection effects.
Table 1 summarizes the 40 Seyfert galaxies observed for this survey. The Seyfert types and recessional velocities were obtained from the NED database. The Hubble classifications were taken either from the RC3 (de Vaucouleurs et al. 1991) or NED and were coded according to the numerical scheme of the RC3. Of the sources that remained after filtering for the redshift and declination limits, we chose not to re-observe four wellstudied galaxies: M51 (Ford et al. 1985; Crane & van der Hulst 1992), NGC 3079 (de Bruyn 1977; Duric et al. 1983; Irwin & Saikia 2003), NGC 3516 (Wrobel & Heeschen 1988; Baum et al. 1993), and NGC 4151 (Baum et al. 1993; Pedlar et al. 1993). These four sources were however included in the sample in the statistical analyses to follow (see Section 4). The more distant Seyfert Mrk 9 was observed serendipitously to accommodate the observing schedule. The breakdown of the entire sample by Seyfert type is 5 type 1s, 15 intermediate type 1s, and 23 type 2s.
2.2 Observations and data reduction
All of the observations were made with the VLA operating in D-configuration and using the standard 5 GHz continuum mode. The observations took place on 20 Feb., 23 Feb, and 3 Mar. 2003. Each source was observed for a total integration time of at least 11 minutes with bracketing scans of a nearby phase reference source. The characteristic angular resolution (synthetic beam) for this observing configuration is 15″−20″, depending on declination (Table 1).
Data reduction was performed using standard procedures in the AIPS software environment. After flagging of obviously spurious data, the flux scale was established based on scans of 3C 48 and 3C 286 and bootstrapped to the phase references. Phases were calibrated against point source models for the phase references, and the solutions were transferred by interpolation in time to the target sources. Initial images were generated using numerical Fourier transforms and CLEAN deconvolution (using the AIPS task IMAGR), and all of the observed Seyfert galaxies were detected. The phases were subsequently self-calibrated based on CLEAN models of the Seyfert radio sources and neighboring sources within the primary beam. The typical background noise level (rms) of the final images is ~0.05 mJy beam−1 (Table 1), or roughly 4 mK in terms of brightness temperature. Note that the brightness temperatures of spiral galaxy disks are of order tens of mK (Hummel 1980; Condon et al. 1982), and so the present observations are sensitive to disk radio emission from the host galaxy.
p. 5

2.2.1 Point source subtraction
Most of the Seyfert galaxies observed for this survey harbor compact radio structure that is unresolved by the VLA C-array and is sufficiently bright to hide fainter, extended structure even outside the beam-width (cf. Baum et al. 1993). To uncover potentially hidden radio continuum structure, we subtracted the brightest unresolved component from each of the CLEAN images. The results are summarized in Table 2. The procedure involved simultaneously fitting three components within a fitting window, usually 4-5 beamwidths on a side. The fitted components comprise (1) a gaussian matched to the beam-size, i.e., a point source; (2) a second, extended gaussian intended to represent any underlying radio structure; and (3) a flat background to account for smooth structure larger than the fitting window. The unresolved component (component 1) was then subtracted from the image, leaving behind only the resolved emission.
2.2.2 Archival VLA data
The present observations of NGC 1241 revealed peculiar, external radio structure that might arise from a background radio source (in the discussion to follow, we show that the peculiar structure is probably a background radio galaxy). To investigate this source in further detail, we retrieved archival VLA observations taken at ν = 1.5 GHz. The observations occurred on 12 March 1984 with the VLA in its B-configuration. The data, which were taken as part of program AD100, were intended to study the radio source B0308-090, but NGC 1241 appears within the field.
We reduced the archival data as described in Section 2.2. The flux scale was set by a scan of 3C 138, and phase corrections were generated based on scans of the nearby calibrator sources B0238-084, B0240-217, and B0336-019. The data were converted to J2000 coordinates using the AIPS task UVFIX. Subsequent imaging was performed by Fourier inversion and CLEAN deconvolution with a forced restoring beam of 7.5″ FWHM, which slightly super-resolves the image along the declination axis.
3 RESULTS
All but five of the observed Seyfert galaxies (and Mrk 9) showed resolved structure after subtraction of the brightest point source. Except for a few cases, however, the radio source morphology alone does not suffice to interpret the nature of the extended radio emission. Distinguishing emission from KSRs and emission related to star-formation in the galaxy disk presents the chief difficulty, particularly for the marginally resolved sources. To help resolve the ambiguity, we laid the radio continuum images over Digitized Sky Survey2 (DSS) images. The results are displayed in Figure 1 − Figure 3.
2 The Digitized Sky Survey was produced at the Space Telescope Science Institute under U.S. Government grant NAG W-2166. The images of these surveys are based on photographic data obtained using the Oschin
p. 6

Based on the radio morphology and comparison with the DSS overlays, the Seyfert radio sources were broadly classified into three categories: (1) point sources, (2) disks, and (3) KSR candidates. The classifications are included in Table 2. Point sources, presented in Figure 1, are defined by a lack of significant residual emission after point source subtraction. Disk sources, displayed in Figure 2, are identified by a good morphological correspondence between the radio source and the optical galaxy; for example, a disk source might be so classified based on a match up of radio and optical spiral arms, or, for poorly resolved sources, a broad match between the shapes of the radio continuum and the brighter optical disk. Finally, the KSR candidates, Figure 3, were classified based on the presence of residual emission that meets or exceeds 1 kpc in total extent but could not be clearly matched to any structure of the optical galaxy. Sources known to have extended radio structure based on other observations but which are not resolved by the present observations were also placed in the KSR category. Several sources were difficult to distinguish owing to the faintness of the residual emission; we conservatively classified these ambiguous cases as disks if the extended radio emission fit within the bright optical disk on the DSS image. Note that these categories are not intended to describe the dominant radio source of the galaxy; if a KSR was detected at all, the source was classified as a KSR even if the disk is brighter in integrated flux density (NGC 4388 is an example).
The breakdown for the 40 Seyferts observed as part of this survey is 5 point sources, 19 disks, and 16 KSR candidates. There were three KSR sources that met the sample selection criteria but were not included in the observations, giving a total detection rate of 19 KSRs out of 43 sample Seyfert galaxies.
3.1 Notes on the KSR candidates
The radio sources from our sample that are classified as KSR candidates and their basic properties are listed in Table 3. The list includes several KSR sources that were detected at lower frequencies than the present observations, or that could not be resolved by the present observations, or which we chose not to re-observe because they were already known KSR candidates. These sources are indicated by references in parentheses. The reported KSR position angles refer specifically to extended radio structure apart from obvious disk emission. The “extent” refers to the end-to-end (or lobe-to-lobe) extent of the radio source, usually measured by image moment analysis, but not necessarily the extent with respect to the nucleus. The morphological classifications include “lobes,” meaning radio peaks that clearly disconnect from the nuclear or disk radio structure; “tongues,” which are extended radio sources that join continuously with the nuclear radio source; “linear,” which refers to sources that appear jet-like or are otherwise narrower than the “tongue” sources, and “loops,” which include figure-8, X-shaped, or other edgebrightened lobe-like structures. The “sidedness” refers to the presence of predominately one-sided or two-sided structures with respect to the nucleus. The morphology and
Schmidt Telescope on Palomar Mountain and the UK Schmidt Telescope. The plates were processed into the present compressed digital form with the permission of these institutions.
p. 7

sidedness classification necessarily depend on the resolution and sensitivity of the observations. In every case, the morphology and sidedness refer to the appearance on the finest resolution radio images that detect the extended KSR structure.
We detail below the properties of the individual KSR radio sources.
3.1.1 Mrk 348 (NGC 262)
Mrk 348 is luminous Seyfert 2 galaxy (Arakelian, Dibay, & Esipov 1972; Koski 1978) that displays clear polarimetric evidence for a hidden Seyfert 1 nucleus (Miller & Goodrich 1990). The radio continuum emission is dominated by a variable, VLBI-scale (~ 0.5 pc) jet (Ulvestad et al. 1999) that feeds into a larger (~ 60 pc) linear radio structure oriented roughly north-south (Neff & de Bruyn 1983; Anton et al. 2002). Baum et al. (1993) discovered large-scale radio lobes (~ 6 kpc extent) that roughly align with the small-scale jet structure. The present observations marginally resolve the large-scale lobes after point-source subtraction.
3.1.2 NGC 1068
NGC 1068 is one of the classical Seyfert galaxies (Seyfert 1943) and is well-known for harboring a hidden type 1 nucleus (Antonucci & Miller 1985). The radio continuum resolves into a 3 kpc diameter, actively star-forming disk (Wynn-Williams, Becklin, & Scoville 1985; Gallimore et al. 1996c) and a kiloparsec-scale radio jet that terminates in parabolic, edge-brightened lobes which straddle the nucleus (Wilson & Ulvestad 1987; Gallimore et al. 1996c). The present observations marginally resolve the twin lobes after point-source subtraction. The central arcsecond further resolves into a 100 pc-scale jet and compact radio knots (Gallimore et al. 1996c; Muxlow et al. 1996). The compact jet appears to have been deflected by interaction with the circumnuclear ISM (Gallimore et al. 1996a; Gallimore et al. 1996b).
3.1.3 NGC 1241
NGC 1241 is a Seyfert 2 galaxy (Dahari 1985) interacting with the neighboring spiral galaxy NGC 1242 (Keel 1996 and references therein). Our observations (Figure 3) show extended radio continuum emission associated with the optical disk, and, peculiarly, two compact radio sources offset by ~ 40″ − 60″ northeast and east of the nucleus that resemble a single-sided ejection structure. The relative brightness and misaligned position angle of the two offset sources however suggest that they might be associated with a background source or sources. Figure 4 shows an archival HST/WFPC-2 image with the archival VLA 1.5 GHz image overlaid as contours. The HST image reveals a resolved, 12th magnitude (in F606W) galaxy, which also appears as a faint knot on the DSS image, located between the two offset radio sources. There are also a handful of background galaxies that might be part of an associated cluster. It therefore seems likely that the double radio source to the northeast is actually a background source.
p. 8

The archival 1.5 GHz image however reveals emission extending ~ 16″ (4.5 kpc) from the nucleus towards the east (PA 100°) that cannot be resolved in the present 5 GHz survey. There is no matching morphology in the surrounding galaxy disk to suggest a significant contribution owing to disk star formation, and so we classify this source as a KSR candidate.
3.1.4 NGC 1320
NGC 1320 is a low luminosity Seyfert 2 galaxy (de Robertis & Osterbrock 1986). Colbert et al. (1996a) first reported the faint radio continuum structure extending 5″ (~ 1 kpc) south of the nucleus. The present observations reveal still fainter radio emission continuing to ~ 14″ (2.6 kpc) south. Longer baseline measurements resolve out the KSR, but the nuclear radio continuum source remains unresolved with size < 0.2″ (< 40 pc) (Thean et al. 2000). An HST/WFPC-2 image in the light of [OIII] λ5007 reveals emission extended ~ 2″ (370 pc) to the northwest along the galaxy major axis but misaligned with the faint radio extension (Ferruit, Wilson, & Mulchaey 2000).
3.1.5 NGC 2639
NGC 2639 has been classified as a type 1 Seyfert, owing to a broad component of Hα (Huchra, Wyatt, & Davis 1982; Keel 1983), or LINER (Ho, Filippenko, & Sargent 1993), and there is a compact, H2O megamaser source argued to trace a molecular accretion disk (Wilson, Braatz, & Henkel 1995; Wilson et al. 1998 and references therein). The nucleus also contains a linear, jet-like radio source, 1.6″ (370 pc) in extent (Ulvestad & Wilson 1984b; Gallimore et al. 1999; Thean et al. 2000). The AGN is marked by a ~ 7 mas (1.6 pc), VLBI-scale radio jet (Hummel et al. 1982; Wilson et al. 1998) that aligns with the arcsecond-scale jet. The present observations show additional radio emission extending ~53″ (~ 12 kpc) and roughly aligned with the compact radio source axis.
3.1.6 NGC 2992
NGC 2992 is an edge-on, type 2 narrow line X-ray galaxy (Ward et al. 1978; Ward et al. 1980; Huchra et al. 1982) with an obscured BLR (e.g., Goodrich, Veilleux, & Hill 1994). Evidence for a large scale outflow, closely aligned with the galaxy minor axis, comes from observations of extended optical emission line gas (e.g., Veilleux, Shopbell, & Miller 2001) and radio continuum (Ward et al. 1980; Hummel et al. 1983). Detected at 1.4 GHz, there is a plume of radio continuum extending up to ~ 81″ (12.5 kpc) east of the radio nucleus. Our present observations fail to detect this plume at 5 GHz (cf. Colbert et al. 1996b). The extended radio source is therefore probably steep-spectrum and wellresolved by the VLA D-array beam; for comparison, the beam area of our observations is about 10% that of the Fleurs synthesis array employed by Ward et al. (1980). The nucleus
p. 9
ObservationsKsrsSampleSourcesGalaxies