1: %\documentclass[manuscript]{aastex}
2: \documentclass[preprint2]{aastex}
3: %\documentclass[12pt,preprint]{aastex}
4: \shorttitle{Merger remnants in QSO host galaxies}
5: \shortauthors{Bennert et al.}
6:
7: \begin{document}
8:
9: \title{EVIDENCE FOR MERGER REMNANTS IN
10: EARLY-TYPE HOST GALAXIES OF LOW-REDSHIFT QSOS\altaffilmark{1}}
11:
12:
13: \author{Nicola Bennert\altaffilmark{2}, Gabriela Canalizo\altaffilmark{2,3},
14: Bruno Jungwiert\altaffilmark{2,4},
15: Alan Stockton\altaffilmark{5}, Fran\c{c}ois Schweizer\altaffilmark{6},
16: Chien Y. Peng\altaffilmark{7,8},
17: Mark Lacy\altaffilmark{9}}
18:
19: \altaffiltext{1}{Based on observations made with the
20: NASA/ESA {\em Hubble Space
21: Telescope}, obtained at the Space Telescope Science Institute, which is
22: operated by the Association of Universities for Research in Astronomy, Inc.,
23: under NASA contract NAS 5-26555. These observations are associated with
24: program \# GO-10421.}
25:
26: \altaffiltext{2}{Institute of Geophysics and
27: Planetary Physics, University of California, Riverside, CA 92521;
28: nicola.bennert@ucr.edu, gabriela.canalizo@ucr.edu, bruno.jungwiert@ucr.edu}
29:
30: \altaffiltext{3}{Department of Physics and Astronomy, University of California, Riverside, CA 92521}
31:
32: \altaffiltext{4}{Astronomical Institute, Academy of Sciences of the Czech Republic,
33: Bo{\v c}n\'\i\ II 1401, 141 31 Prague 4, Czech Republic}
34:
35: \altaffiltext{5}{Institute for Astronomy, University of Hawaii, 2680 Woodlawn Drive,
36: Honolulu, HI 96822; stockton@ifa.hawaii.edu}
37:
38: \altaffiltext{6}{Carnegie Observatories,
39: 813 Santa Barbara Street, Pasadena, CA 91101; schweizer@ociw.edu}
40:
41: \altaffiltext{7}{Space Telescope Science Institute, 3700 San Martin Drive,
42: Baltimore, MD 21218}
43:
44: \altaffiltext{8}{NRC Herzberg Institute of Astrophysics, 5071 West Saanich
45: Road, Victoria, British Columbia, Canada V9E 2E7; cyp@nrc-cnrc.gc.ca}
46:
47: \altaffiltext{9}{{\em Spitzer} Science Center,
48: California Institute of Technology, Pasadena, CA 91125;
49: mlacy@ipac.caltech.edu}
50:
51: \begin{abstract}
52: We present results from a pilot {\em HST} ACS deep
53: imaging study in broad-band $V$ of
54: five low-redshift QSO host galaxies classified in the literature as ellipticals.
55: The aim of our study is to determine whether these early-type hosts formed at
56: high redshift and have since evolved passively, or whether
57: they have undergone relatively recent mergers that may be related to the triggering
58: of the nuclear activity.
59: We perform two-dimensional modeling of the
60: light distributions to analyze the host galaxies' morphology.
61: We find that, while each host galaxy is reasonably
62: well fitted by a de Vaucouleurs profile, the majority of them (4/5) reveal
63: significant fine structure such as shells and tidal tails. These structures
64: contribute between $\sim$5\% and 10\% to the total $V$-band luminosity of each
65: host galaxy within a region of $r \sim$ 3\,$r_{\rm eff}$ and are indicative
66: of merger events that occurred between a few hundred Myr and a Gyr ago.
67: These timescales are comparable to starburst ages in the
68: QSO hosts previously inferred from Keck spectroscopy. Our results thus support
69: a consistent scenario in which most
70: of the QSO host galaxies suffered mergers with accompanying
71: starbursts that likely also triggered the QSO activity in some way,
72: but we are also left with considerable uncertainty on physical
73: mechanisms that might have delayed this triggering for several
74: hundred Myr after the merger.
75: \end{abstract}
76:
77: \keywords{galaxies: active -- galaxies: interactions --- galaxies: evolution --- quasars: general}
78:
79: \section{INTRODUCTION}
80: Quasi-stellar objects (QSOs) are the most luminous active
81: galactic nuclei (AGNs), believed to be powered by accreting supermassive black
82: holes (BH). The growing observational evidence that massive BHs exist not only
83: in the centers of active galaxies, but also in inactive early-type galaxies
84: indicates that more than the mere presence of a massive
85: BH is needed to trigger the activity.
86: And while the ``bright quasar phase'' is now often cited as an essential
87: phase in the evolution of galaxies \citep[e.g.,][]{spr05a},
88: we still do not have
89: a clear picture of what causes the triggering, or even what the
90: necessary conditions are for the triggering to occur.
91:
92: For more than two decades now, mergers have been suspected to be responsible
93: for triggering QSOs \citep{sto82,san88}. There is an abundance of circumstantial
94: evidence in the literature connecting QSO activity at low redshifts with
95: mergers and interactions \citep[e.g.,][]{hec84,hut88,can01,urr07}.
96: Not only do a number of low-redshift QSOs show signs of interactions and mergers,
97: but there is also a tendency for QSO host galaxies
98: to exist in only moderately rich environments
99: \citep[e.g.,][see, however, \citealt{mar07} for a counterexample]{dre85,ell91,wol01}
100: consistent with the results of quasar clustering
101: studies, which suggests that quasars exist in haloes of mass $\sim$10$^{12}$ - 10$^{13}$
102: $M_{\odot}$ \citep{hop07a, daa07}.
103: In such environments, velocity dispersions are
104: moderate, and interactions have their greatest effect.
105:
106:
107: At least in the case of QSOs found in ultraluminous infrared galaxies
108: (ULIRGs), there is a close connection between mergers and QSO activity.
109: ULIRGs are essentially
110: always found in merging systems \citep{san96}, and far too many of them harbor QSOs
111: to be explained by the joint probability of finding a given $L^*$ or
112: brighter galaxy to be both a ULIRG and a QSO host galaxy by chance
113: \citep{can01}. So, at least for this subclass of QSO, some of
114: the gas that the merger feeds into the central starburst also finds its way
115: much deeper to the central black hole.
116:
117: Nevertheless, recent high-resolution
118: imaging studies show that a large fraction of QSOs reside in hosts with
119: relaxed morphologies
120: like elliptical galaxies \citep[e.g.,][]{dis95,bah97,mcl99,flo04}. These
121: observations have elicited claims that tidal interactions in QSO host
122: galaxies are the exception rather than the rule, and that even in those
123: cases in which there are clear signs of interaction, the interactions are not
124: necessarily related to the nuclear activity
125: \citep[e.g.,][hereafter D03]{dun03}.
126: From two-dimensional fitting of a well matched sample of
127: 33 AGNs (radio-loud and radio-quiet QSOs as well as radio galaxies),
128: using images obtained with the Wide Field \& Planetary Camera 2 (WFPC2)
129: on board {\em HST}, D03
130: concluded that ``for nuclear luminosities $M_V < -23.5$, the hosts
131: of both radio-loud {\em and} radio-quiet AGN are virtually all massive
132: elliptical galaxies with basic properties that are indistinguishable
133: from those of quiescent, evolved, low-redshift ellipticals of comparable
134: mass''.
135:
136: Such conclusions conflict with recent findings of
137: significant amounts
138: of dust and cold molecular gas, as well as young stellar populations,
139: in many radio-loud and radio-quiet QSOs \citep[e.g.][]{mar99, can01,
140: eva01, sco03, tad05, bar06}.
141: Although the hosts of most luminous AGNs are bulge-dominated,
142: they seem to be significantly bluer than inactive elliptical galaxies and
143: indeed show evidence for starbursts in the relatively recent
144: past (1--2 Gyr) \citep[e.g.,][]{kau03, san04, can06, sch06, jah07}.
145:
146: In addition, both numerical simulations and observations show that merger remnants
147: that have undergone violent relaxation have surface-brightness profiles that
148: follow $r^{1/4}$ laws \citep[e.g.,][]{too72,bar88,hib96}, even while still
149: showing clear signs of the tidal interactions \citep{rot04}.
150: Thus, if QSO host galaxies are indeed relatively recent merger remnants,
151: they would most likely have elliptical profiles, but they would also show
152: fine structure indicative of past interactions.
153:
154: Detecting fine structure in the already elusive QSO host galaxies
155: is a challenging task. Studies of nearby merger remnants
156: show that their fine structure is often faint compared to the galaxies
157: themselves. For example, the total luminosity of the shells in
158: shell galaxies accounts only for 5--15\% of the total luminosity
159: of the galaxy \citep[e.g.,][]{for86,pri88,wil00},
160: and their detection often requires a resolution
161: of at least $\sim$0.5 kpc \citep[e.g.,][]{sch92}. Therefore,
162: detecting fine structure in QSO hosts requires high angular resolution
163: and higher signal-to-noise (S/N) observations than those obtained in
164: previous studies.
165:
166: Thus, we are conducting a study with deep (5 orbit) HST ACS and
167: WFPC2 observations
168: of a sample of QSO host galaxies that are classified as ellipticals.
169: We presented results for the first object in
170: \citet[][hereafter Paper I]{can07}.
171: In the present paper, we describe results for the four remaining objects from
172: our pilot ACS study of 5 objects.
173: We will present results for the rest of the sample (14 additional objects) in
174: subsequent papers.
175: Throughout this paper, we adopt $\Omega_{\lambda} = 0.7$,
176: $\Omega_{m} = 0.3$ and $H_0$ = 71 km\,s$^{-1}$\,Mpc$^{-1}$.
177:
178:
179: \section{SAMPLE SELECTION}
180: Since the goal of our project is to investigate the possibility
181: that QSO host galaxies are formed through mergers even when
182: they do not show dramatic signs of tidal interactions,
183: we should ideally draw our sample from a sample of hosts
184: that have been classified as undisturbed elliptical galaxies.
185: The best such sample is that of D03.
186:
187: From the detailed morphological study
188: of radio-loud and radio-quiet QSO host galaxies by D03,
189: we selected QSOs for which the host galaxies
190: were classified as elliptical galaxies\footnote{Note that
191: the D03 radio-loud and radio-quiet QSOs
192: were selected to be
193: statistically indistinguishable in their optical luminosity
194: and redshift.}.
195: Moreover, we selected those objects
196: for which existing deep Keck spectra
197: did not reveal extended emission lines
198: that could contaminate
199: our observations in the broad $V$-band filter (F606W).
200: We chose a sample
201: of five QSOs, three radio-quiet and two radio-loud.
202: The sample properties are summarized in Table~\ref{obstable}.
203:
204:
205: \section{OBSERVATIONS AND DATA REDUCTION}
206: All observations were obtained using
207: the ACS Wide Field Channel (WFC) onboard the {\em HST}
208: with the broad $V$-band F606W filter ($\Delta \lambda$ = 2342\AA; 1 pixel
209: corresponds to 0.05\arcsec).
210: Five sets of dithered images were taken (with four subsets of 485--588 s integration
211: time each), yielding a total integration time of 10740--11464 s
212: per target (see Table~\ref{obstable}).
213:
214: \begin{deluxetable}{lccccccc}
215: \tabletypesize{\scriptsize}
216: \tablecolumns{8}
217: \tablewidth{0pc}
218: \tablecaption{Details of Observation}
219: \tablehead{
220: \colhead{QSO} & \colhead{IAU designation} & \colhead{$\alpha$ (J2000)} & \colhead{$\delta$ (J2000)} & \colhead{$z$} & \colhead{Type} & \colhead{Exp. Time} & \colhead{scale}\\
221: & & (hh mm ss) & ($\degr$ $\arcmin$ $\arcsec$) & & & (sec) & (kpc/$\arcsec$)\\
222: \colhead{(1)} & \colhead{(2)} & \colhead{(3)} & \colhead{(4)} & \colhead{(5)}
223: & \colhead{(6)} & \colhead{(7)} & \colhead{(8)}}
224: \startdata
225: PHL\,909 & 0054+144 & 00 57 09.9 & +14 46 10 & 0.172 & RQQ & 10920 & 2.885\\
226: PKS\,0736+01 & 0736+017 & 07 39 18.0 & +01 37 05 & 0.191 & RLQ & 11364 & 3.137\\
227: PG\,0923+201 & 0923+201 & 09 25 54.7 & +19 54 05 & 0.190 & RQQ & 11464 & 3.124\\
228: MC2\,1635+119 & 1635+119 & 16 37 46.5 & +11 49 49 & 0.146 & RQQ & 11432 & 2.520\\
229: OX\,169 & 2141+175 & 21 43 35.5 & +17 43 49 & 0.211 & RLQ & 10740 & 3.392\\
230: \enddata
231: \tablecomments{
232: Col. (1): QSO. Col. (2): IAU designation. Col. (3,4): Optical positions
233: taken from NED. Col. (5): Heliocentric redshift as listed in NED. Col. (6):
234: Classification as either radio-quiet QSO (RQQ) or radio-loud QSO (RLQ).
235: Col. (7): Total integration time in sec. Col. (8): Physical scale in kpc/$\arcsec$.
236: }
237: \label{obstable}
238: \end{deluxetable}
239:
240: In order to achieve background-limited images enabling
241: us to detect any existing faint structure in the QSO host galaxies,
242: we obtained individual frames with long integration times.
243: Consequently, the central pixels of the QSO nuclei are saturated
244: in these images. We discuss
245: the effects of the saturation in detail in the next section.
246:
247: We re-calibrated the data manually,
248: correcting
249: the pipeline flat-fielded single exposures
250: for the bias-level offset between the
251: adjacent quadrants that is still present in
252: the final product of CALACS \citep{pav05}.
253: We used MultiDrizzle \citep{koe02} to combine
254: the individual images to the final
255: distortion and cosmic-ray corrected scientific image
256: (default values plus bits=8578 and a deltashift file with the offsets between
257: the images as determined from stars within the field-of-view (FOV)).
258: Note that no sky subtraction was performed during
259: MultiDrizzle.
260:
261:
262: \section{IMAGE ANALYSIS}
263: \label{process}
264: Different methods such as unsharp masking, creating
265: a structure map \citep{pog02}, and using the two-dimensional
266: galaxy fitting program GALFIT \citep{pen02}
267: were applied to look for any fine structure that might be present.
268: We here describe only the procedure
269: we adopted for the use of GALFIT as the other two methods did not
270: yield additional information
271: (see Paper I for details on these methods).
272:
273: For convolution with the point-spread function (PSF)
274: of the {\em HST} ACS optics, we created both an artificial
275: PSF star from
276: TinyTim\footnote{Available at \\http://www.stsci.edu/software/tinytim/tinytim.html}
277: (Version 6.3) at the same position
278: as each QSO and a PSF star from an observed star
279: on an ACS/WFC F606W image observed close
280: to the positions of the QSOs.
281: For the TinyTim PSF,
282: we adopted a power law $F_{\nu} = \nu^{\alpha}$ with $\alpha$ = $-$0.3
283: which is the average value we obtain for the QSOs.\footnote{$(B-R)$ ranges between
284: 0.1 and 0.7 as measured from nuclear Keck spectra.}
285: The computed radius for this artificial PSF is 10\arcsec.
286:
287: For the observed PSF star, we chose
288: a star with a signal-to-noise ratio (S/N) of 20,000 from the {\em HST}
289: archive. The star was observed
290: on 20 dithered images with a total exposure time of $\sim$8100\,s
291: (GO-9433, data sets j6mf19* and j6mf21*) and at a position of the WFC
292: chip less than 5$\arcsec$ away from the position of each QSO;
293: note that the full ACS FOV is 3.4\arcmin~$\times$ 3.4\arcmin.
294: These images were processed in the same manner as described above
295: for the QSO observations.
296: Finally, we eliminated a few faint objects surrounding the PSF star
297: and modified the PSF image to minimize any noise introduced
298: during PSF subtraction and convolution procedures.
299: To adaptively smooth the PSF,
300: we compared data values to the standard deviation $s$ of the surrounding
301: sky: (1) for data values $>$ 7$s$, we retained the unmodified PSF;
302: (2) for data values between 3$s$ and 7$s$, the image was smoothed with a Gaussian
303: kernel with $\sigma$ = 0.5 pixel; (3) for data values $<3s$, a Gaussian kernel
304: of $\sigma$ = 2.0 pixel was used, and (4) for data values $<1s$ after this
305: last operation, the value was replaced with 0.
306: The radius of the observed PSF star is $\sim$4\arcsec.
307: The full-width-at-half-maximum (FWHM)
308: of the PSF is 0.1\arcsec~in both
309: the artificial star and the observed PSF star,
310: in close agreement with the FWHM of stars
311: on the ACS images close to the QSO.
312:
313: The observed PSF star has $(B-R)$ $\simeq$ 1.4 mag, as inferred from USNO\footnote{United
314: States Naval Observatory} data.
315: To test chromatic effects
316: on the form of the PSF, we created a series of artificial PSF stars using
317: TinyTim. We used power laws with $\alpha$ ranging from -4.0 to 0.9,
318: corresponding to $(B-R)$ values between 2 mag and 0 mag (to account for the
319: observed range of $(B-R)$ of the QSOs and the star, plus possible
320: uncertainties).
321: We also created PSFs using a blackbody spectrum
322: with temperatures between 3000 and 40000 K.
323: We fitted these different PSFs
324: and a host galaxy model to our QSOs.
325: The derived host galaxy parameters vary by 2\%.
326: We thus consider the chromatic effects
327: as negligible compared to the other uncertainties
328: such as the PSF mismatch due to telescope
329: or plate scale breathing, focus changes,
330: PSF saturation (see below) etc.
331:
332: We subtracted the two different PSFs (star and TinyTim) from saturated
333: and unsaturated stars within the ACS FOV, for comparison.
334: The observed PSF star yielded significantly better results than the TinyTim PSF.
335: Thus, we used the observed PSF star for all the subsequent
336: fitting and, in the following, we simply refer to it as PSF or PSF star.
337: This exercise also allowed us to estimate the quality
338: of our PSF fit, taking into account the saturation of the QSO nucleus.
339: The central region within a radius of $\sim$1.7\arcsec~is
340: strongly affected by the PSF subtraction. Any structure
341: within this region may be an artifact.
342:
343: We created artificial images to test the influence of the
344: PSF saturation on the derived QSO host galaxy parameters.
345: Since luminous quasars generally do not vary much over time
346: scales of a few years, we can compare our quasar luminosities
347: with D03 to estimate the degree of saturation in our images and
348: the potential systematic errors that might arise due to saturation.
349: This direct comparison shows that our AGNs may be saturated
350: by as much as 60\%. (Note that we define "60\% saturation"
351: in such a way that 60\% of the flux is lost due to saturation.)
352: In addition, we created model images where
353: the AGN has the same luminosity as in D03. We then saturated
354: the AGN by 60\%. After masking out the central saturated pixels
355: in the same manner as we analyzed
356: our data (see below),
357: we find that our systematic errors can be as large as
358: 7\% for the magnitude, 30\% for the effective radius
359: of the different host galaxy components and up to 30\% for the
360: S{\'e}rsi{\'c} index. The PSF magnitude can be
361: underestimated by up to 90\% due to saturation. This explains the differences
362: in AGN luminosity that we observe compared with D03 (see Table~\ref{results}).
363:
364: GALFIT \citep{pen02} is a 2-dimensional fitting program based on $\chi$$^2$
365: minimization; it allows
366: the user to simultaneously fit one or more objects in an image
367: with different model light distributions, such as S{\'e}rsi{\'c} \citep{ser68},
368: de Vaucouleurs \citep{dev48}, or exponential.
369: The following steps were performed.
370: We created a mask to exclude the saturated pixels in the center,
371: the diffraction spikes, surrounding objects,
372: and any visible fine structure in order to fit only
373: the smooth underlying host galaxy light distribution.
374: For the purpose of the $\chi$$^2$ minimization, pixels were weighted by
375: 1/$\sigma$$^2$ with $\sigma$ based on Poisson statistics as estimated by
376: GALFIT. (Note that the weighting is however irrelevant
377: for the central saturated pixel that were masked.)
378: A PSF and a S{\'e}rsi{\'c} function were fitted.
379: We fitted the background sky and
380: bright close neighboring galaxies simultaneously, while faint neighboring objects
381: were simply masked.
382: This least-square fit was then subtracted
383: from the original image to obtain the residual image in which any
384: fine structure superposed on the smooth host galaxy light distribution is
385: more readily visible.
386:
387: To estimate the uncertainties caused by the sky determination,
388: we used two different approaches.
389: First, different sizes of the fitting region were used
390: (up to the largest possible size)
391: with the sky varying freely.
392: Second, the sky was also determined independently
393: and held fixed to this value during alternative fits.
394:
395: Note that a S{\'e}rsi{\'c} power law is defined as
396: \begin{eqnarray*}
397: \Sigma (r) = \Sigma_{\rm eff} \exp \left[- \kappa_n \left(\left(\frac{r}{r_{\rm eff}}\right)^{1/n}-1\right)\right] \hspace{0.2cm} ,
398: \end{eqnarray*}
399: where $\Sigma_{\rm eff}$ is the pixel surface brightness at the effective radius $r_{\rm eff}$,
400: and $n$ is the S{\'e}rsi{\'c} index ($n=4$ for a de Vaucouleurs
401: profile, $n=1$ for an exponential
402: profile).
403:
404: To fit the host galaxies, we used three different models:
405: (1) a single de Vaucouleurs profile
406: (S{\'e}rsi{\'c} with $n=4$), (2) a single S{\'e}rsi{\'c} profile ($n$ free)
407: and (3) a combination of a de Vaucouleurs plus exponential (S{\'e}rsi{\'c} with $n=1$)
408: profile.
409: If the sky value is kept fixed during the fitting process,
410: the fit for (1) has a total of nine free parameters: three parameters
411: for the fitting of the PSF ($\alpha$, $\delta$, magnitude),
412: and six free parameters for the fitting of
413: the de Vaucouleurs profile ($\alpha$, $\delta$, magnitude,
414: $r_{\rm eff}$, b/a, P.A.).
415: Model (2) has one additional free parameter for the index $n$.
416: Model (3) has six more free parameters than model (1).
417: These parameters correspond to the exponential
418: profile
419: ($\alpha$, $\delta$, magnitude,
420: $r_{\rm eff}$, b/a, P.A.).
421: In all three cases, if the sky is also fitted, there are three additional
422: parameters (background and gradient in $\alpha$ and $\delta$).
423: In some cases, we used a S{\'e}rsi{\'c} profile to fit
424: bright galaxies in close proximity to the QSO hosts (PHL\,909: six galaxies;
425: PG\,0923+201: seven galaxies, MC2\,1635+119: one galaxy).
426: In those cases, the free parameters for these fits add to the number of
427: total free parameters.
428:
429:
430:
431: \section{HOST GALAXY PROPERTIES AND FINE STRUCTURE}
432: In this section, we summarize the results
433: for all five QSOs in the pilot sample.
434: Although MC2\,1635+119 was discussed
435: in detail in Paper I, Figure~\ref{finalsbp}
436: shows its luminosity profile
437: which was not included in Paper I.
438: To facilitate comparisons, we also include the results for MC2\,1635+119 in the
439: tables.
440:
441: \subsection{Host Galaxy Morphology}
442: \label{morphology}
443: Figure~\ref{final} shows the ACS/WFC images for four of the five
444: QSOs in our sample: PHL\,909,
445: PKS\,0736+01, PG\,0923+201, and OX\,169.
446: Including MC2\,1635+119,
447: four of the five host galaxies reveal extended fine structure
448: on different scales and with different morphologies, ranging from
449: spectacular shells and arcs (MC2\,1635+119; Paper I),
450: to tidal tails (PHL\,909) to
451: jet-like (OX\,169) and spiral-like (PKS\,0736+01) structure.
452: Only one object (PG\,0923+201) resides in a host galaxy without
453: any obvious fine structure.
454: We here describe the overall host-galaxy morphology
455: for each object in turn. Further details on each object, including
456: a comparison with results in the literature and
457: a description of neighboring objects, are given in Appendix~\ref{individual}.
458:
459: The host galaxy of PHL\,909 shows
460: a variety of fine structures at different
461: radii. A central ring- or disk-like structure
462: ($r \simeq$1.8\arcsec~-- 2.5\arcsec)
463: can be seen, possibly intersected by a dust lane
464: to the SE (Fig.~\ref{final}).
465: To the NW, two shells are apparent. They occur at distances
466: of 2.5\arcsec~(P.A.~$\simeq$ 328) and 3\arcsec~(P.A.~$\simeq$ 298).
467: Diffuse outer material seems to form another ring
468: or disk-like structure (Fig.~\ref{acs}, inset; $r \simeq$ 3.4\arcsec~-- 6\arcsec).
469: The most prominent fine structure consists of two tidal tails,
470: one extending from the
471: disk-like structure in the SE, curving towards
472: the N, and the other extending from the disk in the NW,
473: bending towards the south (Figs.~\ref{final} and~\ref{acs}). The western tidal
474: tail extends to a galaxy (``a''; Fig.~\ref{acs}) $\simeq$16\arcsec~to
475: the NW (P.A.~$\simeq$ 284$\degr$).
476: The tidal feature most likely corresponds to the faint
477: extended emission reported by \citet{mcl99}
478: which was also seen in the K-band image \citep{dun93}.
479: The companion (``a'') itself has two tidal tails, one reaching out in the direction
480: of the QSO and connecting with its tidal tail,
481: another tidal tail extending to the S up to 5.5\arcsec~(Fig.~\ref{final}).
482: Also the eastern tidal tail seems to be connected
483: to a very faint small companion (``d'') at 14\arcsec~distance
484: from the QSO (P.A.~$\simeq$ 98$\degr$), possibly a dwarf galaxy (Fig.~\ref{acs},
485: inset).
486:
487: PKS\,0736+01 has been observed in the near-infrared
488: K-band by \citet{dun93}.
489: The host is about 10\arcsec~in diameter
490: and is described as highly disturbed with a large area
491: of low surface-brightness nebulosity
492: and several compact companions embedded.
493: In our deep {\em HST} ACS image, this nebulosity
494: is clearly visible and shows resolved structure that looks
495: like a faint face-on spiral.
496: The spiral-like structure
497: extends out to a radius of 16\arcsec~($\simeq$ 50\,kpc),
498: i.e.~very large for a normal spiral galaxy.
499: Also, the pitch angle changes with radius, arguing
500: against a simple spiral-arm interpretation.
501: Instead, what is observed here, may be
502: spatial wrapping of material from a
503: merger event (see discussion in Sec.~\ref{mergerage}).
504:
505: PG\,0923+201 resides in a galaxy group of at least 6 galaxies,
506: some of which form closely interacting pairs, including the two
507: large elliptical galaxies to the south-east of the QSO
508: (see Appendix~\ref{individual}).
509: However, the host galaxy itself does not reveal any
510: obvious fine structure in the deep {\em HST} ACS image.
511:
512:
513: The host galaxy of MC2\,1635+119 has been described
514: in detail in Paper I. It reveals spectacular fine structure,
515: consisting of (at least) five interleaved inner
516: shells (out to $r \simeq$ 13\,kpc) and
517: several extended arcs out to a radius of $\sim$ 65\,kpc.
518:
519: The host galaxy of OX\,169 exhibits an extended linear structure
520: which has been reported by several authors \citep{sto78,hut84,geh84,smi86,hec86}.
521: OX\,169 was studied by \citet{sto91} and \citet{hut92} who
522: described this jet-like structure in detail. \citet{sto91} showed
523: that it consists mainly of old stars at the same redshift.
524: This is supported by the prominence of the structure also
525: in the K-band image \citep{dun93}.
526: While the feature resembles a galactic disk seen edge-on, with a dust lane
527: in the SE, it is not centered on the QSO; almost 2/3 of it
528: lies on the SE side of the QSO.
529: Because of its strong asymmetry, \citet{sto91} favor an
530: interpretation as a tidal tail seen edge-on rather
531: than a normal galactic disk.
532: In our deep {\em HST} ACS image, the linear structure
533: is very prominent and extends over a projected distance of $\sim$68\,kpc.
534: The feature runs from NW to SE (P.A.~$\simeq$ 303$\degr$),
535: out to $\simeq$7.6\arcsec~in the NW and
536: $\simeq$12.5\arcsec~in the SE.
537: It is slightly curved and appears to continue across the nucleus.
538: A dust lane is clearly visible between 7.6\arcsec~and 11.5\arcsec~in the
539: SE side of the tail.
540: In addition to the prominent linear structure, the deep
541: {\em HST} ACS image reveals
542: faint extended structure for the first time:
543: a shell-like feature in the NE (P.A.~$\simeq$ 83$\degr$)
544: at a distance of $\simeq$5\arcsec, plumes to the N of the SE end
545: of the tail (P.A.~$\simeq$ 110$\degr$) at a distance of $\simeq$11.5\arcsec~from the
546: center, and a fan extending from the SE tail towards
547: the SW. The latter extends out to
548: a companion (``a''; Fig.~\ref{acs})
549: to the SW (P.A.~$\simeq$ 235$\degr$)
550: at a distance of 6.1\arcsec~and may be a tidal stream.
551: The companion has an overall spiral-like
552: morphology which seems to be tidally stretched.
553: These findings provide strong support
554: for the interpretation of the linear structure
555: as a tidal tail seen edge on rather than it
556: being a normal galactic disk.
557: It is only natural to assume that all features have a common
558: origin during a merger event.
559:
560: \onecolumn
561: \begin{figure}
562: \epsscale{1}
563: \plotone{f1_small.eps}
564: \caption{ACS/WFC images of four QSO host galaxies, all smoothed with
565: a $\sigma = 0.5$ pixel Gaussian.
566: In each image, north is up and east is to the left.
567: A harder stretch is shown as inset, Gaussian smoothed
568: with $\sigma = 1$ pixel (PHL\,909, OX\,169) and $\sigma = 2$ pixels
569: (PKS\,0736+01, PG\,0923+201), respectively.
570: Including MC2\,1635+119 (Paper I),
571: fine structure can be seen in at least four of the five
572: QSO host galaxies (PHL\,909: tidal tails and shells; PKS\,0736+201: spiral-like
573: structure; OX\,169: disk-like structure,
574: most likely tidal tail seen edge-on;
575: MC2\,1635+119: inner shells and outer arcs, see Paper I).
576: }\label{final}
577: \end{figure}
578: \twocolumn
579:
580:
581:
582: \subsection{Host Galaxy Properties}
583: As described in Section~\ref{process},
584: we modeled the QSO host galaxies with GALFIT
585: and fitted single de Vaucouleurs profiles,
586: single S{\'e}rsi{\'c} components,
587: and a combination of de Vaucouleurs and
588: exponential-disk profiles.
589: Varying the range of the sky parameter shows
590: that the uncertainty in photometry can be up to 20\%,
591: the uncertainty in $r_{\rm eff}$ up to 10\%,
592: and the uncertainty in S{\'e}rsi{\'c} index up to 20\%.
593: These uncertainties add to the uncertainties from the PSF saturation
594: (see Sec.~\ref{process}).
595: The total uncertainties are $\sim$20\% for photometry,
596: $\sim$30\% in $r_{\rm eff}$ and $\sim$35\% in the S{\'e}rsi{\'c} index.
597: The results are summarized in Table~\ref{results}.
598:
599: For all objects, good fits were achieved with all three
600: fitting functions mentioned above.
601: The fits improved slightly in going from a single de Vaucouleurs to
602: a single S{\'e}rsi{\'c} to a combination of de Vaucouleurs
603: plus exponential-disk profiles
604: (see Table~\ref{results}). While this is expected simply
605: because of the increasing number of free parameters,
606: the residual images do look smoother for the two-component
607: fits.
608: In such a decomposition, the de Vaucouleurs
609: profile remains the dominant component, but
610: the exponential-disk component contributes between
611: 6\% and 28\% of the total luminosity in $V$ of the hosts
612: (Table~\ref{results}).
613: The residual image did
614: improve significantly when adding a second component
615: to the host in MC2\,1635+119 (Paper I).
616:
617: Single-S{\'e}rsi{\'c} fits yielded a S{\'e}rsi{\'c} index
618: around 4 for each of the host galaxies (ranging
619: between 2.9 and 6.5).
620: Our results for the surface-brightness profiles
621: of the underlying host galaxies are thus in
622: overall agreement with those of \citet{dun03}.
623: In Figure~\ref{finalall}, we show the residual images
624: derived by subtracting a GALFIT model consisting
625: of a single S{\'e}rsi{\'c} profile from the
626: observed images.
627:
628:
629:
630: \subsection{Surface-Brightness Profiles}
631: We derived
632: surface-brightness profiles for each object using the
633: IRAF\footnote{IRAF (Image Reduction
634: and Analysis Facilities) is distributed by the National Optical Astronomy
635: Observatories, which are operated by AURA, Inc., under cooperative
636: agreement with the National Science Foundation.} task {\em ellipse}.
637: The profiles were obtained for both
638: the observed surface brightness and the best-fitting
639: model using GALFIT and its subcomponents.
640: The resulting 1D profiles for the different fits
641: (single de Vaucouleurs, single S{\'e}rsi{\'c},
642: de Vaucouleurs plus exponential) are only marginally
643: different. Therefore, we only present the
644: best fitting model using a S{\'e}rsi{\'c} component in Figure~\ref{finalsbp}.
645:
646: The deviation of the model from the data
647: within the central 1.7\arcsec~is most likely caused by the PSF subtraction.
648: This is especially pronounced
649: for PG\,0923+201, as this object has the brightest PSF
650: (see Table~\ref{results}), and the low S/N ratio in the outer
651: parts of the PSF star results in additional
652: noise in the data.
653: In general, data points at a higher
654: surface brightness than the model correspond to the
655: fine structure that was masked out during GALFIT fitting
656: but that was included for the creation of the 1D profiles.
657: This is clearly the case for OX\,169, for which
658: the tidal tail causes the deviation between data and
659: model beyond 5\arcsec.
660:
661: The surface-brightness plot of the observed host profile
662: of PG\,0923+201
663: shows a bump peaking around 2\arcsec~that is not fitted by the model.
664: Since this feature lies just outside of the central region that is affected by
665: the PSF, we found it difficult to determine whether it is real or not.
666: However, it could reflect the presence of an inner disk.
667: If so, the outer radius of this disk would be approximately
668: 8\,kpc.
669: The outer ``uplift'' of points over the best-fit profile
670: (around $r \sim 8$\arcsec) is most likely due to the companion
671: galaxy to the SSE and the relatively strong residuals that remain
672: after subtracting the GALFIT model for this galaxy (see also Fig.~\ref{finalall}).
673: The observed surface brightness for
674: PHL\,909 also deviates slightly from the fit with a
675: bump peaking around
676: 2.3\arcsec. This bump coincides with the
677: ring-like structure seen in the image,
678: with an outer radius of $\sim$7\,kpc.
679:
680: \begin{deluxetable}{lccccccccccc}
681: \tabletypesize{\scriptsize}
682: \tablecolumns{13}
683: \tablewidth{0pc}
684: \tablecaption{Results of Modeling the QSO Host Galaxies Using GALFIT}
685: \tablehead{
686: \colhead{QSO} & \colhead {Model} &
687: \colhead{Function} & \colhead{$\Delta$($\alpha$,$\delta$)} & \colhead{$m_{\rm F606W}$}
688: & \colhead{$R$} & \colhead{$r_{\rm eff}$} & \colhead{$r_{\rm eff}$} & \colhead{S{\'e}rsi{\'c}} & \colhead{$b/a$} &
689: \colhead{P.A.} & \colhead{$\chi_\nu^2$} \\
690: \colhead{(fitted area)}& & & \colhead{($\arcsec$, $\arcsec$)} & \colhead{(mag)} & \colhead{(mag)} & \colhead{($\arcsec$)} & \colhead{(kpc)} &
691: \colhead{index} & & \colhead{($\degr$)}\\
692: \colhead{(1)} & \colhead{(2)} & \colhead{(3)} & \colhead{(4)} & \colhead{(5)}
693: & \colhead{(6)} & \colhead{(7)} & \colhead{(8)} & \colhead{(9)} & \colhead{(10)} & \colhead{(11)} & \colhead{(12)} }
694: \startdata
695: PHL\,909 & deV+Exp & PSF & (0,0) & 17.23 & \nodata & \nodata & \nodata & \nodata & \nodata & \nodata & 4.83 \\
696: (950$\times$950) & & deV & ($-$0.1,$-$0.1) & 17.08 & \nodata & 2.35 & 6.77 & 4 (fixed) & 0.57 & 130.4 & \nodata\\
697: & & Exp & ($-$0.96,1.45) & 19.88 & \nodata & 3.14 & 9.05 & 1 (fixed) & 0.57 & 111.0 & \nodata\\
698: & S & PSF & (0,0) & 18.18 & \nodata & \nodata & \nodata & \nodata & \nodata & \nodata & 5.14 \\
699: & & S & (0.1,$-$0.06) & 16.79 & \nodata & 2.53 & 7.30 & 5.6 & 0.63 & 130.2 & \nodata\\
700: & deV & PSF & (0,0) & 17.39 & \nodata & \nodata & \nodata & \nodata & \nodata & \nodata & 5.17 \\
701: & & deV & ($-$0.1,$-$0.05) & 16.96 & \nodata & 2.54 & 7.32 & 4 (fixed) & 0.61 & 130.4& \nodata\\
702: & deV (D03) & PSF & \nodata & \nodata & 15.5 & \nodata & \nodata & \nodata & \nodata& \nodata & \nodata \\
703: & & deV & \nodata & \nodata & 16.6 & 2.76 & 7.93 & 4 (fixed) & 0.61 & 131 & \nodata \\
704: & S (D03) & S & \nodata & \nodata & \nodata & \nodata & \nodata & 4 & \nodata & \nodata & \nodata \\
705: \hline
706: \\*[-0.2cm]
707: PKS\,0736+01 & deV+Exp & PSF & (0,0) & 17.18 & \nodata & \nodata & \nodata & \nodata & \nodata & \nodata & 2.20 \\
708: (1051$\times$1051) & & deV & (0.01,$-$0.05) & 17.70 & \nodata & 1.63 & 5.10 & 4 (fixed) & 0.94 & 2.3& \nodata\\
709: & & Exp & ($-$0.94,$-$0.15) & 18.75 & \nodata & 7.09 & 22.23 & 1 (fixed) & 0.80 & 86.2& \nodata\\
710: & S & PSF & (0,0) & 16.07 & \nodata & \nodata & \nodata & \nodata & \nodata & \nodata & 2.25 \\
711: & & S & ($-$0.04,0) & 17.08 & \nodata & 4.27 & 13.41 & 6.5 & 0.98 & 32.7 & \nodata\\
712: & deV & PSF & (0,0) & 16.86 & \nodata & \nodata & \nodata & \nodata & \nodata & \nodata & 2.27 \\
713: & & deV & ($-$0.04,0) & 17.35 & \nodata & 3.30 & 10.34 & 4 (fixed) & 0.98 & 39.2& \nodata\\
714: & deV (D03) & PSF & \nodata & \nodata & 16.2 & \nodata & \nodata & \nodata & \nodata& \nodata & \nodata\\
715: & & deV & \nodata & \nodata & 16.9 & 3.27 & 10.21 & 4 (fixed) & 0.97 & 26 & \nodata\\
716: & S (D03) & S & \nodata & \nodata & \nodata & \nodata & \nodata & 5.3 & \nodata & \nodata & \nodata \\
717: \hline
718: \\*[-0.2cm]
719: PG\,0923+201 & deV+Exp & PSF & (0,0) & 16.39 & \nodata & \nodata & \nodata & \nodata & \nodata & \nodata & 2.59 \\
720: (1051$\times$1051) & & deV & ($-$0.04,$-$0.02) & 17.40 & \nodata & 1.44 & 4.49 & 4 (fixed) & 0.96 & 174.2 & \nodata\\
721: & & Exp & (1.02,0.75) & 20.45 & \nodata & 2.26 & 7.07 & 1 (fixed) & 0.51 & 146.1 & \nodata \\
722: & S & PSF & (0,0) & 16.15 & \nodata & \nodata & \nodata & \nodata & \nodata & \nodata & 2.76 \\
723: & & S & (0.07,$-$0.03) & 17.48 & \nodata & 1.81 & 5.66 & 2.9 & 0.97 & 42.1 & \nodata\\
724: & deV & PSF & (0,0) & 16.36 & \nodata & \nodata & \nodata & \nodata & \nodata & \nodata & 2.78 \\
725: & & deV & ($-$0.06,0.01) & 17.31 & \nodata & 1.68 & 5.2 & 4 (fixed) & 0.97 & 42.9 & \nodata\\
726: & deV (D03) & PSF & \nodata & \nodata & 15.7 & \nodata & \nodata & \nodata & \nodata & \nodata & \nodata\\
727: & & deV & \nodata & \nodata & 17.2 & 2.02 & 6.30 & 4 (fixed) & 0.98 & 38 & \nodata\\
728: & S (D03) & S & \nodata & \nodata & \nodata & \nodata & \nodata & 3.3 & \nodata & \nodata & \nodata \\
729: \hline
730: \\*[-0.2cm]
731: MC2\,1635+119 & deV+Exp & PSF & (0,0) & 17.58 & \nodata & \nodata & \nodata & \nodata & \nodata & \nodata & 3.49 \\
732: (481$\times$481) & & deV & ($-$0.03,0.04) & 17.50 & \nodata & 1.12 & 2.82 & 4 (fixed) & 0.71 & 57.9& \nodata\\
733: & & Exp & (0.53,0.36) & 18.81 & \nodata & 5.48 & 13.8 & 1 (fixed) & 0.82 & 27.7& \nodata\\
734: & S & PSF & (0,0) & 17.02 & \nodata & \nodata & \nodata & \nodata & \nodata & \nodata & 3.91 \\
735: & & S & ($-$0.03,0.04) & 17.18 & \nodata & 2.47 & 6.23 & 4.95 & 0.71 & 55.9 & \nodata\\
736: & deV & PSF & (0,0) & 17.17 & \nodata & \nodata & \nodata & \nodata & \nodata & \nodata & 3.93\\
737: & & deV & ($-$0.03,0.04) & 17.27 & \nodata & 2.05 & 5.16 & 4 (fixed) & 0.71 & 56.2& \nodata\\
738: & deV (D03) & PSF & \nodata & \nodata & 18.1 & \nodata & \nodata & \nodata & \nodata& \nodata & \nodata\\
739: & & deV & \nodata & \nodata & 16.8 & 2.28 & 5.73 & 4 (fixed) & 0.69 & 56& \nodata\\
740: & S (D03) & S & \nodata & \nodata & \nodata & \nodata & \nodata & 5.6 & \nodata & \nodata & \nodata \\
741: \enddata
742: \tablecomments{Col. (1): QSO and, in parentheses, the size of the
743: fitted area in pixels. Col. (2) GALFIT model (deV = de Vaucouleurs, Exp = Exponential, S = S{\'e}rsi{\'c}).
744: Col. (3): Individual components used.
745: Col. (4): Offsets with respect to the PSF. Col. (5): Integrated
746: apparent magnitude in the F606W filter. Note that the magnitudes $m_{\rm F606W}$
747: given for the PSF underestimate the AGN component as the PSF was saturated
748: in all cases; see text for details.
749: Col. (6): Integrated apparent
750: magnitude converted from F675W to Cousins $R$ band as given by D03.
751: Col. (7,8): Effective radius in \arcsec~and kpc, respectively.
752: Note that although we concluded in Section~\ref{process} that
753: the central region within a radius of $\sim$1.7\arcsec~is
754: still strongly affected by the PSF subtraction, all $r_{\rm eff}$ given
755: in this table are more than 10 times the full-width-at-half-maximum (FWHM)
756: of the ACS PSF.
757: Col. (9): S{\'e}rsi{\'c} index. Col. (10): Axis ratio.
758: Col. (11): Position angle (east of north). Col. (12): $\chi^2$
759: per degree of freedom, $\nu$. $\chi^2_{\nu}$
760: corresponds to the model in column 2.\\
761: Results from D03 are listed for
762: comparison. The $r_{\rm eff}$ given here for the D03 results
763: were recalculated according to the worldmodel
764: adopted in this paper. The P.A. given here for the D03 results were derived
765: by adding the P.A. given in their Table 3 to the orientation of the spacecraft
766: (as the P.A. given in their Table 3 is apparently not corrected for it).
767: }
768: \label{results}
769: \end{deluxetable}
770:
771:
772: \begin{deluxetable}{lccccccccccc}
773: \tabletypesize{\scriptsize}
774: \tablenum{2 - continued}
775: \tablecolumns{13}
776: \tablewidth{0pc}
777: \tablecaption{}
778: \tablehead{
779: \colhead{QSO} & \colhead {Model} &
780: \colhead{Function} & \colhead{$\Delta$($\alpha$,$\delta$)} & \colhead{$m_{\rm F606W}$}
781: & \colhead{$R$} & \colhead{$r_{\rm eff}$} & \colhead{$r_{\rm eff}$} & \colhead{S{\'e}rsi{\'c}} & \colhead{$b/a$} &
782: \colhead{P.A.} & \colhead{$\chi_\nu^2$} \\
783: \colhead{(fitted area)}& & & \colhead{($\arcsec$, $\arcsec$)} & \colhead{(mag)} & \colhead{(mag)} & \colhead{($\arcsec$)} & \colhead{(kpc)} &
784: \colhead{index} & & \colhead{($\degr$)}\\
785: \colhead{(1)} & \colhead{(2)} & \colhead{(3)} & \colhead{(4)} & \colhead{(5)}
786: & \colhead{(6)} & \colhead{(7)} & \colhead{(8)} & \colhead{(9)} & \colhead{(10)} & \colhead{(11)} & \colhead{(12)} }
787: \startdata
788: OX\,169 & deV+Exp & PSF & (0,0) & 17.80 & \nodata & \nodata & \nodata & \nodata & \nodata & \nodata & 3.45\\
789: (900$\times$900) & & deV & (0,0) & 17.68 & \nodata & 1.82 & 6.17 & 4 (fixed) & 0.47 & 120.4& \nodata\\
790: & & Exp & (0,0.15) & 19.34 & \nodata & 1.23 & 4.17 & 1 (fixed) & 0.66 & 18.5& \nodata \\
791: & S & PSF & (0,0) & 17.54 & \nodata & \nodata & \nodata & \nodata & \nodata & \nodata & 3.66\\
792: & & S & (0,0) & 17.42 & \nodata & 1.30 & 4.41 & 3.9 & 0.64 & 122.9 & \nodata\\
793: & deV & PSF & (0,0) & 17.56 & \nodata & \nodata & \nodata & \nodata & \nodata & \nodata & 3.66 \\
794: & & deV & (0,0) & 17.41 & \nodata & 1.29 & 4.36 & 4 (fixed) & 0.64 & 123.0 & \nodata \\
795: & deV (D03) & PSF & \nodata & \nodata & 15.9 & \nodata & \nodata & \nodata & \nodata& \nodata & \nodata\\
796: & & deV & \nodata & \nodata & 17.3 & 1.88 & 6.35 & 4 (fixed) & 0.47 & 121& \nodata \\
797: & S (D03) & S & \nodata & \nodata & \nodata & \nodata & \nodata & 3.6 & \nodata & \nodata & \nodata \\
798: \enddata
799: \tablecomments{Col. (1): QSO and, in parentheses, the size of the
800: fitted area in pixels. Col. (2) GALFIT model (deV = de Vaucouleurs, Exp = Exponential, S = S{\'e}rsi{\'c}).
801: Col. (3): Individual components used.
802: Col. (4): Offsets with respect to the PSF. Col. (5): Integrated
803: apparent magnitude in the F606W filter. Note that the magnitudes $m_{\rm F606W}$
804: given for the PSF underestimate the AGN component as the PSF was saturated
805: in all cases; see text for details.
806: Col. (6): Integrated apparent
807: magnitude converted from F675W to Cousins $R$ band as given by D03.
808: Col. (7,8): Effective radius in \arcsec~and kpc, respectively.
809: Note that although we concluded in Section~\ref{process} that
810: the central region within a radius of $\sim$1.7\arcsec~is
811: still strongly affected by the PSF subtraction, all $r_{\rm eff}$ given
812: in this table are more than 10 times the full-width-at-half-maximum (FWHM)
813: of the ACS PSF.
814: Col. (9): S{\'e}rsi{\'c} index. Col. (10): Axis ratio.
815: Col. (11): Position angle (east of north). Col. (12): $\chi^2$
816: per degree of freedom, $\nu$. $\chi^2_{\nu}$
817: corresponds to the model in column 2.\\
818: Results from D03 are listed for
819: comparison. The $r_{\rm eff}$ given here for the D03 results
820: were recalculated according to the worldmodel
821: adopted in this paper. The P.A. given here for the D03 results were derived
822: by adding the P.A. given in their Table 3 to the orientation of the spacecraft
823: (as the P.A. given in their Table 3 is apparently not corrected for it).
824: }
825: \end{deluxetable}
826:
827: \begin{deluxetable}{lccc}
828: \tabletypesize{\small}
829: \tablecolumns{4}
830: \tablewidth{0pc}
831: \tablecaption{Contribution of Fine Structure to Host Galaxy Light}
832: \tablehead{
833: \colhead{QSO} & \colhead{fs (3\,$r_{\rm eff}$)} & \colhead{$\Sigma_{\rm fs}$ (3\,$r_{\rm eff}$)} & \colhead{$\Sigma_{\rm fs}$ (max)}\\
834: & \colhead{$\%$} & \colhead{mag\,\,arcsec$^{-2}$} & \colhead{mag\,\,arcsec$^{-2}$}\\
835: \colhead{(1)} & \colhead{(2)} & \colhead{(3)} & \colhead{(4)}}
836: \startdata
837: PHL\,909 & 7.8 & 29.2 & 24.8\\
838: PKS\,0736+01 & 8.2 & 29.6 & 26.5 \\
839: MC2\,1635+119 & 6.3 & 28.3 & 25.2 \\
840: OX\,169 & 13.3 & 29.5 & 23.7 \\
841: \enddata
842: \tablecomments{
843: Col (1): QSO. Col. (2): Percentage of flux contained
844: in the observed fine structure within 3\,$r_{\rm eff}$.
845: Col. (3): Surface-brightness magnitude
846: of the fine structure in the F606W filter within 3\,$r_{\rm eff}$.
847: Col. (4): Average surface-brightness magnitude within an area of
848: 2\arcsec$\times$2\arcsec~around the maximum
849: of the fine structure.
850: }
851: \label{percentage}
852: \end{deluxetable}
853:
854: \onecolumn
855: \begin{figure}
856: \epsscale{1}
857: \plotone{f2_small.eps}
858: \caption{Residual images derived
859: by subtracting a GALFIT model consisting of a single S{\'e}rsi{\'c}
860: profile.
861: In each image, north is up, east is to the left.
862: MC2\,1635+119 was presented in Paper I.
863: }\label{finalall}
864: \end{figure}
865: \twocolumn
866:
867: \onecolumn
868: \begin{figure}
869: \epsscale{0.9}
870: \plotone{f3.eps}
871: \caption{Surface-brightness plots of observed and best-fitting model profiles
872: for all five QSO host galaxies.
873: The observed profile is shown as diamonds with error bars.
874: The resulting profile of fitting a single S{\'e}rsi{\'c}
875: profile is shown as a solid line.
876: The residuals (fit $-$ data) are shown in the lower panels.
877: }\label{finalsbp}
878: \end{figure}
879: \twocolumn
880:
881: \subsection{Fine Structure}
882: To estimate the luminosity within the various types of
883: fine structure compared to the total luminosity
884: of the host galaxy, we
885: created a mask that included all the light within
886: an annulus with an inner radius 1.7\arcsec~and an outer radius
887: 3\,$r_{\rm eff}$ (as taken from a single de Vaucouleurs
888: fit, see Table~\ref{results}).
889: The image was multiplied by
890: this mask (good=1, bad=0), and the total counts in
891: the product were summed. This was done both for the GALFIT residual image
892: ($f_{\rm fine structure}$) obtained by subtracting the GALFIT model
893: of a de Vaucouleurs + exponential profile, and for the GALFIT model itself
894: ($f_{\rm galaxy}$). We then computed the ratio $f_{\rm fine structure}$/$f_{\rm galaxy}$.
895: The fractional luminosity of the extended fine structure in such an annulus
896: ranges between 2\% and 5\%.
897: To account for regions that were over-subtracted by GALFIT,
898: we also used a slightly different approach:
899: Applying the same general method as above, we only summed
900: the pixels that are $> -1 \sigma$ of the sky background.
901: Then, the fractional luminosity of the extended fine structure
902: ranges between 6\% and 13\%.
903: The results are summarized in Table~\ref{percentage}.
904:
905: However, note that these values give only the average fraction within 3\,$r_{\rm eff}$.
906: Locally, the surface-brightness magnitude of the fine structure relative to
907: the underlying model can be much higher.
908: The surface-brightness magnitude within an area of 2\arcsec$\times$2\arcsec~centered
909: on the maximum of the fine structure lies
910: between $\sim$24 and $\sim$26 mag\,arcsec$^{-2}$ (as derived from the GALFIT residual image)
911: for the four objects with fine structure (Table~\ref{percentage}).
912:
913:
914: \section{DISCUSSION}
915: \label{mergerage}
916: We find significant fine structure in the deep {\em HST} ACS images
917: of at least four of the five QSO host galaxies
918: (PHL\,909, PKS\,0736+01, MC2\,1635+119, and OX\,169) in our pilot sample.
919: While the prominent linear structure in OX\,169 has been previously reported by several
920: authors \citep{sto78,hut84,geh84,smi86,hec86,mcl99}, the structure in the other
921: three objects is revealed here for the first time.
922: Only one of the objects studied, PG\,0923+201, does not show
923: any obvious fine structure.
924:
925: Our findings are supported by the
926: WFPC2 images of D03 and \citet{mcl99}
927: since some of the structure that we find is also visible in
928: their images:
929: Part of the north-western tidal tail in PHL\,909 is visible (Fig.\ A11,
930: \citealt{mcl99});
931: some of the spiral-like structure in PKS\,0736+01 can be seen
932: (Fig.\ A6 in \citealt{mcl99}); and the shell in MC\,1635+119 labeled
933: ``d'' in Fig.~3 of Paper I
934: is visible to the north-east of the QSO nucleus
935: at a distance of $\sim$4\arcsec~(Fig.\ A18 in \citealt{mcl99}).
936: However, the depth reached in our ACS images
937: is needed to positively identify the ``fuzz'' seen in these
938: earlier images as real.
939: Our results clearly show the need of high S/N imaging for an
940: accurate interpretation
941: of QSO host galaxy morphologies.
942:
943: The host galaxy of PG\,0923+201 does not show obvious signs
944: of fine structure. However, the surface-brightness profile reveals
945: a bump in the inner 2\arcsec~that could reflect the presence
946: of an inner disk. Interestingly, PG\,0923+201 resides in a galaxy group of
947: at least 6 galaxies. It is the only QSO in our sample that has
948: been observed with the {\em Spitzer} Space Telescope, both
949: with the Infrared Spectrograph (IRS) and the
950: Multiband Imaging Photometer for Spitzer (MIPS; 24$\mu$m).
951: Neither FIR emission nor Polycyclic-Aromatic Hydrocarbon (PAH) features
952: have been detected \citep{shi07}.
953: Interestingly, this lack of signs of star-formation activity in the infrared
954: goes hand-in-hand with the lack
955: of obvious signs for recent merger activity.
956:
957: Much of the fine structure observed in our deep {\em HST} ACS
958: images of the other four QSO host galaxies is indicative of merger events in the
959: relatively recent past.
960: The possible type and timescales for the merger
961: in MC2\,1635+119 were discussed in detail in Paper I.
962: Briefly, we found that the fine structure seen in the host galaxy of this object
963: was most likely produced by a strong tidal interaction within the last
964: $\sim$1\,Gyr. On the one hand, the regularity of the inner shell structure
965: is suggestive of a near radial collision of a dwarf galaxy
966: with a giant elliptical galaxy in a minor merger.
967: On the other hand, the arcs observed at much larger distances than
968: the inner shells, and other tidal debris off-axis from the direction of
969: the encounter implied by the inner shells, are more indicative of a major
970: merger. Using simple $N$-body simulations, we estimated that the time scale
971: needed to form the observed structure
972: is less than $\sim$1.7 Gyr (see Paper I for details).
973: In comparison, the spectrum of the host galaxy is dominated
974: by a population of intermediate-age ($\sim$1.4 Gyr) stars, indicative of
975: a strong starburst that may have occurred during the merger event.
976: Thus, the observed QSO activity
977: may have been triggered in the recent past either by a minor merger
978: or by debris from an older ($\sim$Gyr) major merger that is currently
979: ``raining'' back into the central regions of the merger remnant.
980:
981: The prominent extended structure observed in OX\,169 is most likely
982: a tidal tail seen edge on.
983: Prominent tidal tails are also observed in the host of PHL\,909.
984: The tails in both QSO hosts
985: are strongly suggestive of relatively recent major merger events,
986: where at least some of the parent galaxies were disk galaxies.
987:
988: The structure surrounding PKS\,0736+201 resembles that of the
989: Seyfert 1.5 galaxy NGC\,5548, an Sa galaxy with ripples, tidal arms and a faint
990: long tidal tail that has been described in detail
991: by \citet{sch88}. The observed structure in NGC\,5548 has been interpreted
992: in terms of the spatial wrapping of material from an obliquely infalling
993: companion. When seen in projection, such material can appear in the form of
994: relatively sharp-edged ripples \citep{sch88}.
995: Compared with NGC\,5548, the faint spiral-like structure
996: seen in PKS\,0736+01 is more extended ($\sim$ 50\,kpc vs. 10\,kpc)
997: and more regular.
998:
999: Structure similar to that
1000: observed in PKS 0736+201 can be reproduced
1001: in numerical simulations of minor mergers.
1002: For example, the merger of a dwarf elliptical galaxy
1003: with a large spiral and a total mass ratio of 1:8
1004: results in very similar structure $\sim$1\,Gyr after the
1005: first passage of the two galaxies, but also again
1006: at later stages in the merger event (\citealt{you07};
1007: T. J. Cox, private communication).
1008:
1009: By comparing the structure observed in these hosts to
1010: results from numerous published $N$-body simulations of mergers
1011: (e.g., \citealt{hib95}), we know that the ages of these
1012: tidal features are likely of the order of a few orbital times.
1013: Given the masses of giant elliptical galaxies and
1014: the distances at which we observe these features,
1015: this time span ranges from a few hundred Myr to about 1 Gyr.
1016: In the cases in which we can clearly see tidal tails, we can infer
1017: that the dynamical timescales cannot be much less than $\sim 250$ Myr,
1018: since that is roughly the amount of time that it would take for stars with
1019: typical orbital velocities ($\sim 200$\,km\,s$^{-1}$) to travel in a straight
1020: line to the projected maximum distances at which we observe the features
1021: (e.g., $\sim$ 50 kpc for PHL\,909 or $\sim$ 42 kpc for OX\,169).
1022: On timescales longer than $\sim$1 Gyr, tidal tails are expected to rapidly
1023: disperse and lose contrast.
1024:
1025: Further evidence that these are relatively evolved mergers is the fact
1026: that enough time has elapsed since the initial tidal encounter
1027: for each host to acquire an essentially elliptical
1028: morphology. While it is possible that some of the objects,
1029: such as MC\,1635+119 and PG\,0923+201, may have started off as
1030: elliptical galaxies, at least OX\,169 and PHL\,909 must have
1031: resulted from the merger of disk galaxies. The fact that
1032: they are at the present time reasonably well fitted by de Vaucouleurs
1033: profiles implies merger timescales on the order of a Gyr.
1034:
1035: To address the question of whether the AGN activity may
1036: indeed have been triggered by the merger events, we need to obtain
1037: estimates of the different timescales involved.
1038: How do the ages of the stellar populations compare to the time elapsed since
1039: the merger events, and what does that imply for the QSO duty cycle?
1040: What is the connection between
1041: merger, star formation, and BH accretion?
1042:
1043: As mentioned above, the timescales for the merger events observed in
1044: our sample is of the order of a Gyr or less. G. Canalizo \& A. Stockton
1045: (2008, in preparation) find similar timescales for strong starburst events in
1046: the host galaxies of these QSOs. By modeling deep Keck spectra of these
1047: objects, they find traces of major starburst episodes in intermediate-age
1048: starburst populations that involve a substantial fraction of the stellar mass.
1049: These starburst ages range from $\sim0.7$ Gyr (OX\,169) to
1050: 2.2 Gyr (PKS\,0736+017). Hence, it seems possible that the starbursts were
1051: triggered by the merger events.
1052:
1053: The most difficult timescale to estimate is that of the duration
1054: of the AGN activity.
1055: Current estimates of QSO lifetimes range between 10$^6$ and
1056: 10$^8$ years and are derived primarily from demographics
1057: (see review by \citealt{mar04}).
1058: For example, from theoretical calculations \citet{yu02}
1059: estimate a mean lifetime
1060: for a luminous QSO ($L_{\rm bol}$ $\ge$ 10$^{46}$\,erg\,s$^{-1}$)
1061: of (3--13) $\times$ 10$^7$ yr which is comparable
1062: to the Salpeter time scale ($\sim$ 4.5 $\times$ 10$^7$ yr
1063: for $\epsilon$ = 0.1 and $L/L_{\rm edd}$ = 1; \citealt{sal64,mar04}).
1064: Apart from demographics, there have been several
1065: attempts to deduce the AGN duty cycle for individual galaxies.
1066: For example, the length of radio jets
1067: can yield a lower estimate of QSO lifetimes,
1068: depending on the expansion speed of the jet
1069: \citep[e.g.,][]{wil78,sch95,blu99}. However, this method cannot
1070: be applied to our sample as the two radio-loud QSOs in the sample
1071: each feature only a compact flat-spectrum radio source.
1072: Also, the size of the NLR has been proposed as a
1073: straightforward geometric measure
1074: of AGN lifetimes \citep{mar04}.
1075: However, the NLR sizes for the five QSOs of our
1076: sample have not yet been determined.
1077:
1078: Thus, as we do not have a direct way to estimate the time elapsed since
1079: the triggering of the nuclear activity for the objects in our sample, it is
1080: difficult to work out the physical relationship between the mergers we detect
1081: and the fueling of the QSOs. An overly simplistic scenario, in which each
1082: QSO was triggered by the merger at the time of the starburst episode would
1083: imply QSO lifetimes on the order of a Gyr, in stark contrast to the
1084: theoretical estimates cited above. Instead, our observations may provide
1085: evidence in support of significant time
1086: delays between tidal interactions and the fueling of the
1087: central black holes, as predicted by hydrodynamic simulations
1088: (see, e.g., \citealt{bar98,spr05b,hop07b}).
1089:
1090:
1091: However, while QSOs often reside in bulge-dominated galaxies like
1092: the ones presented in this study, they are also frequently found in
1093: young mergers (e.g., \citealt{urr07,guy06,can01,hut94}). If, as the
1094: accumulating evidence suggests, mergers are indeed essential for
1095: the triggering of QSOs, then our results imply at least one of two
1096: scenarios:
1097: Either QSO activity is episodic \citep[e.g.,][]{nor88,hop06} and occurs
1098: over longer timescales than previously speculated, or there is a large range of
1099: values for the time delays between the merger and the onset of the nuclear
1100: activity.
1101:
1102:
1103: \acknowledgments
1104: We thank the anonymous referee for the valuable comments
1105: helping to improve the paper.
1106: Support for the {\em HST} program 10421 was provided by NASA through
1107: a grant from the Space Telescope Science Institute,
1108: which is operated by The Association of Universities for
1109: Research in Astronomy, Inc., under NASA contract No.\ NAS526555.
1110: Additional support was provided by the National Science Foundation,
1111: under grant AST 0507450.
1112: B. Jungwiert is also supported by the grant LC06014 of the Czech Ministry
1113: of Education and by Research Plan AV0Z10030501 of the Academy
1114: of Sciences of the Czech Republic.
1115: C. Y. Peng is grateful for support from the HIA Plaskett
1116: fellowship and the STScI Institute/Giacconi fellowship.
1117: This research has made use of the NASA/IPAC Extragalactic Database (NED)
1118: which is operated by the Jet Propulsion Laboratory,
1119: California Institute of Technology,
1120: under contract with the National Aeronautics and Space Administration.
1121:
1122:
1123:
1124:
1125:
1126: \appendix
1127:
1128: \section{NOTES ON INDIVIDUAL OBJECTS}
1129: \label{individual}
1130: Here, we give a more detailed description of
1131: each of the four QSOs.
1132: (Note that MC2\,1635+119 was discussed in detail in Paper I.)
1133: Specifically, we review and discuss published
1134: data, compare them with our findings, and include
1135: a description of neighboring objects.
1136: The redshifts of companion galaxies and other galaxies within
1137: the ACS FOV are taken from NED unless stated
1138: otherwise. The full ACS/WFC images are shown in Figure~\ref{acs},
1139: with labels for the galaxies to which we refer in the following discussion.
1140: Note that we cannot show detailed images for each such galaxy;
1141: instead, we refer the interested reader to the {\em HST} archive.
1142:
1143: \subsection{PHL\,909}
1144: Based on older images, the host galaxy of the radio-quiet QSO PHL\,909
1145: (z=0.172, 1\arcsec~$\simeq$ 2.89\,kpc)
1146: was described as an elliptical galaxy \citep{bah96,bah97,tay96,mcl99,
1147: ham02}.
1148: The age of the stellar population from
1149: off-nuclear optical spectra is not well constrained,
1150: but the spectra are suggestive of an old age \citep{nol01}.
1151: Note, however, that \citet{hug00} report a significant
1152: increase of flux bluewards of a weak 4000-\AA~break,
1153: but cannot determine whether it is due to scattered
1154: light from the QSO or indicative of a younger stellar
1155: population.
1156:
1157: BH mass estimates (making use
1158: of $\sigma$, FWHM$_{\rm H\beta}$ and $L_{\rm 5100}$,
1159: $R$-band luminosity of the host spheroid)
1160: cover a
1161: range of 0.17--2.5 $\times$ 10$^9$ $M_{\odot}$ \citep{mcl01, wu02, dun03, wu04}.
1162: \citet{dun03} predict an Eddington ratio ($L_{\rm nuc}$/$L_{\rm edd}$) of 5.2\%.
1163: \citet{chu06} classify PHL\,909 as Lyman limit system.
1164: They mention a number of faint small objects surrounding
1165: the QSO that could be companions. In a K-band image, the galaxy extends
1166: $\sim$10\arcsec~towards a western companion \citep{dun93}.
1167: The optical spectrum of PHL\,909 shows double-peaked
1168: Balmer emission lines \citep{str03}.
1169: In the radio, PHL\,909 is an unresolved point source
1170: \citep{kuk98}.
1171:
1172: The arc-like features at $\simeq$2\arcsec~from the QSO
1173: mentioned by \citet{bah96}
1174: correspond most likely to the
1175: central ring- or disk-like structure ($r \simeq$1.8\arcsec~-- 2.5\arcsec)
1176: seen in our deep {\em HST} ACS image.
1177: However, \citet{bah96} were unable to determine whether the features
1178: were artifacts due to the much lower S/N of their images.
1179:
1180: The neighborhood of PHL\,909 is populated with several
1181: objects, some of which look like dwarf galaxies.
1182: Seven brighter, nearby galaxies make PHL\,909
1183: appear to reside in a galaxy group.
1184: However, redshifts are known for only 3 objects,
1185: of which only one object is at the same redshift as the QSO,
1186: the large spiral galaxy (``h'') to the S
1187: (P.A.~$\simeq$ 183$\degr$) 28\arcsec~away ($z$ = 0.169).
1188: The two other objects are foreground galaxies,
1189: one an elliptical galaxy (``b''; $z$ = 0.102)
1190: to the NE (P.A.~$\simeq$ 28$\degr$) at a distance of 18\arcsec,
1191: and one a spiral (``e''; $z$ = 0.124) to the SE
1192: (P.A.~$\simeq$ 140$\degr$) 20\arcsec~away.
1193: Two lenticular objects with unknown redshift
1194: lie at 31\arcsec~E (``c''; P.A.~$\simeq$ 96$\degr$)
1195: and 12\arcsec~S (``f''; P.A.~$\simeq$ 173$\degr$) projected
1196: distance of the QSO.
1197: A spiral galaxy to the SW (``g''; P.A.~$\simeq$ 200$\degr$)
1198: is 20\arcsec~away.
1199: Another bright object is the galaxy that is apparently interacting
1200: with the QSO (see Section~\ref{morphology}).
1201: In the full FOV of the ACS frame (Fig.~\ref{acs}),
1202: redshifts are known for two other galaxies, both
1203: foreground spirals (``i'', $z$ = 0.104, P.A.~$\simeq$ 29.3$\degr$,
1204: distance $\simeq 1.12\arcmin$ and ``j'', $z$ = 0.082, P.A.~$\simeq$ 76$\degr$,
1205: distance $\simeq 2.34\arcmin$).
1206:
1207: \subsection{PKS\,0736+01}
1208: PKS\,0736+01 (z=0.191, 1\arcsec~$\simeq$ 3.14\,kpc)
1209: was classified as a blazar by \citet{ang80},
1210: as a compact flat-spectrum radio quasar by \citet{wal85},
1211: and as an optically violently variable by \citet{bro89}.
1212: The host galaxy was described as an elliptical by \citet{wri98,mcl99,fal00},
1213: and \citet{ham02}.
1214: Off-nuclear optical spectra have been obtained
1215: and fitted by \citet{hug00} and \citet{nol01},
1216: indicating an age of 12 Gyr.
1217: The estimated BH mass
1218: (from relations with $\sigma$, FWHM$_{\rm H\beta}$ and $L_{\rm 5100}$,
1219: $R$-band luminosity of the host spheroid) lies between
1220: 0.14 and 1.3 $\times$ 10$^9$ $M_{\odot}$ \citep{mcl01, wu02, dun03}.
1221: \citet{dun03} give 3.5\% as an estimate of the Eddington ratio.
1222: The optical {\em HST} WFPC2 image of \citet{mcl99} shows three
1223: companions (two to the NW and one to the SE),
1224: but fails to reveal the low surface-brightness nebulosity,
1225: described as highly disturbed, seen in the K-band \citep{dun93}.
1226: In the near-infrared H-band, the host is resolved but round \citep{kot98}.
1227:
1228: Several faint objects surrounding PKS\,0736+01 underlie
1229: the isophotes of the ``spiral''.
1230: Two of these are background objects
1231: as determined from Keck spectra
1232: (G. Canalizo \& A. Stockton 2008, in preparation): a lenticular-shaped
1233: galaxy at 5.9\arcsec~to the SE (``a''; Fig.~\ref{acs};
1234: P.A.~$\simeq$ 147$\degr$; $z$ = 0.8500)
1235: and an irregular galaxy at a distance of
1236: 4.4\arcsec~to the NW (``b''; P.A.~$\simeq$ 307$\degr$; $z$ = 0.3785).
1237: The latter shows a tidal tail that
1238: is extending toward another object to the NW
1239: (``c''; P.A.~$\simeq$ 297$\degr$; distance $\simeq$ 9\arcsec) of
1240: unknown redshift.
1241:
1242: Whether the observed faint spiral-like structure
1243: in PKS\,0736+01
1244: indeed indicates the presence of a faint spiral host galaxy,
1245: whether it is formed from debris
1246: of a merger event as speculated in Section~\ref{mergerage},
1247: or whether it is actually a combination of both
1248: (a spiral disturbed by accreted material),
1249: our observations clearly show that the host galaxy
1250: of PKS\,0736+01 cannot be considered to be an undisturbed elliptical galaxy.
1251:
1252: \subsection{PG\,0923+201}
1253: The radio-quiet QSO PG\,0923+201 (z=0.19, 1\arcsec~$\simeq$ 3.12\,kpc)
1254: resides in an elliptical host galaxy \citep{bah97,mcl99,mcl00,ham02,guy06}.
1255: It is not detected as a radio source \citep{kuk98}.
1256: From a noisy optical off-nuclear spectrum,
1257: \citet{nol01} estimate a stellar-population age
1258: of about 12 Gyr.
1259: BH mass estimates (using $\sigma$, FWHM$_{\rm H\beta}$ and $L_{\rm 5100}$,
1260: $R$-band luminosity of the host spheroid) yield
1261: $M_{\rm BH}$ $\simeq$ 0.11--2.6 $\times$ 10$^9$ $M_{\odot}$
1262: \citep{mcl01, wu02, dun03}.
1263: \citet{dun03} estimate an
1264: Eddington ratio $L_{\rm nuc}$/$L_{\rm edd}$ of 8.8\%.
1265:
1266: PG\,0923+201 is a member of a galaxy group:
1267: In the ACS FOV, at least 5 galaxies
1268: are at a comparable
1269: redshift of $z \simeq 0.19$,
1270: (pair ``a'', \citet{hec84}; ``c'', \citet{ell91}; pair ``d'',
1271: Keck spectra; G. Canalizo \& A. Stockton 2008, in preparation),
1272: four of which form two interacting pairs (``a'', ``d''; Fig.~\ref{acs}).
1273: The closest pair (``a'') at $\sim$ 12\arcsec~to the east
1274: consists of two large elliptical galaxies
1275: [one galaxy at 10\arcsec~(P.A.~$\simeq$ 153$\degr$) and
1276: one galaxy at 15\arcsec~~(P.A.~$\simeq$ 161$\degr$) to the SE].
1277: \citet{hut89} speculated that the QSO may be interacting
1278: with these. The northern galaxy of this pair shows a shell-like
1279: structure.
1280: The other pair (``d'') lies at a projected distance of $\simeq 1.6\arcmin$ to
1281: the north-east (P.A.~$\simeq$ 21$\degr$). Both galaxies have
1282: clear tidal tails on one side and are connected by
1283: a tidal bridge.
1284: The fifth galaxy (``c'') at a comparable redshift lies 18\arcsec\ to
1285: the east (P.A.~$\simeq$ 83$\degr$), a small galaxy with overall elliptical
1286: morphology but faint extended tail-like structures.
1287: The tidal tails make it resemble a more distant and faint
1288: version of the Antennae galaxies
1289: (NGC4038/4039), the best-studied and
1290: nearest ($\sim$ 19.2 Mpc) example of a major interaction
1291: between two massive gas-rich spiral galaxies.
1292: However, this object is possibly
1293: at a later stage of the merger as the two nuclei
1294: seem to have coalesced. The SE arm
1295: ends in a knot, a compact object that could be an
1296: interacting dwarf galaxy.
1297:
1298: Note that within the ACS FOV,
1299: another galaxy pair (``e'') at a distance of $\simeq 2\arcmin$~to the east
1300: (P.A.~$\simeq$ 92$\degr$) is at an advanced stage of merging,
1301: with curved tidal tails and the nuclei appearing disrupted in several
1302: fragments.
1303: While \citet{ell91} list this pair at a redshift of $z = 0.1899$,
1304: Keck spectra show that it is rather at a redshift of $z = 0.2318$
1305: (G. Canalizo \& A. Stockton 2008, in preparation).
1306: The redshift is known for one other galaxy within the ACS FOV, a background spiral
1307: (``f'', $z$ = 0.2328, P.A.~$\simeq$ 87$\degr$,
1308: distance $\simeq 2.6\arcmin$).
1309: There is another very close galaxy to the SW
1310: (P.A.~$\simeq$ 224$\degr$) of the QSO,
1311: (``b''; distance $\simeq$ 11\arcsec)
1312: with peculiar, tidally disrupted morphology, but unknown redshift.
1313:
1314: There are at least two arcs close to the QSO (labeled as ``1'' and ``2''
1315: in Fig.~\ref{acs}) that could be gravitationally lensed galaxies.
1316: However, they may also be tidal debris, especially in the case of ``1'',
1317: which lies close to the interacting pair marked ``a''.
1318:
1319:
1320: \subsection{OX\,169}
1321: The host galaxy of the radio-loud QSO OX\,169 (z=0.211, 1\arcsec~$\simeq$ 3.39\,kpc)
1322: has been described as being an elliptical \citep{hut92,tay96,mcl99}.
1323: \citet{nol01} infer an old age for the stellar population of
1324: OX\,169 from their optical off-nuclear spectrum, although
1325: the spectrum has a low S/N ratio.
1326: BH mass estimations using different relations
1327: ($\sigma$, FWHM$_{\rm H\beta}$ and $L_{\rm 5100}$,
1328: $R$-band luminosity of the host spheroid) cover a
1329: range of 0.22--1.7 $\times$ 10$^9$ $M_{\odot}$ \citep{mcl01, wu02, dun03}.
1330: The Eddington ratio is estimated to be 6.1\% \citep{dun03}.
1331: From its radio emission, OX\,169 is classified as a compact ($< 4\arcsec$)
1332: flat-spectrum source \citep{fei84}.
1333:
1334: OX\,169 seems to have several faint satellite galaxies,
1335: but their redshifts are unknown.
1336: The closest one at 3.7\arcsec~to the NE
1337: (``c''; P.A.~$\simeq$ 43$\degr$) merges with the outer isophotes of the QSO host galaxy and
1338: resembles a dwarf elliptical galaxy.
1339: Two others lie towards the SW at distances of
1340: 4.6\arcsec~(``b''; P.A.~$\simeq$ 245$\degr$) and
1341: 6.1\arcsec~(``a''; P.A.~$\simeq$ 235$\degr$). The former
1342: has an irregular shape and resembles a tidally disrupted
1343: galaxy, while the latter object is the companion already described in Section~\ref{morphology}.
1344:
1345: \onecolumn
1346: \begin{figure}
1347: \epsscale{1}
1348: \plotone{f4_small.eps}
1349: \caption{Full ACS/WFC images of four QSO host galaxies.
1350: The two bars mark the position of the QSO.
1351: North is up, east is to the left.
1352: The FOV of the ACS/WFC frame is 3.4\arcmin~$\times$ 3.4\arcmin.
1353: As inset, a smaller area around each QSO is shown,
1354: with the image size identical to the images
1355: in Figs.~\ref{final} and~\ref{finalall}.
1356: Letters mark the
1357: galaxies we refer to in
1358: the text. The letters refer to the closest object north
1359: or south of them. In the case of PG\,0923+201, the two numbers
1360: refer to the arcs east of them.
1361: }\label{acs}
1362: \end{figure}
1363: \twocolumn
1364:
1365:
1366: \begin{thebibliography}{}
1367: \bibitem[Angel \& Stockman(1980)]{ang80} Angel, J. R. P., \& Stigma, H. S. 1980, ARA\&A, 18, 321
1368: \bibitem[Bahcall et al.(1996)]{bah96} Bahcall, J. N., Kirhakos, S., \& Schneider, D. P. 1996,
1369: ApJ, 457, 557
1370: \bibitem[Bahcall et al.(1997)]{bah97} Bahcall, J. N., Kirhakos, S.,
1371: Saxe, D. H., \& Schneider, D. P. 1997, ApJ, 479, 642
1372: \bibitem[Barnes(1988)]{bar88} Barnes, J. E. 1988, ApJ, 331, 699
1373: \bibitem[Barnes(1998)]{bar98} Barnes, J. E. 1998, in Galaxies: Interactions
1374: and Induced Star Formation, ed. Friedli, D., Martinet, L., \& Pfenniger, D.
1375: (Berlin: Springer), 275
1376: \bibitem[Barthel(2006)]{bar06} Barthel, P. D. 2006, A\&A, 458, 107
1377: \bibitem[Blundell et al.(1999)]{blu99}
1378: Blundell, K. M., Rawlings, S., \& Willot, C. J. 1999, AJ, 117, 677
1379: \bibitem[Brown et al.(1989)]{bro89} Brown, L. M. L., et al. 1989, ApJ, 340, 129
1380: \bibitem[Canalizo \& Stockton(2001)]{can01} Canalizo, G., \& Stockton, A. 2001, ApJ, 555, 719
1381: \bibitem[Canalizo et al.(2006)]{can06} Canalizo, G., Stockton, A., Brotherton, M. S.,
1382: \& Lacy, M. 2006, NewAR, 50, 650
1383: \bibitem[Canalizo et al.(2007)]{can07} Canalizo, G., Bennert, N., Jungwiert, B., Stockton, A., Schweizer,
1384: F., Lacy, M., \& Peng, C. 2007, ApJ, 669, 801
1385: \bibitem[Chun et al.(2006)]{chu06} Chun, M. R., Gharanfoli, S., Kulkarni, V. P., \& Takamiya, M. 2006,
1386: AJ, 131, 686
1387: \bibitem[da Angela et al.(2007)]{daa07} da Angela, J. et al. 2007, MNRAS, accepted (arXiv:astro-ph/0612401)
1388: \bibitem[de Vaucouleurs(1948)]{dev48} de Vaucouleurs, G. 1948, Ann. d'Astrophys., 11, 247
1389: \bibitem[Disney et al.(1995)]{dis95} Disney, M. J., et al. 1995, Nature, 376, 150
1390: \bibitem[Dressler et al.(1985)]{dre85} Dressler, A., Thompson, I. B., \&
1391: Shectman, S. A. 1985, ApJ, 288, 481
1392: \bibitem[Dunlop et al.(1993)]{dun93}
1393: Dunlop, J. S., Taylor, G. L., Hughes, D. H., \& Robson, E. I. 1993,
1394: MNRAS, 264, 455
1395: \bibitem[Dunlop et al.(2003)]{dun03}
1396: Dunlop, J. S., McLure, R. J., Kukula, M. J., Baum, S. A.,
1397: O'Dea, C. P., \& Hughes, D. H. 2003, MNRAS, 340, 1095 (D03)
1398: \bibitem[Ellingson et al.(1991)]{ell91} Ellingson, E., Green, R. F., \& Yee, H. K. C. 1991,
1399: ApJ, 378, 476
1400: \bibitem[Evans et al.(2001)]{eva01} Evans, A. S., Frayer, D. T., Surace, J. A.,
1401: \& Sanders, D. B. 2001, AJ, 121, 1893
1402: \bibitem[Falomo \& Ulrich(2000)]{fal00} Falomo, R., \& Ulrich, M.-H. 2000,
1403: A\&A, 357, 91
1404: \bibitem[Feigelson et al.(1984)]{fei84} Feigelson, E. D., Isobe, T., \& Kembhavi, A.
1405: 1984, AJ, 89, 1464
1406: \bibitem[Floyd et al.(2004)]{flo04} Floyd, D. J. E., Kukula, M. J., Dunlop, J. S.,
1407: McLure, R. J., Miller, L., Percival, W. J., Baum, S. A., \& O'Dea, C. P. 2004, MNRAS,
1408: 355, 196
1409: \bibitem[Fort et al.(1986)]{for86} Fort B. P., Prieur J.-L., Carter D., Meatheringham S. J. \& Vigroux L., 1986, ApJ, 319, 110
1410: \bibitem[Gehren et al.(1984)]{geh84} Gehren, T., Fried, J., Wehinger, P. A., \& Wyckoff, S. 1984,
1411: ApJ, 278, 11
1412: \bibitem[Guyon, Sanders \- \& Stockton(2006)]{guy06} Guyon, O., Sanders, D. B., \& Stockton, A. 2006, ApJS, 166, 89
1413: \bibitem[Hamilton et al.(2002)]{ham02} Hamilton, T. S., Casertano,
1414: S., \& Trunshek, D. A. 2002, ApJ, 576, 61
1415: \bibitem[Heckman et \- al.(1984)]{hec84} Heckman, T. M., Bothun, G. D., Balick, B., \&
1416: Smith, E. P. 1984, ApJ, 89, 7
1417: \bibitem[Heckman \- et al.(1986)]{hec86} Heckman, T. M., Smith, E. P., Baum, S. A., van Breugel, W. J. M.,
1418: Miley, G. K., Illingworth, G. D., Bothun, G. D., \& Balick, B. 1986, ApJ, 311, 526
1419: \bibitem[Hibbard \& Mihos(1995)]{hib95} Hibbard, J. E., \& Mihos, J. C. 1995, ApJ, 110, 140
1420: \bibitem[Hibbard \- \& van Gorkum(1996)]{hib96} Hibbard, J. E., \& van Gorkom, J. H.\ 1996, AJ, 111, 655
1421: \bibitem[Hopkins et al.(2007)]{hop07a} Hopkins, P. F., Hernquist, L., Cox, T. J.,
1422: \& Keres, D. 2007, ApJ, submitted (arXiv:0706.1243v2)
1423: \bibitem[Hopkins et al.(2007)]{hop07b} Hopkins, P. F., Lidz, A., Hernquist, L., Coil, A. L.,
1424: Myers, A. D., Cox, TJ., Spergel, D. N., ApJ, 662, 110
1425: \bibitem[Hopkins et al.(2006)]{hop06} Hopkins, P. F., Hernquist, L., Cox, T. J., Di Matteo, T., Robertson, B., \& Springel, V. 2006, ApJS, 163, 1
1426: \bibitem[Hughes et al.(2000)]{hug00} Hughes, D. H., Kukula, M. J., Dunlop, J. S., \& Boroson, T.
1427: 2000, MNRAS, 316, 204
1428: \bibitem[Hutchings et al.(1984)]{hut84} Hutchings, J. B., Crampton, D., Campbell, B., Duncan, D., \&
1429: Glendenning, B. 1984, ApJS, 55, 319
1430: \bibitem[Hutchings et al.(1988)]{hut88} Hutchings, J. B., Johnson, I., \& Pyke, R. 1988, ApJS, 66, 361
1431: \bibitem[Hutchings et al.(1989)]{hut89}
1432: Hutchings, J. B., Janson, T., \& Neff, S. G. 1989, ApJ, 342, 660
1433: \bibitem[Hutchings \& Neff(1992)]{hut92} Hutchings, J. B., \& Neff, S. G. 1992, AJ, 104, 1
1434: \bibitem[Hutchings et al.(1994)]{hut94} Hutchings, J. B., Holtzman, J., Sparks, W. B., Morris, S. C., Hanisch, R. J., \& Mo, J. 1994, ApJ, 429, L1
1435: \bibitem[Jahnke et al.(2007)]{jah07} Jahnke, K., Wisotzki, L., Courbin, F. \& Letawe, G. 2007,
1436: MNRAS, 378, 23
1437: \bibitem[Kauffmann et al.(2003)]{kau03} Kauffmann, G., et al. 2003, MNRAS, 346, 1055
1438: \bibitem[Koekemoer et al.(2002)]{koe02} Koekemoer A. M., Fruchter,
1439: A. S., Hook, R. N., \& Hack, W. 2002, in Proc. 2002 {\em HST}
1440: Calibration Workshop, ed. S. Arribas, A. Koekemoer, \& B.
1441: Whitmore (Baltimore: STScI), 337
1442: \bibitem[Kotilainen et al.(1998)]{kot98} Kotilainen, J. K., Falomo, R., \& Scarpa, R. 1998, A\&A, 332, 503
1443: \bibitem[Kukula et al.(1998)]{kuk98} Kukula, M. J., Dunlop,
1444: J. S., Hughes, D. H., \& Rawlings, S. 1998, MNRAS, 297, 366
1445: \bibitem[Malkan(1984)]{mal84}Malkan, M. A. 1984, ApJ, 287, 555
1446: \bibitem[Martel et al.(1999)]{mar99} Martel, A. R., et al. 1999, ApJS, 122, 81
1447: \bibitem[Martini(2004)]{mar04} Martini, P. 2004,
1448: Coevolution of Black Holes and Galaxies, from the Carnegie Observatories
1449: Centennial Symposia.
1450: Published by Cambridge University Press, as part of the
1451: Carnegie Observatories Astrophysics Series. Editor Ho, L. C. 2004, 169.
1452: \bibitem[Martini et al.(2007)]{mar07} Martini, P., Mulchaey, J. S.,
1453: Kelson, D. D. 2007, ApJ, 664, 761
1454: \bibitem[McLure et al.(1999)]{mcl99} McLure, R. J., Kukula, M. J., Dunlop, J. S., Baum, S. A.,
1455: O'Dea, C. P., \& Hughes, D. H. 1999, MNRAS, 308, 377
1456: \bibitem[McLure et al.(2000)]{mcl00} McLure, R. J., Dunlop, J. S., \&
1457: Kukula, M. J. 2000, MNRAS, 318. 693
1458: \bibitem[McLure \& Dunlop(2001)]{mcl01} McLure, R. J., \& Dunlop, J. S.
1459: 2001, MNRAS, 327, 199
1460: \bibitem[Nolan et al.(2001)]{nol01} Nolan, L. A., Dunlop, J. S., Kukula, M. J.,
1461: Hughes, D. H., Boroson, T. \& Jimenez, R. 2001, MNRAS, 323, 308
1462: \bibitem[Norman \& Scoville(1988)]{nor88} Norman, C. A., \& Scoville, N. Z. 1988,
1463: ApJ, 332, 124
1464: \bibitem[Pavlovsky et al.(2005)]{pav05}
1465: Pavlovsky, C., et al. 2005,
1466: "ACS Data Handbook", Version 4.0, (Baltimore: STScI)
1467: \bibitem[Peng et al.(2002)]{pen02}
1468: Peng, C. Y., Ho, L. C., Impey, C. D., \& Rix, H.-W. 2002, AJ, 124, 266
1469: \bibitem[Pogge \& Martini(2002)]{pog02}
1470: Pogge, R. W., \& Martini, P. 2002, ApJ, 569, 624
1471: \bibitem[Prieur(1988)]{pri88} Prieur J.-L., 1988, ApJ, 326, 596
1472: \bibitem[Rothberg \& Joseph(2004)]{rot04} Rothberg, B. \& Joseph, R.D.\ 2004, ApJ, 128, 2098
1473: \bibitem[Salpeter(1964)]{sal64} Salpeter, E. E. 1964, ApJ, 140, 796
1474: \bibitem[S{\'a}nchez et al.(2004)]{san04} S{\'a}nchez, S. F., et al. 2004,
1475: ApJ, 614, 586
1476: \bibitem[Sandage \& Bedke(1994)]{san94} Sandage, A., \& Bedke, J. 1994, Carnegie Institution of Washington with The Flintridge Foundation, Washington DC
1477: \bibitem[Sanders et al.(1988)]{san88} Sanders, D. B., Soifer, B. T., Elias, J. H., Madore, B. F., Matthews, K., Neugebauer, G., \& Scoville, N. Z. 1988, ApJ, 325, 74
1478: \bibitem[Sanders \& Mirabel(1996)]{san96} Sanders, D. B., \& Mirabel, I. F. 1996,
1479: ARA\&A, 34, 749
1480: \bibitem[Scheuer(1995)]{sch95} Scheuer, P. A. G. 1995, MNRAS, 277, 331
1481: \bibitem[Schweitzer et al.(2006)]{sch06} Schweitzer, M., et al. 2006,
1482: ApJ, 649, 79
1483: \bibitem[Schweizer(1980)]{sch80} Schweizer, F., 1980, ApJ, 237, 303
1484: \bibitem[Schweizer \& Seitzer(1988)]{sch88} Schweizer, F., \& Seitzer, P. 1988, ApJ, 328, 88
1485: \bibitem[Schweizer \& Seitzer(1992)]{sch92} Schweizer, F., \& Seitzer, P. 1992, AJ, 104, 1039
1486: \bibitem[Scoville et al.(2003)]{sco03} Scoville, N. Z., Frayer, D. T., Schinnerer, E., \&
1487: Christopher, M. 2003, ApJ, 585, L105
1488: \bibitem[S{\'e}rsi{\'c}(1968)]{ser68} S{\'e}rsi{\'c}, J. L. 1968, Atlas de Galaxias Australes
1489: (C{\'o}rdoba: Obs. Astron., Univ. Nac. C{\'o}rdoba)
1490: \bibitem[Shi et al.(2007)]{shi07} Shi., Y., et al. 2007, ApJ, accepted
1491: (arXiv:0707.2806v1)
1492: \bibitem[Smith et al.(1986)]{smi86} Smith, E. P., Heckman, T. M., Bothun, G. D., Romanishin, W.
1493: \& Balick, B. 1986, ApJ, 306, 64
1494: \bibitem[Springel et al.(2005a)]{spr05a} Springel, V., Di
1495: Matteo, T., Hernquist, L. 2005a, ApJ, 620, L79
1496: \bibitem[Springel et al.(2005b)]{spr05b} Springel, V., Di Matteo, T.,
1497: Hernquist, L. 2005b, MNRAS, 361, 776
1498: \bibitem[Stockton(1978)]{sto78} Stockton, A. 1978, ApJ, 223, 747
1499: \bibitem[Stockton(1982)]{sto82} Stockton, A. 1982, ApJ, 257, 33
1500: \bibitem[Stockton \& \- Farnham(1991)]{sto91} Stockton, A., \& Farnham, T. 1991, ApJ, 371, 525
1501: \bibitem[Strateva et al.(2003)]{str03} Strateva, I. V., Brandt, W. N., Eracleous, M., Schneider, D. P., \&
1502: Chartas, G. 2003, ApJ, 651, 749
1503: \bibitem[Tadhunter \- et al.(2005)]{tad05} Tadhunter, C., Robinson, T. G., Gonz{\'a}lez Delgado, R. M.,
1504: Wills, K., \& Morganti, R. 2005, MNRAS, 356, 480
1505: \bibitem[Taylor et al.(1996)]{tay96} Taylor, G. L., Dunlop, J. S., Hughes, D. H.,
1506: \& Robson, E. I. 1996, MNRAS, 283, 930
1507: \bibitem[Toomre \& Toomre(1972)]{too72} Toomre, A., \& Toomre, J. 1972, ApJ, 178, 623
1508: \bibitem[Urrutia et al.(2007)]{urr07} Urrutia, T., Lacy M., \& Becker, R. H. 2007, ApJ, in press (arXiv:0709.2805)
1509: \bibitem[Wake et al.(2004)]{wak04}
1510: Wake, D. A., et al. 2004, ApJ, 610, L85
1511: \bibitem[Wall \& Peacock(1985)]{wal85} Wall, J. V., \& Peacock, J. A. 1985, MNRAS, 216, 173
1512: \bibitem[Wilkinson et al.(2000)]{wil00} Wilkinson, A., Prieur, J.-L., Lemoine, R., Carter, D., Malin, D., Sparks, W. B. 2000,\mnras, 319, 977
1513: \bibitem[Willis \& Strom(1978)]{wil78} Willis, A. G., \& Strom, R. G. 1978, A\&A, 62, 375
1514: \bibitem[Wold et al.(2001)]{wol01} Wold, M., Lacy, M., Lilje, P. B., Serjeant, S. 2001, MNRAS, 323, 231
1515: \bibitem[Wright et al.(1998)]{wri98} Wright, S. C., McHardy, I. M., \& Abraham, R. G. 1998, MNRAS, 295, 799
1516: \bibitem[Wu et al.(2002)]{wu02} Wu, X.-B., Liu, F. K., Zhang, T. Z. 2002, A\&A, 389, 742
1517: \bibitem[Wu \& Liu(2004)]{wu04} Wu, X.-B., \& Liu, F. K. 2004, ApJ, 614, 91
1518: \bibitem[Younger et al.(2007)]{you07} Younger, J. D., Cox, T. J.,
1519: Seth, A. C., \& Hernquist, L. 2007, ApJ, accepted (arXiv:0707.4481v1)
1520: \bibitem[Yu \& Tremaine(2002)]{yu02} Yu, Q., \& Tremaine, S. 2002, MNRAS, 335, 965
1521: \end{thebibliography}
1522:
1523: \end{document}
1524: