1: %% \documentclass[12pt,preprint]{aastex}
2: %% \documentclass[manuscript]{aastex}
3: %% \documentclass[preprint2]{aastex}
4: %% \documentclass[preprint2,longabstract]{aastex}
5: \documentclass{emulateapj}
6:
7: \slugcomment{To appear in the {\em Astrophysical Journal Letters}}
8: \shorttitle{Intermediate-Mass Black Hole in NGC\,4395}
9: \shortauthors{Wrobel \& Ho}
10: \begin{document}
11: \title{Radio Emission on Sub-Parsec Scales from the
12: Intermediate-Mass Black Hole in NGC\,4395}
13: \author{J. M. Wrobel\altaffilmark{1} and L. C. Ho\altaffilmark{2}}
14:
15: \altaffiltext{1}{National Radio Astronomy Observatory, P.O. Box O,
16: Socorro, NM 87801; jwrobel@nrao.edu}
17:
18: \altaffiltext{2}{The Observatories of the Carnegie Institution of
19: Washington, 813 Santa Barbara Street, Pasadena, CA 91101;
20: lho@ociw.edu}
21:
22: \begin{abstract}
23: The Seyfert 1 nucleus of NGC\,4395 is energized by a black hole of
24: mass $3.6 \times 10^5 M_\odot$ (Peterson et al.), making it one of
25: only two nuclear black holes of intermediate mass, $10^3 - 10^6
26: M_\odot$, detected in the radio regime. Building upon UV and X-ray
27: evidence for outflows from this Seyfert nucleus, the VLBI High
28: Sensitivity Array was used at 1.4~GHz to search for extended structure
29: on scales greater than 5~mas (0.1~pc). Elongated emission was
30: discovered, extending over 15~mas (0.3~pc) and suggesting an outflow
31: on sub-parsec scales from this intermediate-mass black hole. The
32: Seyfert nucleus is located at the center of an elliptical star
33: cluster, and the elongation position angle of the sub-parsec radio
34: structure is only 19\arcdeg\, from the star cluster's minor axis.
35: \end{abstract}
36:
37: \keywords{galaxies: active ---
38: galaxies: individual (NGC\,4395) ---
39: galaxies: nuclei ---
40: galaxies: Seyfert ---
41: radio continuum: galaxies}
42:
43: \section{Motivation}
44:
45: Dynamical studies have established that supermassive black holes, with
46: masses $10^6 - 10^9 M_\odot$, occur in the nuclei of most nearby
47: galaxies with stellar bulges. The focus has now turned to searches
48: for intermediate-mass black holes (IMBHs), with masses $10^3 - 10^6
49: M_\odot$, in nearby galactic nuclei. Finding IMBHs will help to
50: define the nature of the seed black holes that figure in models for
51: the growth of black holes over cosmic time and to predict the
52: gravitational radiation background expected for {\em LISA\/} due to
53: mergers of IMBHs. In this mass regime, dynamical searches currently
54: fail beyond the Local Group and are being replaced by surveys for
55: signatures of active galactic nuclei \citep[AGN;][]{gre04}. Such AGN
56: searches are strongly guided by the discovery of a candidate IMBH in
57: \object{NGC 4395} \citep{fil03}, recently established as a bona fide
58: nuclear IMBH with mass (3.6$\pm$1.1)$\times 10^5 M_\odot$
59: \citep{pet05}.
60:
61: NGC\,4395 is a bulgeless Sdm galaxy at a Cepheid distance of
62: 4.3$\pm$0.3~Mpc with a scale 0.021~pc~mas$^{-1}$ \citep{thi04}. Its
63: nucleus has the emission properties of a Seyfert 1, with broad
64: permitted optical and UV emission lines \citep{fil93,fil03,pet05} and
65: a pointlike, hard X-ray source that is highly time-variable
66: \citep{shi03,mor05,vau05}. The reverberation study by \citet{pet05}
67: reports a characteristic diameter for the UV broad-line-region (BLR)
68: of 7.0$\times 10^{-5}$~pc or 0.0033~mas. There are hints of an
69: outflowing UV absorber close to the nucleus of this AGN \citep{cre04}.
70: Also, in luminous AGNs, warm X-ray absorbers close to the nucleus are
71: often associated with outflows; in NGC\,4395 such absorbers have been
72: inferred from variations in their ionization \citep{shi03} and/or
73: column densities \citep{mor05}. While these X-ray and UV hints of
74: outflows are intriguing, seeking structural evidence for an outflow is
75: an essential next step.
76:
77: NGC\,4395 is one of only two nuclear IMBHs that have been detected in
78: the radio regime \citep{gre06}, important because VLBI on mas scales
79: can then be used to search for structural evidence for an outflow on
80: sub-parsec scales. At 1.4~GHz, the Very Long Baseline Array
81: \citep[VLBA;][]{nap94} detected one source, with flux density
82: 0.53$\pm$0.13~mJy, spectral power $1.2 \times 10^{18}$~W~Hz$^{-1}$,
83: diameter less than 11~mas (0.2~pc), and brightness temperature more
84: than 2 million K \citep{wro01}. While these traits are consistent
85: with the VLBA detection being accretion-powered by the IMBH, the
86: resolution and sensitivity of the VLBA image failed to provide
87: quantitative structural information. This Letter reports improved
88: imaging of the Seyfert nucleus of NGC\,4395 at 1.4~GHz, using a VLBI
89: High Sensitivity Array (HSA) consisting of the VLBA, the Very Large
90: Array \citep[VLA;][]{tho80} operating in its phased mode, and the
91: Robert C. Byrd Green Bank Telescope
92: \citep[GBT;][]{jew04}.\footnote{The VLBA, the VLA, and the GBT are
93: operated by the National Radio Astronomy Observatory, which is a
94: facility of the National Science Foundation, operated under
95: cooperative agreement by Associated Universities, Inc.} The HSA
96: imaging is described in Section 2. The implications for the Seyfert
97: nucleus are explored in Section 3, with an emphasis on the discovery
98: of elongated structure on sub-parsec scales.
99:
100: \section{HSA Imaging}
101:
102: The HSA was used to observe NGC\,4395 and calibrators on 2005 May 1
103: 0000-0800 UT. Data were acquired in dual circular polarizations with
104: 4-level sampling and at a center frequency of 1.43849~GHz with a
105: bandwidth per polarization of 32~MHz, formed from four contiguous
106: baseband channels each of width 8~MHz. Antenna separations spanned
107: 50-8600 km. Phase-referenced observations were made in the nodding
108: style. A 200 s observation of NGC\,4395 was preceded and followed by
109: a 100 s observation of the phase, rate, and delay calibrator
110: J1220+3431 about 1.5\arcdeg\, from NGC\,4395. Observation and
111: correlation assumed a coordinate equinox of 2000. All {\em a
112: priori\/} calibrator positions were taken from the Goddard VLBI global
113: solution 2004 f,
114: \footnote{http://gemini.gsfc.nasa.gov/solutions/2004f\_astro/} and had
115: one-dimensional errors at 1 $\sigma$ better than 1 mas. The {\em a
116: priori\/} position for NGC\,4395 was from \citet{wro01} after
117: adjusting for the solution 2004 f position for J1220+3431.
118:
119: Data editing, calibration, and imaging were done using the 2005
120: December 31 release of the NRAO AIPS software, following strategies
121: outlined in the AIPS
122: Cookbook\footnote{http://www.aoc.nrao.edu/aips/cook.html}.
123: Correlation was affected by inappropriate Earth Orientation
124: Parameters, so corrections to those parameters were made
125: \citep{wal05}. Data deletion was based on system flags recorded at
126: observation and media weights recorded at correlation. Data
127: compromised by radio frequency interference were also deleted, as were
128: data on baselines involving any antenna observing below an elevation
129: of 20\arcdeg. For NGC\,4395, these steps yielded a total integration
130: time of 4 hr. Corrections for the dispersive delays caused by the
131: Earth's ionosphere were made using electron-content models from the
132: Jet Propulsion Laboratory. Ancillary HSA data were used to set the
133: amplitude scale to an accuracy of about 5\%, after first correcting
134: for sampler errors. The visibility data for J1220+3431 were used to
135: generate phase-referenced visibility data for NGC\,4395 and for the
136: check source J1215+3448, observed to assess both the differential
137: astrometry and the coherence losses. All calibrators were
138: phase-self-calibrated. No self-calibrations were performed on
139: NGC\,4395. No polarization calibration was performed, as only an
140: upper limit on the linear polarization percentage of NGC\,4395 was
141: sought. AIPS task {\em imagr\/} was used to image all calibrators in
142: Stokes $I\/$ and NGC\,4395 in Stokes $I\/$, $Q\/$, and $U\/$.
143:
144: The visibility data for NGC\,4395 were imaged with natural weighting,
145: achieving the expected theoretical rms noise within the field-of-view
146: limits set by time and bandwidth averaging. The most constraining
147: limit follows from accepting, at the field edge, a 5\% drop in the
148: peak amplitude due to averaging over each 8 MHz baseband channel; the
149: resulting field-of-view diameter is about 590~mas. A region of
150: diameter 550~mas (12~pc) was searched in Stokes $I\/$ to match the
151: upper limit of the unresolved VLA detection \citep{ho01}. The Seyfert
152: nucleus was easily detected above 5 $\sigma$ in the cleaned image and
153: found to be slightly extended, with an integrated flux density of
154: 0.74$\pm$0.04~mJy being recovered (Fig.~1, left). There were also
155: hints of 5 $\sigma$ emission about 250~mas (5.2~pc) to the west of the
156: nucleus. Nothing else was detected in Stokes $I\/$ above 5 $\sigma$
157: or a brightness temperature threshold of 250,000~K. Nothing was
158: detected in Stokes $Q\/$ or $U\/$ so those images were not cleaned.
159: Near the Seyfert nucleus, the image of linearly polarized intensity $P
160: = \sqrt{Q^2 + U^2}$ revealed no emission above 5 $\sigma$, implying an
161: upper limit of 2\% in the polarization percentage.
162:
163: The visibility data for NGC\,4395 were also imaged with a
164: ``robustness'' parameter of -0.3 for an optimal balance between
165: resolution and sensitivity \citep{bri95}.
166: Guided by the 4 $\sigma$ contours in the
167: cleaned Stokes $I\/$ image (Fig.~1, right), emission extends from the
168: peak for about 15~mas (0.3~pc) along a position angle (PA) of about
169: 28\arcdeg. Analysis of the check source J1215+3448, observed
170: 0.9\arcdeg\, from J1220+3431 and imaged with the same robustness,
171: implied a correction of about 2\% for residual errors in the phase
172: calibration and a one-dimensional error of about 3 mas in the
173: differential astrometry. NGC\,4395 was observed 1.5\arcdeg\, from
174: J1220+3431, so for that still-desirable geometry these corrections are
175: conservatively doubled to about 4\% and 6 mas, respectively. A
176: quadratic fit to the peak in the right panel of Figure~1 gave a
177: corrected value of 0.24~mJy~beam$^{-1}$ and a position of
178: $\alpha(J2000) = 12^{h} 25^{m} 48^{s}.8774$ and $\delta(J2000) =
179: 33^{\circ} 32' 48''.715$, with a one-dimensional error of 6~mas (the
180: quadratic sum of the errors due to the signal-to-noise ratio, in the
181: position for J1220+3431, and in the differential astrometry).
182:
183: A standard analysis of the data from the internal VLA baselines
184: yielded an integrated flux density of 1.69$\pm$0.08~mJy with a
185: deconvolved diameter of less than 2\arcsec\, (42~pc), plus no evidence
186: for adjacent emission on larger scales. For comparison, \citet{ho01}
187: reported an integrated flux density of 1.68$\pm$0.09~mJy with a
188: deconvolved diameter of less than 550~mas (12~pc) on 1999 Aug 29 UT,
189: and a spectral index $\alpha = -0.60\pm0.08$ ($S \propto
190: \nu^{\alpha}$) between 1.4 and 4.9~GHz. The HSA imaging thus recovers
191: about 0.4 of the flux density available at 1.4~GHz.
192:
193: \section{Implications and Future Directions}
194:
195: The IMBH that energizes the Seyfert nucleus in NGC\,4395 has an
196: Eddington ratio of only $\sim 1.2 \times 10^{-3}$ \citep{pet05}. By
197: analogy with supermassive black holes and stellar-mass black holes,
198: such an Eddington ratio suggests that this IMBH is in a low/hard or
199: quiescent state and, thus, is expected to launch a steady polar
200: outflow \citep{ho05,jes05,nag05,nar05}. The elongated structure in
201: Figure~1 is indeed suggestive of a radio outflow on sub-parsec scales
202: from the IMBH in NGC\,4395. This radio structure extends for about
203: 15~mas (0.3~pc) along a PA of about 28\arcdeg. From the observed
204: smoothness of the H$\alpha$ line, \citet{lao06} suggest that the BLR
205: gas is structured as a smooth, rotationally-dominated flow, most
206: likely in a geometrically thick configuration. This geometrically
207: thick structure may be involved in launching the suspected radio
208: outflow.
209:
210: For perspective, 20 of the 44 low-luminosity AGN (LLAGN) surveyed at
211: mas resolution by \citet{nag05} do show structure on sub-parsec
212: scales. Six of those 20 have sub-parsec scale structures with weak or
213: no known larger scale jets, perhaps because the jet does not propagate
214: beyond the inner few parsecs. For NGC\,4395 at 1.4~GHz, the VLA
215: source lacks adjacent emission and has the same integrated flux
216: densities for deconvolved sizes of 10-40~pc, whereas the HSA imaging
217: traces a suspected outflow on sub-parsec scales. Given these VLA and
218: HSA properties, NGC\,4395 can now be added to the subset of LLAGN
219: whose emission does not extend beyond the inner few parsecs. The
220: spectral power of the emission recovered in the left panel of Figure~1
221: is $P_{1.4} = 1.6 \times 10^{18}$~W~Hz$^{-1}$, while that for the
222: corrected peak in the right panel of Figure~1 is $P_{1.4}= 5.3 \times
223: 10^{17}$~W~Hz$^{-1}$. As expected, these powers bracket the value
224: derived from the VLBA discovery image \citep{wro01}. The sub-parsec
225: scale of the suspected outflow in NGC\,4395 corresponds to 10$^7$
226: Schwarszchild radii, very different from the upper limits of 10$^3$ to
227: 10$^4$ Schwarszchild radii inferred for several LLAGN with flat radio
228: spectra \citep{and04}.
229:
230: The potential importance of outflows in mediating black-hole growth
231: has been highlighted by \citet{pel05}, motivating an estimate of the
232: kinetic luminosity of the suspected outflow from NGC\,4395. But
233: unless the environs of the LLAGN are spatially resolved, such
234: estimates can be very model-dependent. The environs of the LLAGN in
235: M51, at a distance of 8.4~Mpc, are spatially resolved in the radio,
236: optical, and X-ray regimes \citep{ter01}. Moreover, VLA imaging at
237: 1.4~GHz shows that the LLAGN in M51 is only eight times as powerful as
238: the LLAGN in NGC\,4395, and similarly compact \citep{ho01}. For M51,
239: \citet{ter01} estimate that the kinetic luminosity of its jet is
240: comparable to the bolometric luminosity of its LLAGN. For NGC\,4395,
241: the bolometric luminosity of its LLAGN is $\sim 5.4 \times
242: 10^{40}$~ergs~s$^{-1}$ \citep{pet05}, so by analogy with M51, a
243: kinetic luminosity of this order could be available within the inner
244: few parsecs of NGC\,4395. Approximating the radio luminosity of
245: NGC\,4395 as the product of the frequency times power, the luminosity
246: of the emission recovered in the left panel of Figure~1 is $L_{1.4} =
247: 2.4 \times 10^{34}$~ergs~s$^{-1}$, while that for the corrected peak
248: in the right panel of Figure~1 is $L_{1.4} = 7.6 \times
249: 10^{33}$~ergs~s$^{-1}$. Although these radio luminosities are about
250: six orders of magnitude below the bolometric luminosity of the Seyfert
251: nucleus, the HSA imaging helps set a characteristic length scale for
252: the available kinetic luminosity.
253:
254: The absorption-corrected, time-averaged 2-10 KeV luminosity of the
255: Seyfert nucleus in NGC\,4395 is $L_X = 8.8 \times
256: 10^{39}$~ergs~s$^{-1}$ \citep{mor05}. Combining this with the 5-GHz
257: luminosity of $L_5 = 8.8 \times 10^{34}$~ergs~s$^{-1}$ \citep{ho01},
258: the nucleus of NGC\,4395 is formally radio-quiet, with $log(L_5/L_X) =
259: -5$ \citep{ter03}. Similarly, the nucleus is formally radio-quiet
260: when cast in terms of its radio-to-optical ratio \citep{gre06}.
261: Still, this Seyfert nucleus seems able to support a radio outflow,
262: suggesting that the physically defining characteristic is the
263: Eddington ratio of its energizing IMBH. Hence while NGC\,4395 can be
264: classed as a narrow-line Seyfert 1 galaxy energized by an IMBH, its
265: radio properties might be expected to differ from those for a sample
266: of similar galaxies selected to have Eddington ratios near unity
267: \citep{gre06}.
268:
269: The PA of the sub-parsec scale radio structure, about 28\arcdeg, can
270: be compared to the other nuclear structures in NGC\,4395 imaged at
271: lower resolution with the {\em Hubble Space Telescope} \citep[{\em
272: HST};][]{fil93,mat99,fil03}. In a WFPC2 I-band image, the Seyfert
273: nucleus is located at the photometric center of a nuclear star cluster
274: with a half-light diameter of 380~mas (8.0~pc), an axis ratio of 0.83,
275: and a minor-axis PA of 9\arcdeg\, \citep{fil03}, confirming the
276: earlier findings in the I band by \citet{mat99}. In contrast, the
277: galactic disk has a minor-axis PA of 54\arcdeg\, as traced by HI
278: \citep{swa99} and 57\arcdeg\, as traced by starlight \citep{ho01}.
279: The elongation PA of the sub-parsec radio structure in NGC\,4395 is
280: thus closer to the minor axis of the star cluster than that of the
281: galactic disk. But the significance of this is unclear, given the
282: propensity of jet directions in Seyfert galaxies to be uncorrelated
283: with the minor axes of their galactic disks \citep{kin00}.
284:
285: In a WFPC F502N image centered on the [OIII] $\lambda$5007 line, the
286: nuclear structure is somewhat resolved with emission-line gas
287: extending to the west of the Seyfert nucleus \citep{fil93}. Using
288: their WFPC2 B-band and I-band images, \citet{mat99} do find evidence
289: for two structures to the west, namely a small blue arc offset by
290: 250~mas (5.2~pc) and an extended blue plume offset by about 1\arcsec\,
291: (21~pc). \citet{mat99} argue that these two blue features are
292: probably due to [OIII] emission from gas within a nuclear ionization
293: cone. The opening angle of the cone structure covers PAs of
294: 265-280\arcdeg, whereas the sub-parsec scale radio structure defines
295: PAs of about 28/208\arcdeg. The relation between these two structures
296: is not presently understood, perhaps due to projection effects for
297: both structures.
298:
299: The naturally-weighted HSA image shows hints of 5 $\sigma$ emission
300: about 250~mas (5.2~pc) to the west of the nucleus, thus in the
301: vicinity of the small blue arc reported by \citet{mat99}. The reality
302: of this faint extranuclear emission at 1.4~GHz is, at present, far
303: from certain. But if it is real, it has peak brightness temperatures
304: above 250,000~K in the vicinity of probable [OIII] emission
305: \citep{mat99}. As the HSA imaging of the nucleus has recovered only
306: about forty percent of the flux density available at 1.4~GHz, the
307: faint extranuclear emission could be providing the first clues about
308: the location of the approximately 1~mJy still missing from the HSA
309: image. HSA imaging with higher sensitivity could test the reality of
310: the faint extranuclear emission and should be done. Also, {\em HST\/}
311: spectroscopic imaging of the small blue arc and the extended blue
312: plume discovered by \citet{mat99} should be done to test the
313: suggestion by those authors that these blue structures trace [OIII]
314: emission.
315:
316: The nuclear star clusters in other late-type spiral galaxies are in
317: most respects very similar to that in NGC\,4395, and within each star
318: cluster a mixture of stellar populations of different ages is found,
319: suggesting a prolonged formation period \citep{wal06}. For NGC\,4395,
320: few constraints on the cluster's stellar populations are available.
321: \citet{fil93} find no UV evidence for the P Cygni profiles that would
322: be expected from winds from massive OB stars. \citet{fil03} do detect
323: the calcium infrared triplets, indicating that red supergiants are
324: present. The half-light diameter of the nuclear star cluster is
325: 380~mas (8.0~pc), so the HSA imaging samples most of the cluster's
326: volume. To further constrain the stellar populations, as well as to
327: assist with the radio-optical registration, the naturally-weighted HSA
328: image was searched for cluster emitters within the half-light
329: diameter. None were found above a peak flux density of 0.05~mJy per
330: beam, corresponding to a brightness temperature threshold of 250,000~K
331: and a power threshold of $P_{1.4}= 1.1 \times 10^{17}$~W~Hz$^{-1}$.
332: For comparison, an analog to the radio counterpart to an ultraluminous
333: X-ray source, at a distance of 4.8~Mpc in NGC\,5408 \citep{kar03},
334: would emit at the 0.3-mJy level at 4.8~GHz at the distance of
335: NGC\,4395. Such emission would be consistent with the VLA photometry
336: of NGC\,4395 at 4.9~GHz \citep{ho01}. However, given its
337: steep-spectrum nature, such a radio analog would be boosted well above
338: the HSA detection threshold at 1.4~GHz and, thus, cannot be present
339: within the star cluster's half-light diameter. HSA imaging with
340: higher sensitivity would help constraint other classes of cluster
341: emitters.
342:
343: %% Included in this acknowledgments section are examples of the
344: %% AASTeX hypertext markup commands. Use \url without the optional [HREF]
345: %% argument when you want to print the url directly in the text. Otherwise,
346: %% use either \url or \anchor, with the HREF as the first argument and the
347: %% text to be printed in the second.
348:
349: \acknowledgments The authors thank J. Ulvestad for useful feedback on
350: the draft manuscript.
351:
352: {\it Facilities:} \facility{GBT}, \facility{VLA}, \facility{VLBA}.
353:
354: \begin{thebibliography}{}
355: \bibitem[Anderson et al.(2004)]{and04} Anderson, J. M., Ulvestad,
356: J. S., \& Ho, L. C. 2004, \apj, 603, 42
357: \bibitem[Briggs(1995)]{bri95} Briggs, D. S. 1995, Ph. D. thesis, New
358: Mexico Institute of Mining and Technology
359: \bibitem[Crenshaw et al.(2004)]{cre04} Crenshaw, D. M., et al. 2004,
360: \apj, 612, 152
361: \bibitem[Filippenko, Ho, \& Sargent(1993)]{fil93} Filippenko, A. V.,
362: Ho, L. C., \& Sargent, W. L. W. 1993, \apj, 410, L75
363: \bibitem[Filippenko \& Ho(2003)]{fil03} Filippenko, A. V., \& Ho,
364: L. C. 2003, \apj, 588, L13
365: \bibitem[Greene \& Ho(2004)]{gre04} Greene, J. E., \& Ho, L. C. 2004,
366: \apj, 610, 722
367: \bibitem[Greene, Ho, \& Ulvestad(2006)]{gre06} Greene, J. E., Ho,
368: L. C., \& Ulvestad, J. S. 2006, \apj, 636, 56
369: \bibitem[Ho(2005)]{ho05} Ho, L. C. 2005, \apss, 300, 219
370: \bibitem[Ho \& Ulvestad(2001)]{ho01} Ho, L. C., \& Ulvestad,
371: J. S. 2001, \apjs, 133, 77
372: \bibitem[Jester(2005)]{jes05} Jester, S. 2005, \apj, 625, 667
373: \bibitem[Jewell \& Prestage(2004)]{jew04} Jewell, P. R., \& Prestage,
374: R. M. 2004, SPIE, 5489, 312
375: \bibitem[Kaaret et al.(2003)]{kar03} Kaaret, P., Corbel, S.,
376: Prestwich, A. H., \& Zezas, A. 2003, Science, 299, 365
377: \bibitem[Kinney et al.(2000)]{kin00} Kinney, A. L., Schmitt, H. R.,
378: Clarke, C. J., Pringle, J. E., Ulvestad, J. S., \& Antonucci,
379: R. R. J. 2000, \apj, 537, 152
380: \bibitem[Laor et al.(2006)]{lao06} Laor, A., Barth, A. J., Ho, L. C.,
381: \& Filippenko, A. V. 2006, \apj, 636, 83
382: \bibitem[Matthews et al.(1999)]{mat99} Matthews, L. D., et al. 1999,
383: \aj, 118, 208
384: \bibitem[Moran et al.(2005)]{mor05} Moran, E. C., Eracleous, M.
385: Leighly, K. M., Chartas, G., Filippenko, A. V., Ho, L. C., \& Blanco,
386: P. R. 2005, \aj, 129, 2108
387: \bibitem[Nagar et al.(2005)]{nag05} Nagar, N. M., Falcke, H., \& Wilson,
388: A. S. 2005, \aap, 435, 521
389: \bibitem[Napier et al.(1994)]{nap94} Napier, P. J., Bagri, D. S.,
390: Clark, B. G., Rogers, A. E. E., Romney, J. D., Thompson, A. R., \&
391: Walker, R. C. 1994, Proc.\ IEEE, 82, 658
392: \bibitem[Narayan(2005)]{nar05} Narayan, R. 2005, \apss, 300, 177
393: \bibitem[Pellegrini(2005)]{pel05} Pellegrini, S. 2005, \apj, 624,
394: 155
395: \bibitem[Peterson et al.(2005)]{pet05} Peterson, B. M., et al. 2005,
396: \apj, 632, 799
397: \bibitem[Shih, Iwasawa, \& Fabian(2003)]{shi03} Shih, D. C., Iwasawa,
398: K., Fabian, A. C. 2003, \mnras, 341, 973
399: \bibitem[Swaters et al.(1999)]{swa99} Swaters, R. A., Shoenmakers,
400: R. H. M., Sancisi, R., \& van Albada, T. S. 1999, \aap, 304, 330
401: \bibitem[Terashima \& Wilson(2001)]{ter01} Terashima, Y., \& Wilson,
402: A. S. 2001, \apjl, 560, 139
403: \bibitem[Terashima \& Wilson(2003)]{ter03} Terashima, Y., \& Wilson,
404: A. S. 2003, \apj, 583, 145
405: \bibitem[Thim et al.(2004)]{thi04} Thim, F., Hoessel, J. G., Saha, A.,
406: Claver, J., Dolphin, A., \& Tammann, G. A. 2004, \aj, 127, 2322
407: \bibitem[Thompson et al.(1980)]{tho80} Thompson, A. R., Clark, B. G.,
408: Wade, C. M., Napier, P. J., 1980, \apjs, 44, 151
409: \bibitem[Vaughan, et al.(2005)]{vau05} Vaughan, S., Iwasawa, K.,
410: Fabian, A. C., \& Hayashida, K. 2005, \mnras, 356, 524
411: \bibitem[Walcher et al.(2006)]{wal06} Walcher, C. J., Boker, T.,
412: Charlot, S., Ho, L. C., Rix, H.-W., Rossa, J., Shields, J. C., \& van
413: der Marel, R. P. 2006, \apj, submitted (astro-ph/0604138)
414: \bibitem[Walker et al.(2005)]{wal05} Walker, R. C., Dhawan, V.,
415: Brisken, W., Benson, J., Kogan, L, \& Romney, J. 2005, VLBA Test
416: Memo 69
417: \bibitem[Wrobel, Fassnacht, \& Ho(2001)]{wro01} Wrobel, J. M.,
418: Fassnacht, C. D., \& Ho, L. C. 2001, \apj, 553, L23
419: \end{thebibliography}
420:
421: \clearpage
422:
423: \begin{figure}
424: \epsscale{1.0}
425: \plottwo{f1a.eps}{f1b.eps}
426: \caption{ HSA images of Stokes $I\/$ emission from NGC\,4395 at a
427: frequency of 1.4~GHz, centered on the {\em a priori\/} observation
428: position, and spanning 70~mas (1.5~pc) per coordinate. Negative
429: contours are dashed and positive ones are solid.
430: {\em Left:\/} Natural weighting, giving an rms noise of
431: 0.010~mJy~beam$^{-1}$ (1 $\sigma$) and beam dimensions at FWHM of
432: 17~mas (0.4~pc) by 7.2~mas (0.2~pc) with elongation PA -8\arcdeg\,
433: (hatched ellipse). Linear grey scale spans -0.04~mJy~beam$^{-1}$ to
434: the peak of 0.36 ~mJy~beam$^{-1}$. Contours are at -6, -4, -2, 2, 4,
435: 6, 8, 10, 12, 14, ... 36 times $\sigma$.
436: {\em Right:\/} Robustness parameter of -0.3, giving an rms noise of
437: 0.019~mJy~beam$^{-1}$ (1 $\sigma$) and beam dimensions at FWHM of
438: 11~mas (0.2~pc) by 5.4~mas (0.1~pc) with elongation PA 0\arcdeg\,
439: (hatched ellipse). Linear grey scale spans -0.04~mJy~beam$^{-1}$ to
440: the peak of 0.23 ~mJy~beam$^{-1}$. Contours are at -6, -4, -2, 2, 4,
441: 6, 8, 10, and 12 times $\sigma$.}
442: \label{fig1}
443: \end{figure}
444:
445: \end{document}
446: