astro-ph0008093/ms.tex
1: % mnsample.tex
2: %
3: % v1.2 released 5th September 1994 (M. Reed)
4: % v1.1 released 18th July 1994
5: % v1.0 released 28th January 1994
6: 
7: \documentstyle[psfig]{mn}
8: 
9: % If your system has the AMS fonts version 2.0 installed, MN.sty can be
10: % made to use them by uncommenting the line: %\AMStwofontstrue
11: %
12: % By doing this, you will be able to obtain upright Greek characters.
13: % e.g. \umu, \upi etc.  See the section on "Upright Greek characters" in
14: % this guide for further information.
15: %
16: % If you are using AMS 2.0 fonts, bold math letters/symbols are available
17: % at a larger range of sizes for NFSS release 1 and 2 (using \boldmath or
18: % preferably \bmath).
19: 
20: \newif\ifAMStwofonts
21: %\AMStwofontstrue
22: 
23: %%%%% AUTHORS - PLACE YOUR OWN MACROS HERE %%%%%
24: 
25: 
26: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
27: 
28: \ifoldfss
29:   %
30:   \newcommand{\rmn}[1] {{\rm #1}}
31:   \newcommand{\itl}[1] {{\it #1}}
32:   \newcommand{\bld}[1] {{\bf #1}}
33:   %
34:   \ifCUPmtlplainloaded \else
35:     \NewTextAlphabet{textbfit} {cmbxti10} {}
36:     \NewTextAlphabet{textbfss} {cmssbx10} {}
37:     \NewMathAlphabet{mathbfit} {cmbxti10} {} % for math mode
38:     \NewMathAlphabet{mathbfss} {cmssbx10} {} %  "   "    "
39:   \fi
40:   %
41:   \ifAMStwofonts
42:   %
43:     \ifCUPmtlplainloaded \else
44:       \NewSymbolFont{upmath} {eurm10}
45:       \NewSymbolFont{AMSa} {msam10}
46:       \NewMathSymbol{\upi}     {0}{upmath}{19}
47:       \NewMathSymbol{\umu}     {0}{upmath}{16}
48:       \NewMathSymbol{\upartial}{0}{upmath}{40}
49:       \NewMathSymbol{\leqslant}{3}{AMSa}{36}
50:       \NewMathSymbol{\geqslant}{3}{AMSa}{3E}
51:       \let\oldle=\le     \let\oldleq=\leq
52:       \let\oldge=\ge     \let\oldgeq=\geq
53:       \let\leq=\leqslant \let\le=\leqslant
54:       \let\geq=\geqslant \let\ge=\geqslant
55:     \fi
56:     %
57:   \fi
58: %
59: \fi % End of OFSS
60: 
61: \ifnfssone
62:   %
63:   \newmathalphabet{\mathit}
64:   \addtoversion{normal}{\mathit}{cmr}{m}{it}
65:   \addtoversion{bold}{\mathit}{cmr}{bx}{it}
66:   %
67:   \newcommand{\rmn}[1] {\mathrm{#1}}
68:   \newcommand{\itl}[1] {\mathit{#1}}
69:   \newcommand{\bld}[1] {\mathbf{#1}}
70:   %
71:   \def\textbfit{\protect\txtbfit}
72:   \def\textbfss{\protect\txtbfss}
73:   \long\def\txtbfit#1{{\fontfamily{cmr}\fontseries{bx}\fontshape{it}%
74:     \selectfont #1}}
75:   \long\def\txtbfss#1{{\fontfamily{cmss}\fontseries{bx}\fontshape{n}%
76:     \selectfont #1}}
77:   %
78:   \newmathalphabet{\mathbfit} % math mode version of \textbfit{..}
79:   \addtoversion{normal}{\mathbfit}{cmr}{bx}{it}
80:   \addtoversion{bold}{\mathbfit}{cmr}{bx}{it}
81:   %
82:   \newmathalphabet{\mathbfss} % math mode version of \textbfss{..}
83:   \addtoversion{normal}{\mathbfss}{cmss}{bx}{n}
84:   \addtoversion{bold}{\mathbfss}{cmss}{bx}{n}
85:   %
86:   \ifAMStwofonts
87:     %
88:     \ifCUPmtlplainloaded \else
89:       %
90:       % Make NFSS 1 use the extra sizes available for bold math italic and
91:       % bold math symbol. These definitions may already be loaded if your
92:       % NFSS format was built with fontdef.max.
93:       %
94:       \UseAMStwoboldmath
95:       %
96:       \makeatletter
97:       \new@mathgroup\upmath@group
98:       \define@mathgroup\mv@normal\upmath@group{eur}{m}{n}
99:       \define@mathgroup\mv@bold\upmath@group{eur}{b}{n}
100:       \edef\UPM{\hexnumber\upmath@group}
101:       %
102:       \new@mathgroup\amsa@group
103:       \define@mathgroup\mv@normal\amsa@group{msa}{m}{n}
104:       \define@mathgroup\mv@bold\amsa@group{msa}{m}{n}
105:       \edef\AMSa{\hexnumber\amsa@group}
106:       \makeatother
107:       %
108:       \mathchardef\upi="0\UPM19
109:       \mathchardef\umu="0\UPM16
110:       \mathchardef\upartial="0\UPM40
111:       \mathchardef\leqslant="3\AMSa36
112:       \mathchardef\geqslant="3\AMSa3E
113:       %
114:       \let\oldle=\le     \let\oldleq=\leq
115:       \let\oldge=\ge     \let\oldgeq=\geq
116:       \let\leq=\leqslant \let\le=\leqslant
117:       \let\geq=\geqslant \let\ge=\geqslant
118:       %
119:     \fi
120:   \fi
121:   %
122: \fi % End of NFSS release 1
123: 
124: \ifnfsstwo
125:   %
126:   \newcommand{\rmn}[1] {\mathrm{#1}}
127:   \newcommand{\itl}[1] {\mathit{#1}}
128:   \newcommand{\bld}[1] {\mathbf{#1}}
129:   %
130:   \def\textbfit{\protect\txtbfit}
131:   \def\textbfss{\protect\txtbfss}
132:   \long\def\txtbfit#1{{\fontfamily{cmr}\fontseries{bx}\fontshape{it}%
133:     \selectfont #1}}
134:   \long\def\txtbfss#1{{\fontfamily{cmss}\fontseries{bx}\fontshape{n}%
135:     \selectfont #1}}
136:   %
137:   \DeclareMathAlphabet{\mathbfit}{OT1}{cmr}{bx}{it}
138:   \SetMathAlphabet\mathbfit{bold}{OT1}{cmr}{bx}{it}
139:   \DeclareMathAlphabet{\mathbfss}{OT1}{cmss}{bx}{n}
140:   \SetMathAlphabet\mathbfss{bold}{OT1}{cmss}{bx}{n}
141:   %
142:   \ifAMStwofonts
143:     %
144:     \ifCUPmtlplainloaded \else
145:       \DeclareSymbolFont{UPM}{U}{eur}{m}{n}
146:       \SetSymbolFont{UPM}{bold}{U}{eur}{b}{n}
147:       \DeclareSymbolFont{AMSa}{U}{msa}{m}{n}
148:       \DeclareMathSymbol{\upi}{0}{UPM}{"19}
149:       \DeclareMathSymbol{\umu}{0}{UPM}{"16}
150:       \DeclareMathSymbol{\upartial}{0}{UPM}{"40}
151:       \DeclareMathSymbol{\leqslant}{3}{AMSa}{"36}
152:       \DeclareMathSymbol{\geqslant}{3}{AMSa}{"3E}
153:       %
154:       \let\oldle=\le     \let\oldleq=\leq
155:       \let\oldge=\ge     \let\oldgeq=\geq
156:       \let\leq=\leqslant \let\le=\leqslant
157:       \let\geq=\geqslant \let\ge=\geqslant
158:       %
159:     \fi
160:   \fi
161:   %
162: \fi % End of NFSS release 2
163: 
164: \ifCUPmtlplainloaded \else
165:   \ifAMStwofonts \else % If no AMS fonts
166:     \def\upi{\pi}
167:     \def\umu{\mu}
168:     \def\upartial{\partial}
169:   \fi
170: \fi
171: 
172: %nearby.tex 
173:  
174: \title{The 3.4 micron absorption feature towards three obscured 
175: active galactic nuclei
176: }
177: \author[Imanishi]
178:        {M.~Imanishi $^{1,2}$ \\
179: $^1$ National Astronomical Observatory, Mitaka, Tokyo 181-8588, Japan \\
180: $^2$ Institute for Astronomy, University of Hawaii, 
181: 2680 Woodlawn Drive, Honolulu, Hawaii 96822, USA\\
182: } 
183: \date{} 
184: 
185: %\pagerange{\pageref{firstpage}--\pageref{lastpage}}
186: \pubyear{2000}
187: 
188: \begin{document} 
189: 
190: \maketitle
191: 
192: %\label{firstpage}
193: 
194: \begin{abstract} 
195: 
196: The results of 3--4 $\mu$m spectroscopy towards the nuclei of NGC 3094, 
197: NGC 7172, and NGC 7479 are reported.  
198: In ground-based 8--13 $\mu$m spectra, all the sources have strong 
199: absorption-like features at $\sim$10 $\mu$m, but they do not have 
200: detectable polycyclic aromatic hydrocarbon (PAH) emission features.  
201: The 3.4 $\mu$m carbonaceous dust absorption features are detected 
202: towards all nuclei.
203: NGC 3094 shows a detectable 3.3 $\mu$m PAH emission feature, while 
204: NGC 7172 and NGC 7479 do not.
205: Nuclear emission whose spectrum shows dust absorption features but 
206: no PAH emission features is thought to dominated by 
207: highly obscured active galactic nuclei (AGNs) activity.
208: For NGC 7172, NGC 7479, and three other such nuclei in the literature, 
209: we investigate the optical depth ratios between the 3.4 $\mu$m 
210: carbonaceous dust and 9.7 $\mu$m silicate dust absorption 
211: ($\tau_{3.4}$/$\tau_{9.7}$).
212: The $\tau_{3.4}$/$\tau_{9.7}$ ratios towards three highly obscured AGNs 
213: with face-on host galaxies are systematically larger than the ratios 
214: in the Galactic diffuse interstellar medium or the ratios for 
215: two highly obscured AGNs with edge-on host galaxies.
216: We suggest that the larger ratios can be explained if the 
217: obscuring dust is so close to the central AGNs that 
218: a temperature gradient occurs in it.
219: If this idea is correct, our results may provide spectroscopic 
220: evidence for the presence of the putative ``dusty tori'' 
221: in the close vicinity of AGNs.
222: 
223: \end{abstract} 
224:  
225: \begin{keywords}
226: galaxies: active --- galaxies: nuclei --- galaxies: 
227: individual: NGC 3094, NGC 7172, and NGC 7479 --- infrared: galaxies
228: \end{keywords} 
229: 
230: \section{INTRODUCTION} 
231: This is the second paper in a series that investigates the 3.4 $\mu$m
232: carbonaceous dust absorption feature in sources that show a strong
233: absorption-like feature at $\sim$10 $\mu$m but no detectable
234: polycyclic aromatic hydrocarbon (PAH) emission features in
235: ground-based 8--13 $\mu$m (i.e., whole the $N$-band) spectra.
236: The first paper discusses NGC 5506 (Imanishi 2000).
237: Roche et al. (1991) performed 
238: extensive ground-based 8--13 $\mu$m spectroscopy of the nuclei of 
239: nearby active galaxies (mostly $z$ $<$ 0.05),
240: and classified them into the following three groups:
241: (1) those dominated by the family of emission features from small
242: molecules called PAHs (20--40 carbon atoms per molecule; Allamandola,
243: Tielens \& Barker 1989),
244: (2) those with a featureless continuum, and 
245: (3) those that display an absorption-like feature at $\sim$10 $\mu$m.
246: Each group of sources is regarded as, respectively, 
247: (1) those powered by star-forming activity, 
248: (2) those powered by unobscured active galactic nuclei (AGNs) activity, 
249: and 
250: (3) those powered by obscured AGN activity, for which the silicate dust 
251: absorption feature at $\sim$9.7 $\mu$m is expected to be present.
252: Sources in the third group, in particular those with a strong silicate 
253: dust absorption feature, are expected to be highly obscured AGNs, 
254: and thus can be used to investigate the properties of the intervening 
255: medium along our line-of-sight through the study of infrared absorption 
256: features.
257: 
258: However, the attribution of the strong absorption-like feature 
259: at $\sim$10 $\mu$m detected in many ground-based 8--13 $\mu$m spectra 
260: to the 9.7 $\mu$m silicate dust absorption has been 
261: questioned based on the {\it Infrared Space Observatory} ({\it ISO}) 
262: spectra (Clavel et al. 1999; Genzel et al. 1998).
263: Strong PAH emission at 7.7 $\mu$m from star-forming activity and 
264: continuum flux that increases with wavelength 
265: could produce an apparent strong absorption-like feature at 
266: $\sim$10 $\mu$m. 
267: One can attempt to estimate the strength of the 7.7 $\mu$m PAH emission 
268: based on the observed flux of the PAH emission at 8.6 $\mu$m and 
269: 11.3 $\mu$m in ground-based 8--13 $\mu$m spectra 
270: (Dudley 1999) by assuming that the 7.7 $\mu$m PAH emission 
271: is weak in sources with no detectable 8.6 $\mu$m and 11.3 $\mu$m PAH 
272: emission. 
273: However, this estimate could sometimes be uncertain 
274: (Rigopoulou et al. 1999).
275: 
276: Spectroscopy at 3--4 $\mu$m can be a powerful method for finding 
277: a highly obscured AGN, 
278: not only because the effect of extinction is small at 3--4 
279: $\mu$m, but also because there exist emission and absorption features
280: for distinguishing between highly obscured AGN and star-forming 
281: activities.
282: If a source is powered by star-forming activity, PAH emission features
283: should be detected at 3.3 $\mu$m and sometimes at 3.4 $\mu$m 
284: (Tokunaga et al. 1991).  
285: If a source is powered by
286: obscured AGN activity, a 3.4 $\mu$m absorption feature associated with
287: carbonaceous dust grains (Pendleton et al. 1994) should be detected.
288: An advantage of a ground-based 3--4 $\mu$m spectrum over a
289: ground-based 8--13 $\mu$m spectrum is that there is no serious
290: uncertainty in determining a continuum level for the former. This is
291: because both the long and short wavelength sides of the features at
292: 3.3--3.4 $\mu$m are observable in the $L$-band atmospheric window 
293: (2.8--4.2 $\mu$m) for a nearby ({\it z} $<$ 0.18) source.  Furthermore, 
294: if both (less obscured) star-forming activity and obscured AGN activity 
295: are associated in a galaxy, the 3.4 $\mu$m absorption feature is 
296: suppressed 
297: because the observed spectrum is dominated by emission from the less 
298: obscured star-forming activity and/or the emission at 3.4 $\mu$m 
299: by PAHs could veil the absorption feature.  
300: Therefore, sources with a strong 3.4 $\mu$m absorption feature that 
301: also lacks a PAH emission feature are the ones most likely to be 
302: highly obscured AGNs with little star-forming activity.  
303: Unfortunately, high-quality 3--4 $\mu$m spectra for sources with a 
304: strong absorption-like feature at
305: $\sim$10 $\mu$m but no detectable PAH emission features in
306: ground-based 8--13 $\mu$m spectra are available only for a few
307: objects.  Examples include NGC 1068 (Bridger, Wright \& Geballe 1994;
308: Imanishi et al. 1997), IRAS 08572+3915 
309: (Pendleton 1996; Wright et al. 1996) and NGC 5506
310: (Imanishi 2000).
311: 
312: Here, we have performed 3--4 $\mu$m spectroscopy on three such sources, 
313: with the aim of detecting the 3.4 $\mu$m carbonaceous dust
314: absorption feature.
315: 
316: 
317: \section{TARGETS}
318: 
319: \subsection{NGC 3094}
320: 
321: NGC 3094 ({\it z} = 0.008) is a spiral galaxy (Huang et al. 1996).
322: Its nuclear optical spectrum is classified as an AGN-type 
323: (Armus, Heckman \& Miley 1989). 
324: A ground-based 8--13 $\mu$m spectrum taken
325: with a 3$\farcs$8 aperture displays a strong absorption-like feature
326: at $\sim$10 $\mu$m, but no detectable PAH emission features (Roche et
327: al. 1991).
328: 
329: \subsection{NGC 7172}
330: 
331: NGC 7172 ({\it z} = 0.009) is an edge-on galaxy of type S0--Sa with a
332: clear equatorial dust lane along roughly the east-west direction
333: (Anupama et al. 1995).  An obscured AGN was found in the hard X-ray
334: (Piccinotti et al. 1982).  A ground-based 8--13 $\mu$m spectrum taken
335: with a 4$\farcs$2 aperture displays a strong absorption-like feature
336: at $\sim$10 $\mu$m, but no detectable PAH emission features (Roche et
337: al. 1991).
338: 
339: \subsection{NGC 7479}
340: 
341: NGC 7479 ({\it z} = 0.008) is a spiral galaxy of type SBbc (Sandage \&
342: Tammann 1987).  Its grand design spiral arms are clearly seen in the
343: Digitized Sky Survey image, and therefore the viewing direction towards
344: the galaxy is estimated to be far from edge-on (Ma, Peng \& Gu 1998).
345: Recent high-quality optical to near-infrared spectra suggest the
346: presence of an obscured AGN in NGC 7479 (Ho, Filippenko \& Sargent
347: 1997; Larkin et al. 1998).  A ground-based 8--13 $\mu$m spectrum taken
348: with a 4$\farcs$3 aperture displays a strong absorption-like feature
349: at $\sim$10 $\mu$m, but no detectable PAH emission features (Roche et
350: al. 1991).
351: 
352: \section{OBSERVATION AND DATA ANALYSIS}  
353: 
354: The 3--4 $\mu$m spectra towards NGC 3094, NGC 7172, and NGC 7479 were 
355: obtained at the 3.8-m United Kingdom Infrared Telescope (UKIRT) 
356: on Mauna Kea, Hawaii, using the cooled grating spectrometer 
357: (CGS4; Mountain et al. 1990).  
358: An observing log is summarized in Table 1.
359: The sky was photometric throughout the observations.  
360: The detector was a 256$\times$256 InSb array.  
361: A 40 l mm$^{-1}$ grating with a 2 pixel wide slit (= 1$\farcs$2) was used.
362: 
363: Spectra were obtained towards the flux peak at $\sim$3.5 $\mu$m.  
364: The slit position angles were set along the east-west direction 
365: for NGC 3094, and along the north-south direction for NGC 7172 
366: and NGC 7479. 
367: A standard nodding technique along the slit direction with an 
368: amplitude of 11$\farcs$6 was employed to subtract the signal
369: from the sky.  
370: The air masses during the observations and total on-source integration 
371: time are summarized in Table 1.
372: Standard stars (Table 1) were observed with almost the same 
373: air masses as the target objects to correct for the 
374: transmission of Earth's atmosphere.  
375: The main features in the spectra to be investigated were 
376: the 3.3 $\mu$m PAH emission and the 3.4 $\mu$m
377: carbonaceous dust absorption features.  
378: Type-A rather than type-G stars were chosen as standard stars, 
379: in order to avoid possible small effects of stellar photospheric 
380: OH absorption at 3.0--3.6 $\mu$m in late-type stars 
381: (Smith, Sellgren \& Brooke 1993).
382: 
383: Standard data analysis procedures were employed using IRAF 
384: %-------------
385: \footnote{
386: IRAF is distributed by the National Optical Astronomy Observatories, 
387: which are operated by the Association of Universities for Research 
388: in Astronomy, Inc. (AURA), under cooperative agreement with the 
389: National Science Foundation. 
390: }.  
391: %-------------
392: After the obtained frames were bias subtracted, they were
393: divided by a flat image.  The spectra of the targets and the standard
394: stars were extracted using an optimal extraction algorithm.
395: Wavelength calibration was performed using a Krypton or Argon lamp.  
396: The signals of the targets were divided by those of the standard stars 
397: and then multiplied by the spectrum of the blackbody with a 
398: temperature corresponding to individual standard stars (Table 1) 
399: to obtain the final spectra.
400: 
401: As a consequence of selecting type-A stars as standard stars, the
402: resulting spectra are affected by strong stellar absorption features
403: at 3.30 $\mu$m (Pf$\delta$) and at 3.74 $\mu$m (Pf$\gamma$).  Since
404: all targets are nearby, the (relatively narrow) Pf emission lines are
405: affected by these absorption features, and therefore will not be
406: discussed.  When making the final spectra, the data at 3.31--3.32
407: $\mu$m have been removed, because the signals there are extremely
408: small due to the strong methane absorption by Earth's atmosphere.
409: Since the data at 3.45--3.46 $\mu$m and 3.49--3.50 $\mu$m for 
410: NGC 7172 and NGC 7479 are affected by large noise levels 
411: that are difficult to correct for completely,
412: these data points have also been removed.  Since the 3.3 $\mu$m
413: PAH emission and the 3.4 $\mu$m carbonaceous dust absorption features
414: have broad spectral profiles, their discussion is not seriously affected 
415: by the removal of these data points.
416: 
417: \section{RESULTS}
418: 
419: \subsection{NGC 3094}
420: 
421: A flux-calibrated spectrum of NGC 3094 is shown in Fig. 1.
422: Our data give an $L$-band magnitude of 8.4 mag (1$\farcs$2
423: $\times$ 5$''$ aperture), which is almost consistent to the 
424: photometric magnitude at $L'$ measured with a 5$''$ aperture 
425: (Dudley 1998; $L'$ = 129.8 mJy or 8.2 mag).
426: 
427: In Fig. 1, a broad absorption-like feature is recognizable at
428: $\sim$3.35--3.55 $\mu$m.  
429: We interpret the feature as the 3.4 $\mu$m carbonaceous dust absorption.
430: In the Galaxy, the wavelength range of the 3.4 $\mu$m carbonaceous
431: dust absorption feature is 3.33--3.55 $\mu$m in the rest-frame
432: (Pendleton et al. 1994), which corresponds to 3.36--3.58 $\mu$m at the
433: redshift of {\it z} = 0.008.  The data points at $>$3.58 $\mu$m are
434: therefore not affected by this absorption feature.  To avoid possible
435: contamination by the Pf$\gamma$ emission line of NGC 3094, the data
436: points at 3.58--3.72 $\mu$m are regarded as the continuum level on the
437: long wavelength side of the absorption feature.  On the short
438: wavelength side of the absorption feature, the data points at $<$3.26
439: $\mu$m are regarded as the continuum, because they are affected
440: neither by the possible 3.3 $\mu$m PAH emission feature (3.25--3.35
441: $\mu$m in the rest-frame; Tokunaga et al. 1991) nor by the Pf$\delta$
442: emission line.  The best linear fit to the data points at $<$3.26
443: $\mu$m and at 3.58--3.72 $\mu$m is adopted as the continuum line for
444: the 3.4 $\mu$m carbonaceous dust absorption feature.  The adopted
445: continuum level is displayed in Fig. 1 as a solid line.  The resulting
446: optical depth relative to this continuum level is $\tau_{3.4}$ $\sim$
447: 0.04 near the absorption peak.  There may be some uncertainties in
448: determining the continuum level.  
449: We also tried to determine the continuum level
450: by varying the wavelength range slightly,
451: and the resulting values for
452: $\tau_{3.4}$ were within 10\% of the value derived above for any
453: reasonable continuum levels adopted.
454: 
455: In Fig. 1, a broad emission-like feature is found at $\sim$3.3 $\mu$m.
456: This feature is interpreted as the 3.3 $\mu$m PAH emission. 
457: Its rest-frame equivalent width is estimated to be 
458: 3.5$\pm$0.2 $\times$ 10$^{-3}$ $\mu$m.
459: 
460: \subsection{NGC 7172}
461: 
462: A flux-calibrated spectrum of NGC 7172 is shown in Fig. 2.
463: Our data give an $L$-band magnitude of 9.1 mag (1$\farcs$2
464: $\times$ 5$''$ aperture), which is fainter than the values of 8.0--9.0
465: mag measured with 5$''-$8$''$ apertures (Sharples et al. 1984; Glass
466: \& Moorwood 1985; Lawrence et al. 1985; Kotilainen et al. 1992).
467: The remaining $L$-band flux may originate from extended stellar
468: emission roughly along the east-west direction, but a quantitative
469: discussion of the origin of the $L$-band emission is difficult due to
470: the large time variation of the $L$-band flux (Sharples et al. 1984).
471: 
472: In Fig. 2, a broad absorption-like feature is recognizable at
473: $\sim$3.35--3.55 $\mu$m.  
474: We regard the feature as the 3.4 $\mu$m carbonaceous dust absorption.
475: A continuum level is determined by finding the best fitted line to the
476: data points at $<$3.26 $\mu$m and at 3.58--3.72 $\mu$m, as was done 
477: for NGC 3094.  
478: The adopted continuum level is shown in Fig. 2 as a solid
479: line.  The $\tau_{3.4}$ against this continuum level is $\sim$0.08.
480: Any other reasonable continuum levels provide consistent values 
481: for $\tau_{3.4}$ within 20\%. 
482: 
483: In this figure, the presence of the 3.3 $\mu$m PAH emission
484: feature is less clear than in the spectrum of NGC 3094.
485: At the redshift of {\it z} = 0.009, the wavelength of the peak of the
486: 3.3 $\mu$m PAH emission feature ($\sim$3.29 $\mu$m in the rest-frame; 
487: Tokunaga et al. 1991) is redshifted to $\sim$3.32 $\mu$m, 
488: where the signals are 
489: extremely small due to the strong methane absorption by Earth's
490: atmosphere.  Hence, it is difficult to quantitatively estimate the 3.3
491: $\mu$m PAH emission flux in this spectrum.  
492: We tentatively estimate the rest-frame equivalent width of the 
493: 3.3 $\mu$m PAH emission as $<$1.8 $\times$ 10$^{-3}$ $\mu$m.
494: 
495: It should be noted that Moorwood (1986) reported the detection of the
496: 3.3 $\mu$m PAH emission feature with a 7$\farcs$5 aperture, although
497: the significance level of the detection was $<$3 $\sigma$.  
498: Moorwood's aperture size is larger than
499: that of Roche et al.  (1991; 4$\farcs$2) and of the present data
500: (1$\farcs$2 $\times$ 5$''$).  The radio map of this galaxy shows
501: extended emission in addition to unresolved emission (Unger et
502: al. 1987).  Since the direction of the extended radio emission is
503: roughly in the east-west direction, the same as that of the edge-on
504: host galaxy (Sharples et al. 1984; Anupama et al. 1995), 
505: the extended radio emission is
506: thought to originate from star-forming regions in the host galaxy.
507: These star-forming regions could explain the possible detection of the
508: 3.3 $\mu$m PAH emission by Moorwood (1986) using a large aperture.
509: 
510: \subsection{NGC 7479}
511: 
512: A flux-calibrated spectrum of NGC 7479 is shown in Fig. 3.  
513: Our data give {\it L} $=$ 10.5 mag.  
514: Nearly all of the {\it L}-band flux within a 5$''$ aperture 
515: ({\it L} = 10.4 mag; Willner et al. 1985) is detected in the 
516: spectrum (1$\farcs$2 $\times$ 5$''$ aperture).
517: 
518: An absorption-like feature at $\sim$3.35--3.55 $\mu$m is recognizable 
519: in the spectrum in Fig. 3.  
520: This feature is interpreted as the 3.4 $\mu$m carbonaceous dust 
521: absorption.
522: A continuum level is determined in the same way as was done for
523: NGC 3094 and NGC 7172.  
524: The adopted continuum level is shown in Fig. 3 as a solid
525: line.  The $\tau_{3.4}$ against this continuum level is $\sim$0.25.
526: 
527: The presence of the 3.3 $\mu$m PAH emission is not clear.
528: The rest-frame equivalent width of the 3.3 $\mu$m PAH emission 
529: is estimated to be $<$7.2 $\times$ 10$^{-3}$ $\mu$m.
530: 
531: \section{DISCUSSION} 
532: 
533: \subsection{Relation between the $\tau_{3.4}$/$\tau_{9.7}$ ratios 
534: and the inclination of the host galaxies}
535: 
536: In general, emission at $>$3 $\mu$m from the inner few arcsec of 
537: (moderately highly-luminous) obscured AGNs is  
538: dominated by a compact AGN component and not by extended stellar 
539: emission (Alonso-Herrero et al. 1998).
540: In this case, the observed optical depths of dust absorption features 
541: reflect the column density of dust in front of a background emission 
542: associated with AGN activity.
543: 
544: In the spectrum of NGC 3094, both the 3.4 $\mu$m carbonaceous 
545: dust absorption and the 3.3 $\mu$m PAH emission are detected.
546: Thus, its nuclear 3--4 $\mu$m emission is a composite of a 
547: highly obscured AGN and detectable star-forming activity.
548: In this case, the absorption-like feature at $\sim$10 $\mu$m is 
549: caused not only by the 9.7 $\mu$m silicate dust absorption but also 
550: by the 7.7 $\mu$m PAH emission.
551: The estimate of the column density of carbonaceous dust in front of 
552: a background obscured AGN based on the observed $\tau_{3.4}$ value 
553: is uncertain because star-forming activity contaminates the 3--4 $\mu$m 
554: continuum emission.
555: 
556: In the spectra of NGC 7172 and NGC 7479, 
557: the 3.4 $\mu$m carbonaceous dust absorption is detected, but 
558: no 3.3 $\mu$m PAH emission is detected.
559: Hence, their nuclear spectra are thought to be dominated by 
560: the emission of highly obscured AGNs.
561: In this case, the absorption-like feature at
562: $\sim$10 $\mu$m detected in ground-based 8--13 $\mu$m spectra is
563: ascribed mostly to the 9.7 $\mu$m silicate dust absorption.  
564: The ground-based 8--13 $\mu$m spectra of Roche et al. (1991) give the
565: optical depths $\tau_{9.7}$ of the 9.7 $\mu$m silicate dust absorption
566: feature as $>$2 and $\sim$2.5, respectively, for NGC 7172 and NGC 7479
567: (Table 2).  Using these $\tau_{9.7}$ values, the ratios
568: $\tau_{3.4}$/$\tau_{9.7}$ $<$ 0.04 and $\tau_{3.4}$/$\tau_{9.7}$
569: $\sim$ 0.10 are obtained towards the nuclei of 
570: NGC 7172 and NGC 7479, respectively.  
571: The $\tau_{3.4}$/$\tau_{9.7}$ ratios are 
572: 0.06--0.07 towards any direction in the Galactic diffuse inter-stellar 
573: medium (Pendleton et al. 1994; Roche \& Aitken 1984, 1985).
574: 
575: In the literature, we find three more nuclei with detected dust 
576: absorption at 3.4 $\mu$m and 9.7 $\mu$m, and no detectable PAH 
577: emission features.
578: These three sources, together with NGC 7172 and NGC 7479, are summarized 
579: in Table 2. 
580: Among the five sources, NGC 7172 and NGC 5506 show
581: $\tau_{3.4}$/$\tau_{9.7}$ ratios smaller than the ratios in the 
582: Galactic diffuse inter-stellar medium, 
583: while NGC 7479, NGC 1068, and IRAS 08572+3915 show larger
584: $\tau_{3.4}$/$\tau_{9.7}$ ratios.
585: 
586: Since the Galactic extinction is negligible towards all five nuclei 
587: ($A_{\rm V}$ $<$ 1 mag; Burstein \& Heiles 1984; 
588: the Einstein On-Line Service
589: %-------------
590: \footnote{The ftp address is 131.142.11.73.}
591: ), 
592: %-------------
593: the dust extinction must be outside the Galaxy.
594: The host galaxies of NGC 7172 and NGC 5506, the two sources with 
595: smaller $\tau_{3.4}$/$\tau_{9.7}$ ratios, are seen from 
596: an almost edge-on direction 
597: (Sharples et al. 1984; Anupama et al. 1995; Whittle 1992), 
598: while the viewing directions towards the host galaxies of 
599: NGC 7479, NGC 1068, and IRAS 08572+3915, the three sources with 
600: larger $\tau_{3.4}$/$\tau_{9.7}$ ratios, are relatively face-on 
601: (Ma, Peng \& Gu 1998; Whittle 1992; Surace et al. 1998). 
602: A trend is found, that the $\tau_{3.4}$/$\tau_{9.7}$ ratios are smaller 
603: than the Galactic value in two nuclei with edge-on host galaxies, 
604: while the ratios are larger in three nuclei with face-on host galaxies.
605: 
606: We interpret all five nuclei as powered predominantly by 
607: obscured AGN activity, based on the absence of detectable 
608: PAH emission features at 3--4 $\mu$m and 8--13 $\mu$m.
609: However, since the detection feasibility of the PAH emission depends on 
610: the spectral quality of individual sources, 
611: emission of star-forming activity could be non-negligible in some 
612: of the five nuclei.
613: In this case, the $\tau_{3.4}$ value could be suppressed 
614: and the apparent $\tau_{9.7}$ value could increase (section 1), 
615: that is, the $\tau_{3.4}$/$\tau_{9.7}$ ratio could decrease.
616: Hence, although the interpretation of smaller $\tau_{3.4}$/$\tau_{9.7}$ 
617: ratios requires some cautions, the larger $\tau_{3.4}$/$\tau_{9.7}$ 
618: ratios cannot be caused by the possible contamination of the emission 
619: of star-forming activity.
620: 
621: \subsection{Possible explanations for the smaller and larger  
622: $\tau_{3.4}$/$\tau_{9.7}$ ratios}
623: 
624: First, smaller $\tau_{3.4}$/$\tau_{9.7}$ ratios are found towards 
625: two highly obscured AGNs whose host galaxies are seen from an 
626: edge-on direction. 
627: It seems reasonable that dust in the host galaxies are responsible 
628: for the obscuration 
629: (Young et al. 1996; Murayama, Mouri \& Taniguchi 2000).
630: The galaxy types of these host galaxies are S0--Sa 
631: (Anupama et al. 1995; Kinney et al. 1991), earlier than that of 
632: the Galaxy (Sb/bc; Kerr 1993).
633: If the contribution by carbonaceous dust to dust extinction is smaller 
634: in these host galaxies, the smaller $\tau_{3.4}$/$\tau_{9.7}$ ratios 
635: could be explained.
636: 
637: Next, larger $\tau_{3.4}$/$\tau_{9.7}$ ratios are found in 
638: three highly obscured AGNs for which the viewing directions towards 
639: their host galaxies are relatively face-on.
640: Carbonaceous dust is more easily destroyed than silicate dust 
641: (Draine \& Salpeter 1979). 
642: If the obscuring dust towards the three AGNs suffers less dust 
643: destruction than the Galactic diffuse inter-stellar medium, then 
644: the larger $\tau_{3.4}$/$\tau_{9.7}$ ratios could be explained.
645: However, we suggest that the larger ratios can be explained more 
646: naturally by the presence of a temperature gradient in the obscuring 
647: dust.
648: If the obscuring dust is located close to a central compact energy 
649: source, a temperature gradient is predicted to occur, 
650: with the temperature of the dust decreasing with increasing distance 
651: from the central engine (Pier \& Krolik 1992).  
652: The temperature of the innermost dust ($<$ a few pc) is expected 
653: to be $\sim$1000 K, close to the dust sublimation temperature.  
654: Since emission at $\sim$3 $\mu$m is dominated by dust at $\sim$1000 K, 
655: the extinction estimated using the $\sim$3 $\mu$m data 
656: (i.e., observed $\tau_{3.4}$) should reflect the value towards the 
657: innermost dust around the central energy source.
658: On the other hand, dust at $\sim$300 K, 
659: a dominant emission source at 10 $\mu$m, is located further out 
660: than the $\sim$1000 K dust, and thus the extinction estimated using 
661: the $\sim$10 $\mu$m data (i.e., observed $\tau_{9.7}$) is only 
662: towards the outer region.
663: Hence, the $\tau_{3.4}$/$\tau_{9.7}$ ratios can be significantly larger.
664: If this suggestion is correct, our results may provide 
665: spectroscopic evidence for the presence of the putative ``dusty tori'' 
666: in the close vicinity ($<$ a few pc in inner radius) of AGNs 
667: (Antonucci 1993).
668: 
669: \section{SUMMARY}  
670: 
671: The following main results have been found.
672: 
673: \begin{enumerate}
674: \item The 3.4 $\mu$m carbonaceous dust absorption feature has been
675: detected towards the nuclei of NGC 3094, NGC 7172, and NGC 7479.
676: These nuclei are those with a strong absorption-like feature 
677: at $\sim$10 $\mu$m but with no detectable PAH emission features 
678: in ground-based 8--13 $\mu$m spectra.
679: NGC 3094 shows a detectable 3.3 $\mu$m PAH emission feature, while 
680: NGC 7172 and NGC 7479 do not.
681: The detection of the 3.4 $\mu$m absorption, together with the
682: non-detection of the 3.3 $\mu$m PAH emission, suggests that the 
683: nuclei of NGC 7172 and NGC 7479 are powered by 
684: highly obscured AGN activity. 
685: Their strong absorption-like feature at $\sim$10 $\mu$m should be 
686: attributed mostly to 9.7 $\mu$m silicate dust absorption.
687: On the other hand, the detection of both 3.4 $\mu$m absorption and 
688: 3.3 $\mu$m PAH emission in NGC 3094 suggests that the 
689: nucleus is powered both by highly obscured AGN activity and detectable 
690: star-forming activity.
691: Both the 9.7 $\mu$m silicate dust absorption and the 7.7 $\mu$m PAH 
692: emission could be responsible for the strong absorption-like feature 
693: at $\sim$10 $\mu$m.
694: 
695: \item We compared the $\tau_{3.4}$/$\tau_{9.7}$ ratios for five 
696: highly obscured AGNs with detected 3.4 $\mu$m and 9.7 $\mu$m 
697: absorption and without detectable PAH emission features.
698: We found that the ratios in the three highly obscured AGNs whose 
699: host galaxies are seen from a face-on direction are larger than 
700: the ratios in the Galactic diffuse inter-stellar medium or the ratios 
701: in the two highly obscured AGNs whose host galaxies are seen 
702: from an edge-on direction.
703: For the three sources, the larger ratios can be explained if the 
704: obscuring dust is so close to the central AGNs that a temperature 
705: gradient occurs in it. 
706: Our results may provide spectroscopic evidence for the presence of 
707: the putative ``dusty tori'' around AGNs.
708: 
709: \end{enumerate}
710: 
711:   
712: \section*{Acknowledgments}      
713: 
714: We thank Dr. T. Kerr, Dr. J. Davies, T. Carroll, and T. Wold for their 
715: support during the UKIRT observing run, and Dr. C. C. Dudley for his 
716: careful reading of the manuscript.
717: L. Good kindly proofread this manuscript.
718: The anonymous referee gave useful comments.
719: The United Kingdom Infrared Telescope is operated by the Joint Astronomy 
720: Centre on behalf of the U.K. Particle Physics and Astronomy Research 
721: Council.
722: Drs. A. T. Tokunaga and H. Ando give MI the opportunity to work at 
723: the University of Hawaii.
724: MI is financially supported by the Japan Society for the Promotion 
725: of Science during his stay at the University of Hawaii.
726: 
727: \begin{thebibliography}{} 
728: \bibitem{}Aitken D. K., Jones B., 1973, ApJ, 184, 127
729: \bibitem{}Allamandola L. J., Tielens A. G. G. M., Barker J. R., 
730:           1989, ApJS, 71, 733 
731: \bibitem{}Alonso-Herrero A., Simpson C., Ward M. J., Wilson A. S., 1998,  
732:           ApJ, 495, 196
733: \bibitem{}Antonucci R., 1993, ARA\&A, 31, 473
734: \bibitem{}Anupama G. C., Kembhavi A. K., Elvis M., Edelson R., 
735:           1995, MNRAS, 276, 125
736: \bibitem{}Armus L., Heckman T. M., Miley G. K., 1989, ApJ, 347, 727 
737: \bibitem{}Bridger A., Wright C. S., Geballe T. R., 1994, 
738:           in McLean I., ed., 
739:           Infrared Astronomy with Arrays: The Next Generation, 
740:           Kluwer, Dordrecht, p. 537
741: \bibitem{}Burstein D., Heiles C., 1984, ApJS, 54, 33
742: \bibitem{}Clavel J., Schulz B., Altieri B., Barr P., Claes P., 
743:           Heras A., Leech K., 1999, 
744:           in Gaskell C. M., Brandt W. N., Dietrich M.,
745:           Dultzin-Hacyan D., Eracleous M., eds., ASP Conf. Ser. 175, 
746:           Structure and Kinematics of Quasar Broad Line Regions. 
747:           Astron.\ Soc.\ Pac., San Francisco, p. 387 (astro-ph/9806054)
748: \bibitem{}Draine B. T., Salpeter E. E., 1979, ApJ, 231, 438
749: \bibitem{}Dudley C. C., 1998, PhD thesis, University of Hawaii
750: \bibitem{}Dudley C. C., 1999, MNRAS, 307, 553
751: \bibitem{}Dudley C. C., Wynn-Williams C. G., 1997, ApJ, 488, 720 
752: \bibitem{}Genzel R., et al., 1998, ApJ, 498, 579
753: \bibitem{}Glass I. S.,  Moorwood A. F. M., 1985, MNRAS, 214, 429
754: \bibitem{}Ho L. C., Filippenko A. V.,  Sargent W. L. W., 
755:           1997, ApJS, 112, 315
756: \bibitem{}Huang J. H., Gu Q. S., Su H. J., Hawarden T. G., Liao X. H., 
757:           Wu G. X., 1996, A\&A, 313, 13
758: \bibitem{}Imanishi M., 2000, MNRAS, 313, 165
759: \bibitem{}Imanishi M., Terada H., Sugiyama K., Motohara K., 
760:           Goto M., Maihara T., 1997, PASJ, 49, 69  
761: \bibitem{}Kerr F. J., 1993, in Holt S. S., Verter F., eds., 
762:           AIP Conf. Proc. 278, Back to the Galaxy, AIP, New York, p. 3 
763: \bibitem{}Kinney A. L., Antonucci R. R. J., Ward M. J., Wilson A. S., 
764:           Whittle M., 1991, ApJ, 377, 100 
765: \bibitem{}Kotilainen J. K., Ward M. J., Boisson C., Depoy D. L., 
766:           Bryant L. R.,  Smith M. G., 1992, MNRAS, 256, 125
767: \bibitem{}Larkin J. E., Armus L., Knop R. A., Soifer B. T., 
768:            Matthews K., 1998, ApJS, 114, 59
769: \bibitem{}Lawrence A., Ward M., Elvis M., Fabbiano G., 
770:           Willner S. P., Carleton N. P.,  Longmore A., 1985, 
771:           ApJ, 291, 117
772: \bibitem{}Ma J., Peng Q. -H.,  Gu Q. -S., 1998, A\&AS, 130, 449
773: \bibitem{}Moorwood A. F. M., 1986, A\&A, 166, 4
774: \bibitem{}Mountain C. M., Robertson D. J., Lee T. J., Wade R., 
775:           1990, Proc. SPIE, 1235, 25
776: \bibitem{}Murayama T., Mouri H.,  Taniguchi Y., 2000, ApJ, 528, 179 
777: \bibitem{}Pendleton Y. J., 1996, in Greenberg J., ed., 
778:           The Cosmic Dust Connection, Kluwer, Dordrecht, p. 71 
779: \bibitem{}Pendleton Y. J., Sandford S. A., Allamandola L. J., 
780:           Tielens A. G. G. M., Sellgren K., 1994, ApJ, 437, 683 
781: \bibitem{}Piccinotti G., Mushotzky R. F., Boldt E. A., 
782:           Holt S. S., Marshall F. E., Serlemitsos P. J., 
783:            Shafer R. A., 1982, ApJ, 253, 485
784: \bibitem{}Pier E. A., Krolik J. H., 1992, ApJ, 401, 99
785: \bibitem{}Rigopoulou D., Spoon H. W. W., Genzel R., Lutz D., 
786:           Moorwood A. F. M., Tran Q. D., 1999, AJ, 118, 2625
787: \bibitem{}Roche P. F., Aitken D. K., 1984, MNRAS, 208, 481 
788: \bibitem{}Roche P. F., Aitken D. K., 1985, MNRAS, 215, 425  
789: \bibitem{}Roche P. F., Aitken D. K., Phillips M. M.,  Whitmore B., 
790:           1984, MNRAS, 207, 35
791: \bibitem{}Roche P. F., Aitken D. K., Smith C. H., Ward M. J., 
792:           1991, MNRAS, 1991, 248, 606 
793: \bibitem{}Sandage A.,  Tammann G. A., 1987, A Revised Shapley-Ames 
794:           Catalog of Bright Galaxies, Carnegie Institution of 
795:           Washington Publication 635, Washington, D. C., Second Edition, 
796:           p. 66
797: \bibitem{}Sharples R. M., Longmore A. J., Hawarden T. G.,  
798:           Carter D., 1984, MNRAS, 208, 15
799: \bibitem{}Smith R. G., Sellgren K., Brooke T. Y., 1993, MNRAS, 263, 749
800: \bibitem{}Surace J. A., Sanders D. B., Vacca, W. D., Veilleux S., 
801:           Mazzarella, J. M., 1998, ApJ, 492, 116
802: \bibitem{}Tokunaga A. T., Sellgren K., Smith R. G., Nagata T., 
803:           Sakata A., Nakada Y., 1991, ApJ, 380, 452 
804: \bibitem{}Unger S. W., Lawrence A., Wilson A. S., Elvis M.,  
805:           Wright A. E., 1987, MNRAS, 228, 521
806: \bibitem{}Whittle M., 1992, ApJS, 79, 49
807: \bibitem{}Willner S. P., Elvis M., Fabbiano G., Lawrence A., 
808:            Ward M. J., 1985, ApJ, 299, 443
809: \bibitem{}Wright G. S., Bridger A., Geballe T. R., Pendleton Y., 1996, 
810:           Block D. L., \& Greenberg J. M., (eds), 
811:           in New Extragalactic Perspectives in the New South Africa, 
812:           Dordrecht, Kluwer, p.143
813: \bibitem{}Young S., Hough J. H., Efstathiou A., Wills B. J., 
814:           Bailey J. A., Ward M. J., Axon D. J., 1996, MNRAS, 
815:           281, 1206  
816: \end{thebibliography} 
817: 
818: %%%%%%%%%%%%%%%%%%%%%%%%%%  Tables %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
819: \clearpage
820: 
821: %%%%%%%%%%%%%%%%%%   Table 1   %%%%%%%%%%%%%%%%%%%%%
822: \small
823: \begin{table*}
824: \begin{center}
825: 
826: \caption{Observing log}
827: \begin{tabular}{ccccccccc} \hline \hline
828: Object & $z$ & Date & Integration & Air & \multicolumn{4}{c}
829: {Standard Star} \\ 
830:  & & (UT) & Time (sec) & Mass & Name & $L$-mag & Type & 
831: Temperature (K) \\ \hline
832: NGC 3094 & 0.008 & Feb 21, 2000 & 2560 & 1.0--1.2 & HR 3975 & 3.3 & A0Ib 
833: & 9700 \\
834: NGC 7172 & 0.009 & Sep 9, 1999  & 3200 & 1.6--1.7 & HR 8087 & 5.3 & A0V  
835: & 9480 \\
836: NGC 7479 & 0.008 & Sep 9, 1999  & 2400 & 1.0--1.1 & HR 8911 & 4.9 & A0 
837: & 9480 \\
838: \hline
839: \end{tabular}
840: \end{center}
841: \end{table*}
842: 
843: \normalsize
844: 
845: 
846: 
847: %%%%%%%%%%%%%%%%%%%%%%  Table 2  %%%%%%%%%%%%%%%%%%%%%%%%%
848: 
849: %\clearpage
850: 
851: \begin{table*}
852: \begin{center}
853: 
854: \caption{The $\tau_{3.4}$, $\tau_{9.7}$ and $\tau_{3.4}$/$\tau_{9.7}$ 
855: towards sources with detected 3.4 $\mu$m and 9.7 $\mu$m absorption features  
856: and with no detectable PAH emission features.}
857: \begin{tabular}{llcccc} \hline \hline
858: Object & $z$ & $\tau_{3.4}$ & $\tau_{9.7}$ & $\tau_{3.4}$/$\tau_{9.7}$ 
859: & References \\ \hline
860: NGC 1068 & 0.004 & 0.12 & 0.52 & 0.23 & 1, 2 \\
861: IRAS 08572+3915 & 0.058 & 0.9  & 5.2 & 0.17 & 3, 4 \\
862: NGC 7479 & 0.008 & 0.25 & $\sim$2.5 & 0.10& 5, 6 \\
863: NGC 5506 & 0.006 & 0.03 & 1.3 & 0.02 & 7, 2 \\
864: NGC 7172 & 0.009 & 0.08 & $>$2 & $<$0.04 & 5, 6 \\
865: \hline
866: \end{tabular}
867: 
868: References. --- (1) Imanishi et al. 1997; 
869: (2) Roche et al. 1984; 
870: (3) Pendleton 1996; 
871: (4) Dudley \& Wynn-Williams 1997; 
872: (5) this work; 
873: (6) estimated based on the spectra of Roche et al. (1991). 
874:     We estimate $\tau_{9.7}$ by (a) using the formula of 
875:     $\tau_{9.7}$ $=$ ln [$\frac{F_{\rm \lambda}(8) + F_{\rm \lambda}(13)}
876:     {2F_{\rm \lambda}(9.7)}$] (Aitken \& Jones 1973), and by (b) 
877:     deriving the optical depth at 9 $\mu$m ($\tau_{9}$) and by 
878:     adopting the relation of $\tau_{9.7}$/$\tau_{9}$ $\sim$ 1.7  
879:     (Dudley \& Wynn-Williams 1997).
880:     Both methods provide consistent results;   
881: (7) Imanishi 2000. 
882: 
883: \end{center}
884: \end{table*}
885: 
886: %%%%%%%%%%%%%%%%%%%%%%% Figure caption  %%%%%%%%%%%%%%%%%%%%%%%%
887: \clearpage
888: 
889: \begin{figure*}
890: \centerline{\psfig{file=fig1.ps,angle=0,width=6.5in}}
891: \caption{
892: A flux-calibrated spectrum of NGC 3094.
893: A spectral resolution is $\sim$350 after binning.
894: The ordinate is F$_{\lambda}$ in W m$^{-2}$ $\mu$m$^{-1}$, and the 
895: abscissa is the observed wavelength in $\mu$m.
896: The solid line is the adopted continuum level to measure the 
897: optical depth of the 3.4 $\mu$m absorption feature (see text).
898: The emission-like feature at 3.74 $\mu$m is caused by 
899: the stellar absorption of the type-A standard star (see text).
900: }
901: \end{figure*} 
902: 
903: \begin{figure*}
904: \centerline{\psfig{file=fig2.ps,angle=0,width=6.5in}}
905: \caption{
906: A flux-calibrated spectrum of NGC 7172, displayed in the same way as 
907: NGC 3094 (Fig. 1).
908: }
909: \end{figure*} 
910: 
911: \newpage
912: 
913: \begin{figure*}
914: \centerline{\psfig{file=fig3.ps,angle=0,width=6.5in}}
915: \caption{
916: A flux-calibrated spectrum of NGC 7479, displayed in the same way as 
917: NGC 3094 (Fig. 1).
918: This spectrum is shown with a spectral resolution of $\sim$170 
919: at 3.5 $\mu$m. 
920: }
921: \end{figure*} 
922: 
923: %\label{lastpage}
924: 
925: \end{document} 
926: