astro-ph0210047/ms.tex
1: \documentclass[preprint]{aastex}
2: %\usepackage{emulateapj5}
3: %\usepackage[]{natbib}
4: %\usepackage{graphics}
5: %\usepackage{apjfonts}
6: \usepackage{psfig}
7: 
8: \newcommand{\Ha}{H$\alpha$}
9: \newcommand{\Hb}{H$\beta$}
10: \newcommand{\HII}{\ion{H}{2}}
11: \newcommand{\OI}{[\ion{O}{1}]}
12: \newcommand{\OII}{[\ion{O}{2}]}
13: \newcommand{\OIII}{[\ion{O}{3}]}
14: \newcommand{\NII}{[\ion{N}{2}]}
15: \newcommand{\SII}{[\ion{S}{2}]}
16: \newcommand{\NeIII}{[\ion{Ne}{3}]}
17: \newcommand{\CaII}{\ion{Ca}{2}}
18: \newcommand{\MgIb}{\ion{Mg}{1}~$b$}
19: \newcommand{\NaID}{\ion{Na}{1}~$D$}
20: \newcommand{\gtwid}{\mathrel{\raise.25ex\hbox{$>$\kern-.80em\lower1ex\hbox{$\sim$}}}}
21: 
22: %\slugcomment{}
23: \shorttitle{``HIDDEN'' SEYFERT 2 GALAXIES}
24: \shortauthors{MORAN ET AL.}
25: 
26: \begin{document}
27: 
28: \title{``Hidden'' Seyfert 2 Galaxies and the X-ray Background}
29: 
30: \author{Edward C.\ Moran}
31: \affil{Astronomy Department, Wesleyan University, Middletown, CT 06459
32:        \\ \smallskip{\rm and}\\}
33: 
34: \author{Alexei V.\ Filippenko and Ryan Chornock}
35: \affil{Department of Astronomy, University of California,
36:                  Berkeley, CA 94720-3411}
37: 
38: \begin{abstract}
39: Obscured active galactic nuclei, which are classified optically as type~2
40: (narrow-line) Seyfert galaxies in the local universe, are by far the most
41: promising candidates for the origin of the hard (2--10 keV) X-ray background
42: radiation.  However, optical follow-up observations of faint X-ray sources
43: in deep {\it Chandra\/} images have revealed surprising numbers of apparently
44: normal galaxies at modest redshift.  Such objects represent $\sim$~40--60\%
45: of the sources classified in deep {\it Chandra\/} surveys, raising the
46: possibility that the X-ray galaxy population has evolved with cosmic time.
47: Alternatively, most of the faint X-ray galaxies in question are so distant
48: that their angular diameters are comparable to the slit widths used in
49: ground-based spectroscopic observations; thus, their nuclear spectral
50: features may be overwhelmed (``hidden'') by host-galaxy light.  To test
51: this hypothesis, we have obtained integrated spectra of a sample of nearby,
52: well-studied Seyfert 2 galaxies.  The data, which accurately simulate
53: observations of distant {\it Chandra\/} sources, demonstrate convincingly
54: that the defining spectral signatures of Seyfert 2s {\it can\/} be hidden
55: by light from their host galaxies.  In fact, 60\% of the observed objects
56: would not be classified as Seyfert 2s on the basis of their integrated
57: spectra, similar to the fraction of faint X-ray sources identified with
58: ``normal'' galaxies.  Thus, the numbers of narrow-line active galaxies in
59: deep {\it Chandra\/} surveys (and perhaps all ground-based spectroscopic
60: surveys of distant galaxies) are likely to have been underestimated.
61: \end{abstract}
62: \keywords{galaxies:~Seyfert --- X-rays:~diffuse background --- X-rays:~galaxies}
63: 
64: 
65: \newpage
66: \section{Introduction}
67: 
68: Identification of the classes of objects responsible for the cosmic X-ray
69: background radiation (XRB) has been one of the fundamental pursuits of X-ray
70: astronomy for the past several decades.  It now appears that different
71: populations produce the majority of the background in different X-ray energy
72: bands.  In the soft 0.5--2 keV range, active galactic nuclei (AGNs) with broad
73: optical emission lines (i.e., Seyfert~1 galaxies and quasars) are the dominant
74: contributors to the XRB (Schmidt et al.\ 1998).  These objects, however,
75: have steep X-ray spectra and cannot explain the flat slope of the XRB spectrum
76: at higher energies.  Type~2 (narrow-line) Seyfert galaxies have thus emerged
77: as the most promising candidates for the origin of the hard (2--10 keV) XRB.
78: The soft X-ray fluxes of these objects are heavily absorbed by dense
79: circumnuclear gas, presumably the same material that obscures their broad
80: emission-line regions.  Models (Madau, Ghisellini, \& Fabian 1994; Comastri
81: et al.\ 1995; Gilli, Salvati, \& Hasinger 2001) and X-ray observations of
82: nearby sources (Moran et al.\ 2001) have demonstrated that Seyfert~2 galaxies
83: have both the space density and X-ray spectral properties necessary to account
84: for the intensity and spectrum of the hard XRB.
85: 
86: The {\it Chandra X-ray Observatory}, which resolves the majority of the XRB
87: in deep exposures, provides an opportunity to confirm the Seyfert~2 hypothesis
88: directly.  Surprisingly, however, follow-up optical observations of faint
89: sources in the deepest {\it Chandra\/} images have revealed a significant
90: population of apparently {\it normal\/} galaxies and far fewer Seyfert 2s
91: than expected (e.g., Mushotzky et al.\ 2000; Barger et al.\ 2001a, 2001b,
92: 2002; Hornschemeier et al.\ 2001).  ``Normal'' galaxies represent $\sim$
93: 40--60\% of the hard X-ray sources classified in deep surveys; many exhibit
94: evidence of absorption in the X-ray band, but emission lines are very weak
95: in their starlight-dominated optical spectra.  The multiwavelength properties
96: of the prototypical example of this class are described by
97: Comastri et al.\ (2002a).  The prevailing interpretation is that optically
98: normal, X-ray--luminous galaxies are common at modest redshifts, perhaps
99: reflecting a tendency for AGNs in the past to be more absorbed or have weaker
100: ionizing continua than present-day Seyfert galaxies (Barger et al.\ 2001a).
101: This hints at the possibility that the X-ray galaxy population has evolved
102: with cosmic~time.
103: 
104: Alternatively, we propose that the results of the deep {\it Chandra\/} surveys
105: may be heavily influenced by the limitations of ground-based observing.  In
106: spectroscopic observations of nearby AGNs, light is collected through a small
107: aperture centered on the nucleus, which excludes most of the starlight from
108: the host galaxy.  The faint sources that produce the hard XRB, in contrast,
109: are very distant (the majority are at redshifts $z \gtwid 0.5$); thus, when
110: they are observed, the entire galaxy (or a large fraction of it) falls within
111: the spectrograph slit.  (At $z \approx 1$, 10 kpc corresponds to
112: $\sim$~1\farcs5.)  Combined with the low spectral resolution that has been
113: employed to date ($\sim$ 12--20 \AA ) and the modest signal-to-noise ratios
114: frequently obtained (the optical counterparts are generally quite faint),
115: the additional galaxy light (from stars and \HII\ regions) could lead
116: to the appearance that some distant {\it Chandra\/} sources are associated
117: with normal galaxies rather than with Seyfert~2s.
118: 
119: To investigate this possibility we have obtained {\it integrated\/} spectra
120: of nearby Seyfert~2 galaxies that are known to be absorbed X-ray sources.
121: Our techniques simulate ground-based spectroscopic observations of distant
122: X-ray galaxies, allowing us to evaluate whether the emission-line signatures
123: of their activity can be overwhelmed (or ``hidden'') in spectra of their
124: integrated light.  This project draws inspiration from a series of papers
125: by Kennicutt (1992a, 1992b), who obtained integrated spectra of nearby
126: galaxies of all types to assist ``the classification of unresolved galaxies
127: at large distances [and] the spectroscopic identification of extended sources
128: detected in infrared, radio, and X-ray surveys'' (Kennicutt 1992a).  Our
129: similarly motivated work focuses on the class of objects believed to be
130: responsible for the hard X-ray background.
131: 
132: 
133: \section{Observations and Analysis}
134: The objects for this study are drawn from the distance-limited sample of
135: Seyfert galaxies compiled by Ulvestad \& Wilson (1989).  There are 31
136: confirmed Seyfert~2s in the sample (Moran et al.\ 2000), almost all of
137: which have been observed in the 1--10 keV band with the {\sl ASCA\/}
138: satellite (Moran et al.\ 2001).  The Ulvestad \& Wilson sample is ideal
139: for this project for several reasons.  First, because of its distance-limited
140: nature, the sample is relatively unbiased.  Second, the hard X-ray
141: luminosities of the objects ($\sim$~$10^{41}$--$10^{43}$ ergs~s$^{-1}$)
142: are comparable to those of the ``normal'' galaxies
143: in deep {\it Chandra\/} surveys (Barger et al.\ 2001a, 2001b; Hornschemeier
144: et al.\ 2001).  And finally, because all of the objects are located within
145: $\sim$~60~Mpc, most are close enough that we can isolate the emission from
146: their nuclei yet distant enough (i.e., with sufficiently small angular
147: diameters) that we can measure their integrated spectra with relative ease.
148: 
149: Our data were acquired with the 3-m Shane reflector at Lick Observatory
150: over the course of three runs in August and November 2001 and February 2002.
151: We employed 600 line mm$^{-1}$ dispersing elements on each arm of the Kast
152: double spectrograph (Miller \& Stone 1993), which provided coverage of the
153: 3400--8100 \AA\ range. For each galaxy, we obtained a 300~s exposure of the
154: nucleus using a $2''$ slit oriented at the parallactic angle (Filippenko
155: 1982).  The nuclear spectra represent $4''$--$5''$ extractions along the
156: slit and have a resolution of 5--6~\AA\ (FWHM).
157: Integrated spectra of each object were obtained by scanning
158: an east-west slit across the galaxy in declination at a constant rate.
159: Six scans were obtained for most galaxies, although a few of the brighter
160: objects were scanned three to five times.  For the scans, the width of
161: the slit was increased to $6''$ to improve our observing efficiency and
162: to degrade the spectral resolution to 15--16~\AA , comparable to that
163: used in observations of faint {\it Chandra\/} sources (Barger et al.\ 
164: 2002; Hornschemeier et al.\ 2001).  The angular distances of the scans
165: ($30''$--$120''$ in declination) were dictated by the diameters of the
166: galaxies in high-contrast images from the Digitized Sky Survey (DSS).
167: Each scan had a duration of 600~s; combined with the slit width and scan
168: distances, this yielded effective exposures of 30~s to 120~s per scan
169: for each object's integrated spectrum.  These spectra include most of the
170: light from each galaxy above the sky background (i.e., $\sim 35''$ to
171: $\sim 90''$ extractions).
172: 
173: We observed a total of 18 objects from our sample --- all those at $\delta
174: > -23^{\circ}$ with projected angular sizes in right ascension (again,
175: estimated from DSS images) of less than $2'$.  The latter restriction
176: ensured that we would be able to perform accurate sky background subtraction
177: with light that entered the slit near the two ends. The galaxies observed
178: and some pertinent data are listed in Table~1.  The optical and X-ray
179: properties of the 13 objects we did not observe do not differ significantly.
180: 
181: 
182: \section{Integrated Spectra of Seyfert 2 Galaxies}
183: Recently published spectroscopic studies of faint {\it Chandra\/} sources are
184: based mainly on data acquired with 8--10~m telescopes (Barger et al.\ 2002;
185: Hornschemeier et al.\ 2001; Tozzi et al.\ 2001).  In most respects, the
186: integrated spectra we have obtained accurately simulate observations of
187: distant type~2 AGNs with these instruments.  The scan distances and
188: extractions employed for our sample include most of the light from the
189: galaxies; they correspond to physical sizes of $\sim$ 8--20 kpc, similar
190: to that which is typically included within a 1.5--$2''$ slit for a distant
191: {\it Chandra\/} source.  Our spectral resolution, as discussed above, is
192: comparable to that used in {\it Chandra\/} source surveys.  In addition,
193: the effective exposure times associated with our scans are equivalent to
194: 1.5--2~hr Keck exposures of objects $\sim$ 7--8 mag fainter, i.e., objects
195: with visual magnitudes in the $m \approx$ 20--22 range.  These roughly
196: correspond to the Keck exposure times and magnitudes of distant
197: {\it Chandra\/} sources that have been classified spectroscopically
198: (Barger et al.\ 2001a, 2001b, 2002).
199: 
200: Examples of our results are displayed in Figure~1.  The nuclear spectrum of
201: NGC 262 (= Mrk 348; lower trace of Fig.~1a) bears all the characteristics
202: of a type~2 Seyfert galaxy: strong \OII\ $\lambda 3727$, \NeIII\ $\lambda
203: 3869$, \Hb\ $\lambda 4861$, \OIII\ $\lambda\lambda 4959,5007$, \OI\ $\lambda
204: 6300$, \Ha\ $\lambda 6563$, \NII\ $\lambda 6584$, and \SII\ $\lambda\lambda
205: 6717,6731$ emission lines with high \OIII /\Hb , \OI /\Ha , \NII /\Ha , and
206: \SII /\Ha\ flux ratios (Veilleux \& Osterbrock 1987). These same features
207: are clearly present in the integrated spectrum of the galaxy (upper trace
208: of Fig.~1a), despite the presence of a significant amount of additional
209: starlight.  Thus, NGC~262 retains its Seyfert~2 classification in the
210: integrated spectrum.
211: 
212: The spectra in Figures 1b--e tell a much different story.  Once again, the
213: nuclear spectra of these galaxies exhibit Seyfert~2 emission-line signatures.
214: In the integrated spectra, however, the lines are far less prominent, and
215: some are completely obliterated.  The strongest spectral features are the
216: \CaII\ $\lambda\lambda 3934,3968$, G band ($\lambda 4300$), \MgIb\ $\lambda
217: 5176$, and Na~I D $\lambda 5893$ stellar absorption lines typically observed
218: in the spectra of normal early-type galaxies.  Note that the \NeIII\ $\lambda
219: 3869$ emission line, which has been used in {\it Chandra\/} source surveys
220: to identify AGNs (Barger et al.\ 2001a, 2001b; Hornschemeier et al.\ 2001),
221: is not visible in the integrated spectra.  Thus, \NeIII\ emission cannot be
222: relied upon to indicate the presence of an active nucleus.
223: 
224: The data for NGC~3982 (Fig.~1f) provide an interesting third alternative.
225: Although the emission lines in both the nuclear and integrated spectra of
226: this object are strong, the line-flux ratios in the two differ dramatically.
227: For example, in the nuclear spectrum, \NII\ $\lambda 6584$/\Ha\ $\approx$ 1
228: and \OIII\ $\lambda 5007$/\Hb\ $\approx$ 16, which is typical for a
229: Seyfert~2.  In the integrated spectrum, however, \NII /\Ha\ = 0.37 and
230: \OIII /\Hb\ = 0.82, consistent with the values observed in \HII\ regions
231: (e.g., Veilleux \& Osterbrock 1987).  The \OI /\Ha\ and \SII /\Ha\ ratios
232: also resemble those of \HII\ regions.  Given only the integrated spectrum
233: of NGC 3982, we would classify it as a starburst galaxy rather than a type~2
234: Seyfert!  We obtain a similar result for another object in our sample, NGC
235: 1667.  Thus, it is possible for the emission from star-forming regions to
236: drown out an active nucleus in a galaxy, which sheds light on the nature of
237: the ``composite'' starburst/Seyfert galaxies uncovered during optical surveys
238: of {\sl ROSAT\/} sources (Moran, Halpern, \& Helfand 1996).
239: 
240: \section{Discussion and Conclusions}
241: The last column of Table 1 indicates our assessment of the classifications
242: of our galaxies based on their integrated spectra.  While seven objects with
243: strong lines in the nucleus remain as Seyfert~2s in the integrated spectra,
244: 11 others (or $\sim$~60\%) would {\it not\/} be classified as AGNs.  This
245: is comparable to the fraction of {\it Chandra\/} sources ($\sim$~40--60\%)
246: identified with ``normal'' galaxies in deep surveys (Barger et al.\ 2001a,
247: 2001b, 2002; Hornschemeier et al.\ 2001).  Among the normal-looking subset
248: of our sample, nine objects resemble early-type galaxies with red continua
249: and strong absorption features, and two objects have spectra reminiscent of
250: starburst (or post-starburst) galaxies.
251: 
252: All of the objects we observed display some evidence of emission lines in
253: their integrated spectra.  While the same is true of all of the ``normal''
254: galaxies whose spectra are presented by Barger et al.\ (2001a, 2001b) and
255: Hornschemeier et al.\ (2001), 19\% of the sources classified by Barger et
256: al.\ (2002) are described as ``absorption-line'' galaxies.  Some of these
257: objects may in fact have faint emission lines --- weak \OII\ and \OIII\
258: emission is visible in the average absorption-line galaxy spectrum shown
259: in Figure~5h of Barger et al.\ (2002).  Others, however, may be devoid
260: of emission features in the data obtained.  Two factors are likely to
261: contribute to the relatively high incidence of absorption-line systems in
262: the {\it Chandra\/} surveys:
263: 
264: \noindent
265: (1) The \Ha\ and \NII\ emission lines can be the strongest lines in the
266: spectra of Seyfert galaxies; as such, they are vital for accurate spectroscopic
267: classifications.  In fact, the classifications of two of the persistent
268: Seyfert~2s in our sample (Mrk~1066 and MCG $+$01-27-020) would be ambiguous
269: without information about the lines near \Ha .  However, $\sim$~70\% of the
270: spectra of supposedly normal galaxies published by Barger et al.\ and
271: Hornschemeier et al.\ do not cover the \Ha\ region because of the high
272: redshifts of the objects.  It is possible that some of the absorption-line
273: galaxies would exhibit emission lines if this region were observed.
274: 
275: \noindent
276: (2) Our relatively small sample of nearby objects does not include all possible
277: types of host galaxies.  In particular, galaxies with high optical luminosities
278: are absent.  Using the expression given in \S~12.5 of Barger et al.\ (2002)
279: and the information listed in Table~1, we find that 14 (78\%) of the objects
280: we observed have $B$-band luminosities below $10^{43}$ ergs~s$^{-1}$, and all
281: have $L_B < 2 \times 10^{43}$ ergs~s$^{-1}$.  In contrast, the majority of the
282: sources spectroscopically classified by Barger et al.\ have $L_B$ in excess of
283: $10^{43}$ ergs~s$^{-1}$.  Compared to the objects in our sample, the integrated
284: light of these high-$L_B$ galaxies could more easily overwhelm the emission
285: from moderately luminous Seyfert nulcei.  Host-galaxy luminosity may also be
286: related to the detection of optically dull, X-ray--bright objects at fairly
287: low redshifts ($z < 0.4$; see Comastri et al.\ 2002b).  The possibility that
288: some host-galaxy light is excluded in the spectra of these objects is probably
289: offset by the fact that their host galaxies are quite luminous --- nine of
290: the ten ``normal'' galaxies in the Comastri et al.\ (2002b) study have $L_B$
291: greater than a few times $10^{43}$ ergs~s$^{-1}$.
292: 
293: Thus, as we have speculated previously (Moran et al.\ 2001), it {\it is\/}
294: possible to ``hide'' the true nature of classical type~2 AGNs in their
295: integrated optical spectra.  Low signal-to-noise ratio and inadequate spectral
296: coverage due to
297: redshift effects can hinder the classification process further.  We conclude,
298: therefore, that the demographics of the distant X-ray galaxy population under
299: investigation in deep {\it Chandra\/} surveys are more likely to be influenced
300: by the limitations of ground-based spectroscopic observations than by some
301: evolution of the population.  The majority of the apparently normal galaxies
302: that have turned up in these surveys, at least those with flat (presumably
303: absorbed) X-ray spectra and/or substantial X-ray luminosities, would probably
304: have optical spectra similar to those of nearby Seyfert~2s if their nuclei
305: could be isolated spatially.  If so, this would verify that Seyfert~2 galaxies
306: are the dominant contributors to the hard XRB, as prior research has suggested.
307: Host galaxy dilution is likely to impact any ground-based spectroscopic survey
308: of distant galaxies, leading to underestimates of the numbers of narrow-line
309: AGNs in such surveys.
310: 
311: \acknowledgments
312: We would like to express our thanks to Bill Holzapfel for generously donating
313: the August 2001 Lick time that enabled us to get this project underway.  This
314: work was supported by NASA grant NAG5-3556 to A.~V.~F.
315: 
316: 
317: \begin{thebibliography}{}
318: 
319: \bibitem{}Barger, A.~J., Cowie, L.~L., Mushotzky, R.~F., \& Richards, E.~A.\
320:           2001a, AJ, 121, 662
321: \bibitem{}Barger, A.~J., et al.\ 2001b, AJ, 122, 2177
322: \bibitem{}Barger, A.~J., et al.\ 2002, AJ, in press
323: \bibitem{}Comastri, A., et al.\ 2002b, in ``New Visions of the X-ray Universe
324: in the XMM-Newton and Chandra Era,'' in press (astro-ph/0203019)
325: \bibitem{}Comastri, A., et al.\ 2002a, ApJ, 575, 137
326: \bibitem{}Comastri, A., Setti, G., Zamorani, G. \& Hasinger, G.\ 1995,
327:           A\&A, 296, 1
328: \bibitem{}Filippenko, A.~V.\ 1982, PASP, 94, 715
329: \bibitem{}Gilli, R., Salvati, M., \& Hasinger, G.\ 2001, A\&A, 366, 407
330: \bibitem{}Hornschemeier, A.~E., et al.\ 2001, ApJ, 554, 742
331: \bibitem{}Kennicutt, R.~C.\ 1992a, ApJ, 388, 310
332: \bibitem{}Kennicutt, R.~C.\ 1992b, ApJS, 79, 255
333: \bibitem{}Madau, P., Ghisellini, G., \& Fabian, A.~C.\ 1994, MNRAS, 270, L17
334: \bibitem{}Miller, J.~S., \& Stone, R.~P.~S.\ 1993, Lick Observatory Tech.\
335:           Rep., No.~66
336: \bibitem{}Moran, E.~C., Barth, A.~J., Kay, L.~E., \& Filippenko, A.~V.\
337:           2000, ApJ, 540, L73
338: \bibitem{}Moran, E.~C., Halpern, J.~P., \& Helfand, D.~J.\ 1996, ApJS, 106, 341
339: \bibitem{}Moran, E.~C., Kay, L.~E., Davis, M., Filippenko, A.~V., \&
340:           Barth, A.~J.\ 2001, ApJ, 556, L75
341: \bibitem{}Mushotzky, R.~F., Cowie, L.~L., Barger, A.~J., \& Arnaud, K.~A.\
342:           2000, Nature, 404, 459
343: \bibitem{}Schmidt, M., et al.\ 1998, A\&A, 329, 495
344: \bibitem{}Tozzi, P., et al.\ 2001, ApJ, 562, 42
345: \bibitem{}Tully, R.~B.\ 1988, Nearby Galaxies Catalog (Cambridge:
346:           Cambridge Univ.\ Press)
347: \bibitem{}Ulvestad, J.~S., \& Wilson, A.~S.\ 1989, ApJ, 343, 659
348: \bibitem{}Veilleux, S., \& Osterbrock, D.~E.\ 1987, ApJS, 63, 295
349: 
350: \end{thebibliography}
351: 
352: 
353: \begin{center}
354: \begin{deluxetable}{lcccc}
355: \tablewidth{0pt}
356: \tablecaption{Seyfert 2 Galaxies Observed}
357: \tablehead{\colhead{Galaxy} &
358:            \colhead{$d$ (Mpc)\tablenotemark{a}} &
359:            \colhead{$m$\tablenotemark{b}} &
360:            \colhead{Exp.\ (s)\tablenotemark{c}} &
361:            \colhead{Class.\tablenotemark{d}}}
362: \startdata
363: MCG $+$01-27-020         &  46.8  & 14.7 & 120~ & Sy2\\
364: Mrk 3                    &  54.0  & 13.6 & 60   & Sy2\\
365: Mrk 1066                 &  48.1  & 13.6 & 82   & Sy2\\
366: NGC 262                  &  60.1  & 13.9 & 75   & Sy2\\
367: NGC 591                  &  60.7  & 13.9 & 120~ & Sy2\\
368: NGC 788                  &  54.4  & 13.1 & 60   & N/e\\
369: NGC 1358                 &  53.8  & 12.7 & 30   & N/e\\
370: NGC 1667                 &  60.7  & 13.1 & 60   & ~~N/sb\\
371: NGC 1685                 &  60.4  & 14.5 & 60   & N/e\\
372: NGC 2273                 &  28.4  & 12.3 & 72   & N/e\\
373: NGC 3081                 &  32.5  & 12.8 & 40   & Sy2\\
374: NGC 3982                 &  27.2  & 11.8 & 36   & ~~N/sb\\
375: NGC 4117                 &  17.0  & 14.0 & 48   & N/e\\
376: NGC 5283                 &  41.4  & 14.2 & 90   & Sy2\\  
377: NGC 5347                 &  36.7  & 13.2 & 60   & N/e\\
378: NGC 5695                 &  56.4  & 13.8 & 60   & N/e\\
379: NGC 5929\tablenotemark{e}&  38.5  & 13.0 & 40   & N/e\\
380: NGC 7672                 &  53.5  & 14.4 & 86   & N/e
381: \enddata
382: \tablenotetext{a}{Distances from Tully (1988).  For the more distant 
383:                   sources ($cz > 3000$ km~s$^{-1}$), we assume
384:                   $d = cz/H_0$, where $H_0$ = 75 km~s$^{-1}$ Mpc$^{-1}$.}
385: \tablenotetext{b}{Optical $B$ band magnitude, from the NASA Extragalactic
386:                   Database (NED).}
387: \tablenotetext{c}{Effective exposure in the scanned observations.}
388: \tablenotetext{d}{Classification based on the integrated spectrum.
389:                   Sy2 = Seyfert 2; N/e = normal galaxy with early-type
390:                   continuum; N/sb = normal galaxy with starburst emission
391:                   lines.}
392: \tablenotetext{e}{Interacting with NGC 5930, which is included in the
393:                   integrated spectrum.}
394: \end{deluxetable}
395: \end{center}
396: 
397: 
398: \begin{figure}
399: \begin{center}
400: \epsscale{1.0}
401: \centerline{\plotone{f1.eps}}
402: %\vskip 0.1truein
403: \caption{Data for six of the observed Seyfert 2 galaxies.  Both nuclear
404: (lower trace) and integrated (upper trace) spectra are shown in each panel.
405: Relative flux densities in units of ergs cm$^{-2}$ s$^{-1}$ \AA $^{-1}$ are
406: plotted on the ordinate; the dotted line indicates the zero flux level.
407: For clarity, the nuclear spectrum of each object has been multiplied by a
408: constant, which appears following the galaxy name.}
409: \end{center}
410: \end{figure}
411: 
412: 
413: \end{document}
414: 
415: 
416: \noindent
417: (3) For a number of the absorption-line galaxies in the Barger et al.\ (2002)
418: sample, it appears that the wavelength of an important emission line
419: (sometimes the only useful line in the observed bandpass) coincides with an
420: atmospheric absorption band or falls near one end of the bandpass where the
421: $S/N$ ratio is low.  These effects could also inhibit the detection of
422: emission lines in some objects.
423: 
424: