1: \documentclass[12pt,preprint]{aastex}

2: 

3: % The following two lines allow emulation of ApJ look and feel (and

4: % page length):

5: %\usepackage{emulateapj5}

6: %\usepackage{apjfonts}

7: 

8: \def\plotfiddle#1#2#3#4#5#6#7{\centering \leavevmode

9: \vbox to#2{\rule{0pt}{#2}}

10: \special{psfile=#1 voffset=#7 hoffset=#6 vscale=#5 hscale=#4 angle=#3}}

11: 

12: \bibliographystyle{apj}

13: 

14: %\received{2003 April 17}

15: %\received{2003 July 12}

16: \begin{document}

17: \title{Interferometer Observations of Subparsec-scale Infrared Emission 

18: in the Nucleus of NGC 4151}

19: 

20: \author {

21: M.~Swain\altaffilmark{1},

22: G.~Vasisht\altaffilmark{1},

23: R.~Akeson\altaffilmark{2}, 

24: J.~Monnier\altaffilmark{3},

25: R.~Millan-Gabet\altaffilmark{2},

26: E.~Serabyn\altaffilmark{1},

27: M.~Creech-Eakman\altaffilmark{1},

28: G.~van Belle\altaffilmark{2},

29: J.~Beletic\altaffilmark{4},

30: C.~Beichman\altaffilmark{1},

31: A.~Boden\altaffilmark{2},

32: A.~Booth\altaffilmark{1},

33: M.~Colavita\altaffilmark{1},

34: J.~Gathright\altaffilmark{4},

35: M.~Hrynevych\altaffilmark{4},

36: C.~Koresko\altaffilmark{2},

37: D.~Le Mignant\altaffilmark{4},

38: R.~Ligon\altaffilmark{1},

39: B.~Mennesson\altaffilmark{1},

40: C.~Neyman\altaffilmark{4},

41: A.~Sargent\altaffilmark{2},

42: M.~Shao\altaffilmark{1},

43: R.~Thompson\altaffilmark{1},

44: S.~Unwin\altaffilmark{1},

45: P.~Wizinowich\altaffilmark{4}

46: }

47: 

48: \altaffiltext{1}

49: {Jet Propulsion Laboratory, California Institute of Technology, 

50: 4800 Oak Grove Dr., Pasadena, CA 91109}

51: \altaffiltext{2}

52: {Michelson Science Center, California Institute of Technology, 

53: 770 S. Wilson Ave., Pasadena, CA 91125}

54: \altaffiltext{3}{Univ. of Michigan, 941 Dennison Bldg, Ann Arbor, MI, 48109}

55: \altaffiltext{4}{W. M. Keck Observatory, California Association for 

56: Research in Astronomy, 65-1120 Mamalahoa Hwy., Kamuela, HI 96743}

57: 

58: 

59: 

60: \begin{abstract}

61:   

62: We report novel, high-angular resolution interferometric measurements

63: that imply the near-infrared nuclear emission in NGC 4151 is

64: unexpectedly compact. We have observed the nucleus of NGC 4151 at 2.2

65: $\mu$m using the two 10-meter Keck telescopes as an interferometer and

66: find a marginally resolved source $\leq0.1$ pc in diameter.  Our

67: measurements rule out models in which a majority of the $K$ band

68: nuclear emission is produced on scales larger than this size.  The

69: interpretation of our measurement most consistent with other

70: observations is that the emission mainly originates directly in the

71: central accretion disk.  This implies that AGN unification models

72: invoking hot, optically thick dust may not be applicable to NGC 4151.

73: 

74: \end{abstract}

75: \keywords{instrumentation: interferometers -- galaxies: individual(NGC 4151) 

76: -- galaxies: Seyfert}

77: 

78: 

79: \section{Introduction}

80: 

81: NGC 4151 is the nearest and brightest Seyfert and it has been

82: identified as the ``archetype'' of Seyfert 1 galaxies \citep{ngs90}.

83: It has also been called ``one of the most enigmatic'' of galaxies

84: \citep{cr96} because the origin of the near-infrared nuclear emission

85: component, which dominates galactic near-infrared emission, is

86: unclear.  The extent to which the observed near-infrared emission

87: comes directly from the accretion disk or jet rather than from clouds

88: of gas or dust that {\it reprocess} energy from the accretion disk is

89: critical to models that attempt to unify various observational classes

90: of AGN \citep{a93}.  Over the past 30 years, arguments for both

91: thermal and nonthermal (synchrotron) emission mechanisms have been

92: proposed (see \citet{ngs90} and references therein).  Evidence of $K$

93: band emission variability on 1--2 month timescales

94: \citep{pps74,rl81,olt99} and weak (0.1--0.5\%) degrees of $H$ and $K$

95: band linear polarization \citep{krl77,ryp03} can be explained by both

96: thermal and nonthermal models.  Currently, the prevailing view is that

97: the near-infrared emission is thermal and arises from the inner edge

98: of a dust torus that is heated primarily by UV and soft x-ray emission

99: from the corona of the accretion disk surrounding a massive black hole

100: \citep{a93}.  In this picture, the source of the near-infrared

101: emission would appear spatially unresolved by the largest single

102: aperture infrared telescopes and would have low degrees of linear

103: polarization.

104: 

105: Observations at 11 $\mu$m \citep{ngs90,rpp03} strongly imply the

106: presence of a $<$35 pc diameter source where dust grains are heated to

107: around 200 K and are responsible for the mid-infrared emission.

108: Estimates for the inner radius of the dust, which could correspond to

109: the inner edge of a torus, have been as small as about 0.03 pc

110: \citep{rl81}, with values in the 0.2--1 pc range presently favored

111: \citep{a93,pk92,etk93,ryp03}.  In some AGN unification schemes, the

112: torus is supposed to be optically and physically thick \citep{a93}.

113: The radius of the broad line region (BLR) is about 0.01 pc

114: \citep{cbb90}, putting the smallest estimates of the dust torus' inner

115: radius very close to the black hole and accretion disk.

116: 

117: The parent galaxy for the NGC 4151 Seyfert nucleus is nearly face-on

118: ($i=21^{\circ}$ to the line of site) and has a weak radio jet at P.A.=

119: $77^{\circ}$ E. of N. \citep{mwp03}.  While the radio feature

120: corresponding to the UV/optical nucleus has been uncertain, the

121: candidate features contain emission knots that are unresolved on

122: angular scales of a few milliarcseconds (mas) and have a surface

123: brightnesses of a few mJy/beam \citep{mwp03,urc98}.  The 2MASS $K$

124: band flux for NGC 4151 is 8.519 $\pm$ 0.018 magnitudes, and at least

125: 90\% of the flux is contained in the central 1 arcsec

126: \citep{krl77}. The bolometric luminosity of the AGN is about $10^{44}$

127: erg s$^{-1}$, and a $10^7$ M$_{\odot}$ black hole is thought to lie at

128: the center of the BLR \citep{ulr00}.  Following \citet{mwp03}, we have

129: taken the distance as 13.3 Mpc (for $H_{0}=75$ km s$^{-1}$), where 1

130: mas corresponds to 0.065 pc (75.3 light days) in projection.

131: 

132: Interferometric combination of the twin Keck telescopes \citep{col02}

133: provides sensitivity to 2.2 $\mu$m emission on angular scales of

134: $\sim$ 1 mas.  Our observations represent a ten-fold improvement in

135: angular resolution when compared to previous near-infrared

136: measurements of AGN and make it possible to test the subparsec-scale,

137: near-infrared emission models of NGC 4151.  Optical/IR interferometry

138: has traditionally been limited to relatively bright stellar objects.

139: These observations represent the first measurement of an extragalactic

140: source with an optical/IR interferometer.

141: 

142: 

143: \section{Observations and Data Reduction}

144: 

145: \subsection{Observations}

146: 

147: The Keck Interferometer combines the light from the two 10 m Keck

148: telescopes, separated by 85 m with the baseline oriented 38$^{\circ}$

149: (E of N).  Both telescopes are equipped with adaptive optics systems

150: \citep{wiz02}, which are required for $K$ band interferometer

151: observations.  The interferometer field of view on the sky is $\sim$50

152: mas in the $K$ band, as set by a single-mode fiber ahead of the

153: detector; the interferometer is not sensitive to emission outside of

154: this field of view.  The fringe tracker uses a 4-bin synchronous fringe

155: demodulation algorithm \citep{vbc02}, similar to that used at the

156: Palomar Testbed Interferometer \citep{col99b}.  For the data presented

157: here, the system operated at a 200~Hz frame rate (the 4-bin

158: acquisition time).

159: 

160: NGC 4151 was observed with the Keck Interferometer on 2003 May 20 and

161: 21.  The data presented here are from the white-light channel

162: ($\lambda_{\rm center}$=2.18~$\mu$m and $\Delta \lambda

163: \sim$~0.3$\mu$m).  Observations consist of a series of interleaved

164: integrations on the source and several calibrators.  Each integration

165: includes 120 seconds of fringe data followed by a background

166: measurement.  Included with the fringe data are periodic measurements

167: of the single-telescope fluxes.  The data are presented as the

168: visibility amplitude squared, normalized such that an unresolved

169: object has V$^2=1.0$.  On May 20, only limited data were collected on

170: NGC~4151 and calibrators due to weather; however, the source was

171: detected and resolved on both nights, with internally consistent

172: visibility.  In the results discussed below we have used data only from

173: May 21.

174: 

175: \subsection{Data Reduction}

176: 

177: The system visibility, the instrument response to a point

178: source, is measured with respect to the calibrator stars

179: HIP58918, HD 105925, and HIP60286.  The calibrator angular sizes were

180: derived by fitting photometry from SIMBAD and 2MASS.  All of the

181: calibrators have estimated angular diameters of less than 0.2 mas and

182: are unresolved by the interferometer.  The calibrator angular size

183: uncertainty was set to 0.1 mas.  Source and calibrator data were

184: corrected for biases using sky calibrations as described by

185: \citet{col99} and averaged into blocks of 5 seconds each.  The

186: calibration procedure includes a per-scan correction for the bias due

187: to mismatched fluxes between the two telescopes (which includes Strehl

188: mismatch) using the single-telescope flux measurements.

189: 

190: The data were then calibrated for the system visibility \citep{bod98}.

191: The calibrated data points for the target source are the average of

192: the 5 seconds blocks in each integration, with an uncertainty given by

193: the quadrature of the internal scatter and the uncertainty in the

194: calibrator size.  In addition to the measurement error, we have

195: estimated any systematic error in the calibrated visibility to be less

196: than 0.05 \citep{col03}.  This systematic error was summed

197: quadratically with the measurement and calibration errors.  The

198: resulting average visibility is 0.84 $\pm$ 0.06 at an average

199: projected baseline of 82.7 m at a P.A. = 37\arcdeg E of N.  The

200: calibrated visibilities and the system visibility estimates are shown

201: in Fig. 1.

202: 

203: 

204: \section{Discussion}

205: 

206: \subsection{Compact Emission}

207: 

208: If all the light in the fringe tracker field of view (approximately

209: the diffraction limit of the single telescope) came from an unresolved

210: point source, the interferometer calibrated $V^2$ would be unity.  Our

211: measured $V^2$ is relatively high, implying the source of this

212: emission is small.  A variety of compact flux distributions could

213: produce the measured visibility; we have used three simple, geometric

214: models to estimate the angular size of the emission.  The models

215: correspond to a ``face-on'' inclination and are constrained to fit the

216: measured $V^{2}$.  If the emission is distributed as a

217: single-component Gaussian, the FWHM = 0.98 mas $\pm$ 0.18 mas.  If the

218: the emission is distributed as a ring, the ring inner diameter = 0.96

219: mas $\pm$ 0.26 with a width = 0.06 mas, where the width is set to

220: match the total $K$ band flux for an optically thick medium with a

221: temperature of 1500 k.  If the emission is distributed as a point

222: source with an extended (over-resolved) component, the point-source

223: fraction of the flux = 0.92 $\pm$ 0.04 of the flux.  Whatever the

224: details of the emission distribution or mechanism, the bulk of the

225: emission comes from a compact ($\leq$1.52 mas at the $3\sigma$ level )

226: region.  Our measurements rule out any scenario with a majority of the

227: $K$ band nuclear emission coming from a region with a radius in

228: projection larger than $0.05$ pc.

229: 

230: \subsection{Possible Emission Mechanisms}

231: 

232: We consider four possible mechanisms for the $K$ band emission.  Because

233: of the combined constraints implied by our measurement and other

234: observations, we find two of the cases to be unlikely.

235: 

236: \noindent {\bf Synchrotron Emission:} VLBA imaging of NGC 4151 at a

237: few mas scales \citep{urc98} reveals that its parsec-scale radio jet

238: is subluminous; it is several orders of magnitude fainter than the

239: parsec-scale radio jets imaged in nearby classical radio galaxies.

240: Measurements at 18 cm and 6 cm of the nuclear region reveal a

241: relatively flat spectrum ($\nu^{\alpha}$ with $\alpha \simeq -0.3$)

242: source with a peak flux density of about 1 mJy mas$^{-2}$. In

243: contrast, the large near-infrared flux density of $\simeq 10^2$ mJy

244: mas$^{-2}$ is inconsistent with the radio fluxes and implies that the

245: nuclear near-infrared source is unrelated to the radio source and, by

246: inference, unrelated to a subparsec-scale synchrotron jet.

247: 

248: \noindent {\bf Star Cluster:} Explaining the compact $K$ band emission

249: as stellar would require $\sim 3\times10^{6}$ O stars in a volume of

250: $\sim0.001$ pc$^3$.  This explanation is unlikely, as it would imply a

251: density for young stars $\sim100,000$ times higher than in our own

252: galactic center \citep{ms96,gnz00}.  A stellar component is also

253: incompatible with the measured infrared intensity fluctuations

254: \citep{rl81,qsa00}.

255: 

256: \noindent {\bf Thermal Dust:} If the dust is distributed as a

257: physically thick torus, the inner radius is much smaller than

258: typically estimated.  The smallest possible scale for thermal dust

259: emission is set by the dust sublimation radius.  Using the model of

260: \citet{bar87}, we find dust with a temperature of 1900 K,

261: corresponding to a sublimation radius of 0.05 pc, is consistent with

262: our measurement.  Thus, if the infrared emission is from centrally heated

263: dust, it is consistent with our observations only if the dust is very

264: hot and the $K$ band emission is localized near the dust sublimation

265: radius.

266: 

267: \noindent {\bf Thermal Gas:} A simple, face-on, geometrically thin

268: accretion disk model, with $T\propto r^{-3/4}$ \citep{ss73} and with

269: an inner temperature of $\sim 3.5\times 10^5$ K (consistent with the

270: soft x-ray emission \citep{ulr00}), predicts a $K$ band magnitude of

271: 8.4, 95\% of which is inside a radius of $\sim0.01$ pc.  This model

272: predicts an unresolved point source, consistent with our size

273: measurement and with approximately the measured near-infrared flux.

274: 

275: Both the thermal dust and thermal gas models are consistent with our

276: observations.  However, centrally heated dust, which is thought to

277: reprocess UV/optical radiation from the central engine, produces a

278: specific signature known as the ``infrared bump'' \citep{bar87,san89}.

279: There is no evidence for a comparable feature in NGC 4151

280: \citep{rl81,em86,alo01}.  Also, a dust torus with a well-defined, but

281: small, inner radius (corresponding to the sublimation radius) would

282: give similar time delays between fluctuations in the UV/optical flux

283: and fluctuations in the $H$ and $K$ band flux.  This is inconsistent

284: with the analysis by \cite{olt99}.  Thus the absence of the ``infrared

285: bump'' and the presence of the $H$/$K$ time delay difference both

286: support the interpretation that the bulk of the infrared flux measured

287: with the interferometer arises from an unresolved accretion disk.

288: 

289: 

290: \section{Summary and Conclusions}

291: 

292: We have measured the angular diameter of the $K$ band emission in the

293: nucleus of NGC 4151.  Our observations rule out any emission mechanism

294: that requires a large fraction of the nuclear $K$ band emission to be

295: produced at a radius of greater than 0.05 pc from the black hole.

296: Taken in the context of other observations, we interpret our

297: measurement as evidence that a majority of the $K$ band emission comes

298: from a central accretion disk.  

299: 

300: \acknowledgments

301: 

302: We dedicate these observations to the memory of Jim Kelley, the late

303: project manager and tireless advocate of the Keck Interferometer.  We

304: thank Robert Antonucci, Roger Blandford, and Julian Krolik for useful

305: discussions.  The Keck Interferometer is funded by the National

306: Aeronautics and Space Administration as part of its Navigator program.

307: Part of this work was performed at the Jet Propulsion Laboratory,

308: the California Institute of Technology, and the Michelson Science Center,

309: under contract with NASA.  Observations presented were obtained at the

310: W.M. Keck Observatory, which is operated as a scientific partnership

311: among the California Institute of Technology, the University of

312: California, and NASA. The authors wish to recognize and acknowledge

313: the very significant cultural role and reverence that the summit of

314: Mauna Kea has within the indigenous Hawaiian community.  We are most

315: fortunate to have the opportunity to conduct observations from this

316: mountain. The Observatory was made possible by the generous financial

317: support of the W.M. Keck Foundation.  This work has made use of

318: software produced by the Michelson Science Center at the California

319: Institute of Technology.  This work has also made use of the SIMBAD

320: database, operated at CDS, Strasbourg, France, and the NASA/IPAC

321: Infrared Science Archive, operated by the JPL under contract with

322: NASA.

323: 

324: % References :

325: \begin{thebibliography}{}

326: 

327: \bibitem[Alonso-Herrero et al.(2001)]{alo01} Alonso-Herrero, 

328: A., Quillen, A.~C., Simpson, C., Efstathiou, A., \& Ward, M.~J.\ 2001, \aj, 

329: 121, 1369 

330: 

331: \bibitem[Antonucci(1993)]{a93}

332: Antonucci, R. 1993, \araa, 31, 473

333: %unification models for AGNs and quasars

334: 

335: \bibitem[Barvainis(1987)]{bar87} Barvainis, R.\ 1987, \apj, 

336: 320, 537 

337: 

338: \bibitem[Boden et al.(1998)]{bod98} 

339: Boden, A.~F., Colavita, M.~M., van Belle, G.~T., \& Shao, M.\ 1998, 

340: \procspie, 3350, 872 

341: 

342: \bibitem[Cassidy \& Raine(1996)]{cr96} Cassidy, I.~\& Raine, 

343: D.~J.\ 1996, \aap, 310, 49 

344: 

345: \bibitem[Clavel et al.(1990)]{cbb90} 

346: Clavel, J.~et al.\ 1990, \mnras, 246, 668 

347: %The ultra-compact broad emission line region in NGC 4151

348: 

349: \bibitem[Colavita et al.(2003)]{col03}

350: Colavita, M.~M., et al., \apjl, submitted

351: 

352: \bibitem[Colavita(1999)]{col99} Colavita, M.~M.\ 1999, \pasp, 

353: 111, 111 

354: 

355: \bibitem[Colavita et al.(1999)]{col99b} Colavita, M.~M.~et 

356: al.\ 1999, \apj, 510, 505 

357: 

358: \bibitem[Colavita \& Wizinowich(2002)]{col02} Colavita, 

359: M.~M.~\& Wizinowich, P.~L. 2002, \procspie, 4838, 79 

360: %Keck Interferometer instrument paper

361: 

362: \bibitem[Edelson \& Malkan(1986)]{em86} Edelson, R.~A.~\& 

363: Malkan, M.~A.\ 1986, \apj, 308, 59 

364: 

365: \bibitem[Evans et al.(1993)]{etk93}

366: Evans, I.~N., Tsvetanov, Z., Kriss, G.~A., Ford, H.~C., Caganoff, S., \&

367: Koratkar, A.~P. 1993, \apj, 417, 82

368: %HST imaging of NGC 4152 NRL - torus inner radius and PA estimates

369: 

370: \bibitem[Genzel et al.(2000)]{gnz00} Genzel, R., Pichon, C., 

371: Eckart, A., Gerhard, O.~E., \& Ott, T.\ 2000, \mnras, 317, 348 

372: %Stellar dynamics in the Galactic Centre: proper motions and anisotropy

373: 

374: \bibitem[Kemp et al.(1977)]{krl77}

375: Kemp, J.~C, Rieke, G.~H., Lebofsky, M.~J, \& Coyne, G.~V. 1977, \apj, 

376: 215, L107 

377: %infrared polarization measurements

378: 

379: \bibitem[Morris \& Serabyn(1996)]{ms96} Morris, M.~\& 

380: Serabyn, E.\ 1996, \araa, 34, 645 

381: 

382: \bibitem[Mundell, Wrobel, Pedlar, \& Gallimore(2003)]{mwp03} 

383: Mundell, C.~G., Wrobel, J.~M., Pedlar, A., \& Gallimore, J.~F.\ 2003, \apj, 

384: 583, 192 

385: %the nuclear regions of the seyfert galaxy ngc 4151;parsec scale HI 

386: %absorption and the remarkable radio jet

387: 

388: \bibitem[Neugebauer, Graham, Soifer, \& Matthews(1990)]{ngs90} 

389: Neugebauer, G., Graham, J.~R., Soifer, B.~T., \& Matthews, K.\ 1990, 

390: \aj, 99, 1456 

391: %The size of NGC 4151 at 11.2 microns

392: 

393: \bibitem[Oknyanskij et al.(1999)]{olt99} Oknyanskij, V.~M., 

394: Lyuty, V.~M., Taranova, O.~G., \& Shenavrin V.~I. 1999, {\it AL}, 8, 563

395: %IR-optical time correlations

396: 

397: \bibitem[Penston et al.(1974)]{pps74} Penston, M.~V., 

398: Penston, M.~J., Selmes, R.~A., Becklin, E.~E., \& Neugebauer, G.\ 1974, 

399: \mnras, 169, 357 

400: %Broadband optical and infrared observations of Seyfert galaxies

401: 

402: \bibitem[Pier \& Krolik(1992)]{pk92}

403: Pier, E.~A., \& Krolik, J.~H. 1992, \apj, 401, 99

404: %torus models

405: 

406: \bibitem[Quillen et al.(2000)]{qsa00} Quillen, A.~C., Shaked, 

407: S., Alonso-Herrero, A., McDonald, C., Lee, A., Rieke, M.~J., \& Rieke, 

408: G.~H.\ 2000, \apjl, 532, L17 

409: 

410: \bibitem[Radomski et al.(2003)]{rpp03} Radomski, J.~T., Pi{\~ n}a,

411: R.~K., Packham, C., Telesco, C.~M., De Buizer, J.~M., Fisher, R.~S.,

412: \& Robinson, A.\ 2003, \apj, 587, 117

413: 

414: \bibitem[Rieke \& Lebofsky(1981)]{rl81} 

415: Rieke, G.~H., \& Lebofsky, M.~J. 1981, \apj, 250, 87 

416: %spectral components of ngc 4151

417: %report variability

418: 

419: \bibitem[Ruiz et al.(2003)]{ryp03} Ruiz, M., Young, S., 

420: Packham, C., Alexander, D.~M., \& Hough, J.~H.\ 2003, \mnras, 340, 733 

421: %near-IR polarization and modeling of NGC 4151

422: 

423: \bibitem[Sanders et al.(1989)]{san89} Sanders, D.~B., 

424: Phinney, E.~S., Neugebauer, G., Soifer, B.~T., \& Matthews, K.\ 1989, \apj, 

425: 347, 2

426: 

427: \bibitem[Shakura \& Sunyaev(1973)]{ss73} Shakura, N.~I.~\& 

428: Sunyaev, R.~A.\ 1973, \aap, 24, 337 

429: 

430: \bibitem[Ulrich(2000)]{ulr00} Ulrich, M.-H.\ 2000, \aapr, 10, 

431: 135 

432: 

433: \bibitem[Ulvestad, Roy, Colbert, \& Wilson(1998)]{urc98} 

434: Ulvestad, J.~S., Roy, A.~L., Colbert, E.~J.~M., \& Wilson, A.~S.\ 1998, 

435: \apj, 496, 196 

436: 

437: \bibitem[Vasisht et al.(2002)]{vbc02}

438: Vasisht, G., Booth, A., Colavita, M.~M., Johnson, R.~L., Ligon, E.~R.,

439: Moore, J.~D. \& Palmer, D.~L. 2002, \procspie, 4838, 824

440: 

441: \bibitem[Wizinowich et al.(2002)]{wiz02} Wizinowich, P., Le Mignant, D.,

442: Stomski, P., Acton, D.S., Contos, A., \& Neyman, C., 2002, \procspie,

443: 4839, 9

444: %Keck AO instrument paper

445: 

446: 

447: \end{thebibliography}

448: 

449: 

450: \clearpage

451: 

452: \begin{figure}

453: \plotfiddle{f1.eps}{6.5in}{360}{100}{100}{-300}{-120}

454: %\rotatebox{0}{

455: %\plotone{f1.eps}}

456: \caption{

457: The raw (bottom), calibrated (middle) and system visibility (top) for

458: NGC4151.  In the raw data plot, the 5 sec data averages are shown with

459: open circles and the scan averages are shown with filled squares.  The

460: measured visisbility is $0.84\pm0.064$; the error bars in the

461: calibrated data contain a 5\% systematic component.  The hour angles

462: for the averaged points are -0.03, 0.33 and 0.71 hours respectively.

463: The projected baseline length ranges from 81.8 to 83.6 meters and the

464: position angle from 39.3 to 33.9 degrees (E of N).}

465: \label{fig:ifvis}

466: \end{figure} 

467: 

468: \end{document}

469: 

470: