Interferometric imaging of the high-redshift radio galaxy, 4C60.07: An SMA, Spitzer and VLA study reveals a binary AGN/starburst
R.J. Ivison, G.E. Morrison, A.D. Biggs, Ian Smail, S.P. Willner, M.A. Gurwell, T.R. Greve, J.A. Stevens, M.L.N. Ashby
aa r X i v : . [ a s t r o - ph ] A ug Mon. Not. R. Astron. Soc. , 000–000 (0000) Printed 28 October 2018 (MN L A TEX style file v2.2)
Interferometric imaging of the high-redshift radio galaxy, 4C 60.07:An SMA,
Spitzer and VLA study reveals a binary AGN/starburst
R. J. Ivison, , G. E. Morrison, , A. D. Biggs, Ian Smail, S. P. Willner, M. A. Gurwell, T. R. Greve, J. A. Stevens and M. L. N. Ashby UK Astronomy Technology Centre, Royal Observatory, Blackford Hill, Edinburgh EH9 3HJ Scottish Universities Physics Alliance, Institute for Astronomy, University of Edinburgh, Royal Observatory, Blackford Hill, Edinburgh EH9 3HJ Institute for Astronomy, University of Hawaii, Honolulu, HI 96822, USA Canada-France-Hawaii Telescope, Kamuela, HI 96743, USA Institute for Computational Cosmology, University of Durham, South Road, Durham DH1 3LE Harvard-Smithsonian Center for Astrophysics, 60 Garden Street, Cambridge, MA 02138, USA Max-Planck Institute for Astronomy, Heidelberg, Germany Centre for Astrophysics Research, University of Hertfordshire, College Lane, Hatfield AL10 9AB
Accepted 2008 August 7. Received 2008 August 5; in original form 2008 June 19
ABSTRACT
High-resolution submillimetre (submm) imaging of the high-redshift radio galaxy (HzRG),4C 60.07, at z = 3 . , has revealed two dusty components of roughly equal integrated flux. Spitzer imaging shows that one of these components (‘B’) is coincident with an extremelyred active galactic nucleus (AGN), offset by ∼ ∼
30 kpc) from the HzRG core. Theother submm component (‘A’) – resolved by our synthesised beam and devoid of emissionat 3.6–8.0- µ m – lies between ‘B’ and the HzRG core. Since the radio galaxy was discoveredvia its extremely young, steep-spectrum radio lobes and the creation of these lobes was likelytriggered by the interaction, we argue that we are witnessing an early-stage merger, prior toits eventual equilibrium state. The interaction is between the host galaxy of an actively-fueledblack hole (BH), and a gas-rich starburst/AGN (‘B’) marked by the compact submm compo-nent and coincident with broad CO(4–3) emission. The second submm component (‘A’) is aplume of cold, dusty gas, associated with a narrow ( ∼
150 km s − ) CO feature, and may repre-sent a short-lived tidal structure. It has been claimed that HzRGs and submm-selected galaxies(SMGs) differ only in the activity of their AGNs, but such complex submm morphologies areseen only rarely amongst SMGs, which are usually older, more relaxed systems. Our studyhas important implications: where a galaxy’s gas reservoir is not aligned with its central BH,CO may be an unreliable probe of dynamical mass, affecting work on the co-assembly of BHsand host spheroids. Our data support the picture wherein close binary AGN are induced bymergers. They also raise the possibility that some supposedly jet-induced starbursts may haveformed co-evally (yet independently of) the radio jets, both triggered by the same interaction.Finally, we note that the HzRG host would have gone unnoticed without its jets and its com-panion, so there may be many other unseen BHs at high redshift, lost in the sea of ∼ × similarly bright IRAC sources – sufficiently massive to drive a > -W radio source, yetpractically invisible unless actively fueled. Key words: galaxies: starburst – galaxies: formation – techniques: interferometric – instru-mentation: interferometers – submillimetre
Radio galaxies are believed to host actively-fueled, spinningBHs which power their immense radio luminosities and fash-ion their characteristic double lobes (Fanaroff & Riley 1974; Rees1978; Blandford & Payne 1982; McCarthy 1993). High-redshiftexamples of the phenomenon (HzRGs) have often been iden- tified as ultra-steep-spectrum (USS) emitters in radio surveys (e.g. Bornancini et al. 2007) and are thus selected during theirextreme youth (
10 Myr – Blundell & Rawlings 1999). Nowa- USS sources, with α < − where S ν ∝ ν α , were often found to lackidentifications on photographic plates (e.g. Tielens, Miley & Willis 1979).c (cid:13) Ivison et al. days, a wealth of unequivocal evidence also links HzRGs withvery massive galaxies in the early Universe. Near- and mid-infrared (-IR) observations show that HzRGs are associatedwith the most massive stellar populations at any given redshift(Best, Longair & R¨ottgering 1998; Seymour et al. 2007). Directevidence of vast reservoirs of atomic and molecular hydrogenhas also been established, via observations of strong H I absorp-tion against luminous, morphologically complex Ly α halos – of-ten extending 100–200 kpc from the central radio galaxy – andvia detections of CO and dust (Hippelein & Meisenheimer 1993;Papadopoulos et al. 2000; Archibald et al. 2001; Reuland et al.2003, 2004; De Breuck et al. 2003, 2005; Klamer et al. 2005;Villar-Mart´ın et al. 2006).If HzRGs pinpoint the most massive galaxies out to the highestredshifts then we expect these systems to signpost the most mas-sive peaks in the primordial Universe and to evolve into today’sbrightest galaxies – cD-type ellipticals . In accord with these ex-pectations, the environments of HzRGs are over-dense in a richvariety of galaxy types, including SMGs. The first example, stillthe most striking, is the field of 4C 41.17 at z = 3 . (Ivison et al.2000; Greve et al. 2007), and Stevens et al. (2003) presented low-resolution submm imaging of a further six HzRGs. Some of thesedisplayed evidence of extended ( ∼ ∼ z ∼ − SMG population as probed by high-resolutionCO and submm–radio continuum imaging ( < ∼ < ∼ > ∼ L ⋆ galaxies forming from SMGs (Smail et al. 2004).Probing the size and morphology of the rest-frame far-IRemission in HzRGs is thus a key piece in the puzzle of galaxyformation. With the ∼ FWHM ) spatial resolution usu-ally available, we have been unable to distinguish between entirelydifferent modes for the formation of the most massive galaxies.One could imagine HzRG host galaxies assembling via a mas-sive, widespread burst with a low overall star-formation density,or within multiple gas-rich sub-components destined to merge. Ev-idence from mm interferometric imaging has been enlightening,though the need for imaging on a range of spatial scales is oftenevident: De Breuck et al. (2005) found evidence for two gas com-ponents in the 4C 41.17 system whilst Papadopoulos et al. (2000)argued for a galaxy-wide starburst in another HzRG, 4C 60.07, withgas being consumed on a scale of ∼
30 kpc, and for a gas reservoirmore commensurate with local ULIRGs in another, 8C 1909+722.4C 60.07 is a powerful double-lobed FR II radio galaxy – oneof the brightest of the HzRGs observed by Archibald et al. (2001)and Reuland et al. (2004) in the submm waveband. The radio spec-trum of 4C 60.07 is unremarkable at low frequencies, with a spec-tral index, α = − . between 38 and 178 MHz (where S ν ∝ ν α ),but its spectrum steepens at higher frequencies, reaching α = − . between 4.9 and 15 GHz, hence its selection as an USS radio emit-ter and its eventual identification with an R Vega ∼ . galaxyat z = 3 . ± . (Chambers et al. 1996; R¨ottgering et al.1997; Reuland et al. 2007). Defined as being centrally located in clusters and very much larger andbrighter than other galaxies in the cluster, the c being analogous to the no-tation for supergiant stars in stellar spectroscopy (Matthews et al. 1964). In this paper we present submm interferometry of 4C 60.07,utilising the Submillimetre Array (SMA – Ho et al. 2004) at awavelength of 890 µ m with a spatial resolution of ∼ z ∼ , sensi-tive to dust on the scales predicted by existing single-dish imaging.The primary goal was to determine the size and nature of the emis-sion. Is it diffuse, or is it due to multiple discrete, compact objectswhose total flux density can account for the high S µ m seen bySCUBA?In the next section we describe an extensive set of observationsand the associated analysis. Presentation of the reduced images fol-lows in §
3, with our interpretation of those data and discussion oftheir implications in § § Data were obtained on Mauna Kea, Hawaii, during 2007 Marchand October using the SMA’s “compact” configuration (baselinesas long as 113 m during March and 70 m during October – seeFig. 1) and the 345-GHz receivers, with 2 GHz of bandwidthin each of the upper and lower sidebands, as outlined in Ta-ble 1. The pointing centre was R.A. 05:12:54.80, Dec. +60:30:51.7J2000 and observations of 4C 60.07 were interspersed with ∼ ◦ from the target. The October data also included obser-vations of 8C 0716+714 (1 Jy, 17 ◦ ) – to calibrate those data takenwhen the target was setting – and NVSS J044923+633208 (0.3 Jy,4 ◦ ) to check the quality of the phase referencing. The position ofNVSS J044923+633208 was found to be accurate to < ∼ MIR , a set of
IDL -basedroutines specifically developed for data from the SMA. The mainsteps consisted of flagging noisy or corrupted data, converting thedata to units of flux density using antenna system temperatures, de-termining the bandpass (with observations of 3C 454.3 or 3C 279)and, finally, calibrating the antenna complex gains (phase and am-plitude) and interpolating the calibrator solutions onto the target.The absolute flux scale is estimated to be accurate to 10 per cent.The data were then read into the Astronomical Image ProcessingSoftware (
AIPS ) package where they were imaged.The duration of our October track was 7.35 hr, with around1 mm of precipitable water vapour (pwv). The resulting noise levelwas 1.4 mJy beam − , with a [2.4 × FWHM synthesisedbeam for naturally weighted uv data. The SMA Beam Calcula-tor and Sensitivity Estimator had predicted a synthesised beamprofile of [2.2 × FWHM and an r.m.s. noise level of1.41 mJy beam − , for a 7.5-hr track with a 20 ◦ minimum elevationand 1 mm of pwv, and so appears to function adequately.We experimented with several permutations of data. All com-binations resulted in images with the same basic characteris-tics. Although combining all the data yielded the lowest noise, The Submillimeter Array is a joint project between the Smithsonian As-trophysical Observatory and the Academia Sinica Institute of Astronomyand Astrophysics and is funded by the Smithsonian Institution and theAcademia Sinica. Additional data were taken in poor weather conditions during 2007February 20 – see Table 1 – but were not used here. http://sma1.sma.hawaii.edu/beamcalc.html c (cid:13) , 000–000 nterferometric imaging of a forming, massive elliptical Table 1.
Observations of 4C 60.07 with the SMA.Date Track length Conditions Comment(2007) /min τ February 20 216 . ± . UnsuitableMarch 02 257 . ± . AdequateMarch 03 268 . ± . MarginalOctober 23 441 . ± . Excellent − , we opted to use only those data taken in gen-uine ‘submm weather’ ( τ < . ), on October 23 andMarch 02, since experience suggests only these produce adequaterepresentations of the submm sky. The resulting noise level was1.38 mJy beam − , before CLEAN ing, with a synthesised beam mea-suring [2.25 × FWHM with the major axis at positionangle (PA), 78 ◦ . Employing 100 interations of the CLEAN algo-rithm, with a gain of 0.1, in a region of radius 45 arcsec centred onthe radio galaxy (nine times the area contained within the SMA’sprimary beam), resulted in a 20 per cent reduction in the measurednoise. The source structure and flux density were unaffected.
Spitzer
4C 60.07 was observed with
Spitzer (Werner et al. 2004) in all thebroadband filters available in the Multiband Imaging Photometerfor Spitzer (MIPS – Rieke et al. 2004) and Infrared Array Camera(IRAC – Fazio et al. 2004) bands, 3.6–8.0 and 24–160 µ m, respec-tively, in 2004 October, i.e. during the cold portion of the mission(programme identification: 3329 – Seymour et al. 2007).The IRAC data consist of a short sequence of four dithered 30-s frames centred on 4C 60.07 in each of the four bands. We elimi-nated scattered light, column pulldown, residual images and multi-plexer bleed from the individual exposures by hand, then mosaicedthese cosmetically superior frames using IRAC proc (Schuster et al.2006) with 0.86-arcsec pixels (half the area of the native IRAC pix-els). The Infrared Spectrograph (IRS – Houck et al. 2004) also tooktwo 61-s exposures at 16 µ m in the 4C 60.07 field, using ‘peak-upimaging’ mode, prior to it being offered to the community. We usethe image reduced by Seymour et al. (2007) here.The reduction of the 24- µ m data was based on version S16.1.0of the MIPS pipeline. We used an object-masked median stack ofall the MIPS exposures of 4C 60.07 to compensate for structurearising from residual images. The remaining structure in the back-grounds was eliminated using IMSURFIT in IRAF and the resultingmosaics were created with 1.2-arcsec pixels.The IRAC, IRS and MIPS data are aligned to less than0.5 arcsec, conservatively.For IRAC and MIPS, photometric measurements were madeusing 4.9- and 7.0-arcsec-diameter circular apertures, respectively,correcting to total flux via the standard aperture correction . Theflux did not vary significantly for apertures ranging from 6 to This work is based in part on observations made with the
Spitzer SpaceTelescope , which is operated by the Jet Propulsion Laboratory, CaliforniaInstitute of Technology under a contract with NASA. Support for this workwas provided by NASA through an award issued by JPL/Caltech. IRAC Data Handbook; http://ssc.spitzer.caltech.edu/irac/dh/
ARC SE C ARC SEC30 20 10 0 -10 -20 -303020100-10-20-30
Figure 1.
Top:
Coverage in the uv plane of SMA data acquired on 2007March 02 and October 23. Bottom:
Greyscale representation of the dirtybeam, with contours at − , − , 5, 10, 15, 20, 40, 60 and 80 per cent ofthe peak. µ mcentroid. An I -band image was obtained during 2000 November 01 usingthe 4.2-m William Herschel Telescope’s (WHT ) Mosaic Cam-era, which comprises two EEV 4096 × σ limiting magnitude in a 4.9-arcsec-diameter aperture is I Vega = 24 . .We also utilise the K ′ imaging of van Breugel et al. (1998),a 1-hr exposure taken using NIRC (Matthews & Soifer 1994) onthe 10-m Keck I telescope during 0.65-arcsec seeing, with a 3- σ limiting magnitude in a 4.9-arcsec-diameter aperture of K ′ Vega =22 . . Based on observations made with the WHT operated on the island of LaPalma by the Isaac Newton Group in the Spanish Observatorio del Roquede los Muchachos of the Instituto de Astrofisica de Canariasc (cid:13) , 000–000
Ivison et al.
Pseudo-continuum data were obtained at 1.4 GHz during 2001 Jan-uary using the NRAO’s Very Large Array (VLA) in its A con-figuration (programmes AD432 and AI84). The raw data wereflagged, calibrated and imaged following the technique describedby Biggs & Ivison (2006) and the resulting images have a synthe-sised beam size of [1.4 × ◦ .We obtained still higher-resolution continuum images, with100-MHz bandwidths centred at 4.7 and 8.2 GHz (programmeAC374) and synthesised beam sizes of ∼ ∼ FWHM from the National Radio Astronomy Observatory’s (NRAO’s) Data Archive System .We also reduced the 24-GHz continuum data taken byGreve et al. (2004) to explore CO J = 1 − emission from4C 60.07 for the velocity range of the broad CO(4–3) componentfound by Papadopoulos et al. (2000): +455 ± km s − . Thesewere obtained between 2001 October and 2003 March and includefive, two and two tracks of data in the D, C and DnC configu-rations, respectively (programmes AI88, AI93 and AI104). Com-bining these data with a natural weighting scheme ( ROBUST = 5)yielded a [2.6 × FWHM ), at PA 102 ◦ ,and σ = 11 . µ Jy beam − after 100 iterations of the CLEAN algo-rithm in
AIPS
IMAGR with a gain of 0.1 – more than twice asdeep as the maps presented by Greve et al. (2004) who used onlythe highest resolution data.We used several methods to quantify the CO J = 1 − emis-sion. First, we aped the approach of Greve et al. (2004): we im-aged the two 50-MHz-wide intermediate frequencies (IFs), repre-senting ‘continuum only’ (IF1, σ = 15 . µ Jy beam − ) and ‘con-tinuum plus line’ (IF2, σ = 15 . µ Jy beam − ), then subtractedIF1 CLEAN components from the uv database for IF2, after check-ing the position and integrated flux density of the brightest lobe inIF1 and IF2. The dirty image made using the continuum-subtracteddata has σ = 16 . µ Jy beam − . For our second method, we sub-tracted a dirty image of IF1 from one of IF2, yielding a ∼ √ × noisier map than the first method ( σ = 23 . µ Jy beam − ). Forour third method, we repeated the second method for each of thenine observational epochs, tapering the data to produce a ∼ FWHM ) beam. The noise-weighted combination of these nine im-ages yielded σ = 30 . µ Jy beam − . The three methods gave a con-sistent picture in which the peak flux density due to broad-line CO J = 1 − emission centred on any of the prominent IR or ra-dio components can be no greater than 0.05 mJy beam − , and nogreater than 0.12 mJy in total (both 3 σ ). Observations using the Submm Common-User Bolometer Array(SCUBA – Holland et al. 1999) on the 15-m James Clerk MaxwellTelescope (JCMT) at a wavelength of 850 µ m were described byStevens et al. (2003). We have returned to those data, and the 450- µ m data obtained simultaneously, reducing them afresh with thesoftware described by Ivison (2006), though without the shift-and-add technique since 4C 60.07 was not detected at high signal-to-noise (S/N) in individual scans. The resulting images, smoothedwith 7-arcsec Gaussians to yield point spread functions with FWHM The National Radio Astronomy Observatory is operated by AssociatedUniversities Inc., under a cooperative agreement with the National ScienceFoundation. https://archive.nrao.edu Table 2.
Submm and radio positions in the 4C 60.07 system.Characteristic Value CommentR.A. (J2000) 05:12:55.13 ± ± a S µ m . ± . mJyTotal a S µ m . ± . mJyR.A. (J2000) 05:12:54.71 ± ± a S µ m . ± . mJyTotal a S µ m . ± . mJyR.A. (J2000) 05:12:55.147 ± ± ± ± a Measured using a twin-Gaussian fit to the dirty image. of ∼
10 and ∼
15 arcsec at 450 and 850 µ m, are shown in Fig. 2.No positional corrections have been applied. Another faint SMG lies to the north, with its reference beams visible, well outside theprimary beam of the SMA’s 6-m antennas at 345 GHz (30 arcsec, FWHM ). As stated earlier, 4C 60.07 is extremely bright at submm wave-lengths, with S µ m = 23 . ± . mJy, as measured via fully-sampled jiggle maps with SCUBA (Fig. 2). Using SCUBA’s pho-tometry mode, which was insensitive to emission larger thanthe telescope beam, the measured flux density was . ± . mJy beam − – the brightest of 47 radio galaxies observed byArchibald et al. (2001) and Reuland et al. (2004).Our new, high-resolution submm interferometry, shownin Fig. 3, reveals unambiguously that the emission first de-tected using SCUBA is spread over a region larger than the < ∼
500 pc ( < ∼ z = 3 . ) usually encountered in lo-cal ULIRGs (Downes & Solomon 1998), or even the < ∼ < ∼ S µ m = 8 . and . mJy). Hereafter, ‘A’, ‘B’ and the radio core will be the principalcomponents discussed.Component ‘A’ is marginally resolved by our ∼ JMFIT in AIPS of [2.2 × ◦ ; its peak S µ m is lower than that of ‘B’,but its integrated S µ m is marginally higher. The total S µ m of Labelled ‘1’ in Stevens et al. (2003), which contains spurious coordi-nates. The object should be referred to as SMM J051254.2+603119. It hasno counterpart at 24 µ m and may lie at a similarly high redshift to 4C 60.07.c (cid:13) , 000–000 nterferometric imaging of a forming, massive elliptical ARC SE C ARC SEC40 30 20 10 0 -10 -20 -30 -40403020100-10-20-30-40
Contours: SCUBA 850umGreyscale: SCUBA 450um
Figure 2.
Greyscale representation of 4C 60.07 observed using SCUBA( § µ m, superimposed with contours of the 850- µ m im-age (shown in its original form by Stevens et al. 2003) plotted at − , − , − , , , , , × σ , where σ is the noise level. The referencebeams can be seen, due to the east-west chopping and nodding of the sec-ondary mirror on the JCMT. The positions of the components detected byour SMA imaging, discussed later, are shown as crosses. these two submm components, . ± . mJy, the equivalent of . ± . mJy at 850 µ m (applying a +13-per-cent correction for a55- K greybody with emissivity index, +1.5). This is consistent withthe . ± . mJy measured by SCUBA at 850 µ m at the ∼ σ level.The submm morphology appears consistent with that seen atlower S/N using the Plateau de Bure Interferometer (PdBI) of theInstitut de Radio Astronomie Millim´etrique (IRAM) at 240 GHz(Fig. 4 – Papadopoulos et al. 2000) and confirms the extended na-ture of the dust emission. The measured size, even taking both com-ponents together, is significantly smaller than the [11 × Papadopoulos et al. (2000) found the CO J = 4 − line emis-sion from 4C 60.07 to be bright and broad, even by the standardsnow known for SMGs (Neri et al. 2003; Greve et al. 2005). Theydescribe two distinct CO(4–3) components in their relatively low-spatial-resolution PdBI data cube (Fig. 4). Together, these cover > ∼ − .One CO(4–3) component is spans only ∼
150 km s − FWHM .Its position is consistent with that of the putative AGN core – takento be the flat-spectrum radio source between the steeper-spectrumhot spots seen most clearly at 1.4 GHz – and it is ∼
220 km s − blueward of the systemic (He II ) velocity of the AGN’s host galaxy. ARC SE C ARC SEC10 5 0 -5 -101050-5-10
Contours: SMA 890umGreyscale: IRAC 3.6+4.5um A B
Figure 3.
Greyscale representation of the
Spitzer averaged 3.6- and 4.5- µ mimages of 4C 60.07. Superimposed on the IRAC data is the high-resolutionSMA 890- µ m image of 4C 60.07, with contours plotted at − . , 2.5, 3.5,4.5, 5.5 and 6.5 × − , the r.m.s. noise level. The size andshape of the SMA’s synthesised beam is shown in the bottom left corner ofthis plot. The FWHM of the SMA’s primary beam extends slightly beyondthe extent of the image, to ∼
30 arcsec
FWHM . The position of the radiocore – as seen at 4.7 and 8.2 GHz – is shown by a square; those of the steep-spectrum hot spots are labelled with diamonds; the submm components, ‘A’and ‘B’, are labelled. The IR, radio and submm images are aligned to betterthan 0.5 arcsec, conservatively.
Greve et al. (2004) detected gas glowing in CO(1–0), coincidentin velocity with the narrow CO(4–3) emission, and consistent spa-tially. On the basis of dynamical mass limits, Papadopoulos et al.argued that this narrow component has yet to form the bulk of itseventual stellar mass.The second CO(4–3) component lies ∼
700 km s − redwardand ∼ > ∼
550 km s − FWHM , with emission likely extending to lower fre-quencies, beyond the observed band (Papadopoulos et al. 2000).Greve et al. (2004) report CO(1–0) in the velocity range of thisbroad CO(4–3) component, with a peak flux density of . ± . mJy, but offset spatially by ∼ − (Fig. 5). The total CO(1–0)flux, integrated over the velocity range to which these data are sen-sitive ( +455 ± km s − ), is σ < . Jy km s − . Adopting X CO = 0 . ( K km s − pc − ) − M ⊙ (Downes & Solomon 1998),this translates into M H < . × M ⊙ , less than half ofthe mass determined on the basis of the CO J = 4 − emis-sion by Papadopoulos et al. who assumed a 4–3/1–0 line ratio of r = 0 . , and less than the mass determined for the narrowCO(1–0) component seen blueward of our passband by Greve et al.(2004). This suggests variations in the bulk gas properties acrossthe 4C 60.07 system. c (cid:13)000
550 km s − FWHM , with emission likely extending to lower fre-quencies, beyond the observed band (Papadopoulos et al. 2000).Greve et al. (2004) report CO(1–0) in the velocity range of thisbroad CO(4–3) component, with a peak flux density of . ± . mJy, but offset spatially by ∼ − (Fig. 5). The total CO(1–0)flux, integrated over the velocity range to which these data are sen-sitive ( +455 ± km s − ), is σ < . Jy km s − . Adopting X CO = 0 . ( K km s − pc − ) − M ⊙ (Downes & Solomon 1998),this translates into M H < . × M ⊙ , less than half ofthe mass determined on the basis of the CO J = 4 − emis-sion by Papadopoulos et al. who assumed a 4–3/1–0 line ratio of r = 0 . , and less than the mass determined for the narrowCO(1–0) component seen blueward of our passband by Greve et al.(2004). This suggests variations in the bulk gas properties acrossthe 4C 60.07 system. c (cid:13)000 , 000–000 Ivison et al.
ARC SE C ARC SEC10 5 0 -5 -101050-5-10
Contours: IRAM 1250umGreyscale: IRAM CO(4−3)
Figure 4.
Greyscale representation of the uniform-weighted, velocity-integrated CO(4–3) emission from 4C 60.07, as detected using IRAM PdBIwith a [8.9 × − . , 2.5, 3.5, 4.5, 5.5 and 6.5 × − , ther.m.s. noise level, with the shape and size of the somewhat elongated 240-GHz synthesised beam shown in the bottom left corner. The image is cen-tred on the position of the radio core – seen at 4.7 and 8.2 GHz – which islabelled with a square; those of the steep-spectrum hot spots are labelledwith diamonds; submm components, ‘A’ and ‘B’ (see Fig. 3), are markedwith crosses. We first concern ourselves with the new data from the SMA. Themost unexpected aspect of the SMA,
Spitzer and VLA imaging(Fig. 3) is the offset between the primary sites of submm emis-sion and that of the radio core’s synchrotron . The submm com-ponents lie 10–30 kpc (on the plane of the sky) from the radio core– a larger distance than can be understood via the positional un-certainties ( ≪ § .Fig. 6 shows how the morphology of the 4C 60.07 systemchanges as we move from rest-frame 0.2 µ m (UV), via rest-frame0.5 µ m ( ∼ V ), 0.8 µ m ( ∼ R ), 0.9 µ m ( ∼ I ), 1.2 µ m ( ∼ J ), 1.7 µ m The positional offset is significant at the 4- σ level. The uncertainty inthe submm position is dominated by the centroiding (see the appendix ofIvison et al. 2007) since systematic uncertainties are limited by the coin-cidence of the southern submm component with IR emission and by thepositional tests described in § Of the six submm-bright HzRGs with deep
Spitzer imaging(Archibald et al. 2001; Reuland et al. 2004; Seymour et al. 2007), all dis-play an accurate alignment of the radio core and the 3.6- µ m emission (to atolerance of ∼ µ m(Fig. 6). ARC SE C ARC SEC10 5 0 -5 -101050-5-10
Contours: VLA CO J=1−0Greyscale: VLA 24GHz
Figure 5.
Greyscale representation of the 24-GHz continuum emission from4C 60.07 superimposed with contours of the CO(1–0) emission in the ve-locity range of the broad CO(4–3) component detected with IRAM PdBI byPapadopoulos et al. ( +455 ± km s − ) as imaged using the VLA witha [2.6 × § − , 2 and 3 × µ Jy beam − , the r.m.s. noiselevel. The emission coincident with component ‘B’ has an overall signifi-cance, using a Gaussian fit with size fixed to that of the synthesised beam,of < ∼ σ . The image is centred on the position of the radio core, which islabelled with a square; those of the steep-spectrum hot spots are labelledwith diamonds; submm components, ‘A’ and ‘B’ (see Fig. 3), are markedwith crosses. ( ∼ H ) and 3.3 µ m ( ∼ L ) to rest-frame 5.0 µ m ( ∼ M ), i.e. λ obs = 0.9–24 µ m. The rest-frame UV emission is centred on the radiocore. In rest-frame V , a red object (‘K’) becomes visible, 2.2 arcsecNNE of the core; the core emission covers ∼ R and I ( λ obs = 3.6–4.5 µ m) and the emissioncoincident with submm component ‘B’ strengthens. By the time wereach the rest-frame J and H filters at λ obs = 5.8–8.0 µ m, ‘B’ hasbecome the dominant component, with only very weak emissionassociated with the radio core and ‘K’. Component ‘A’ appears tobe devoid of significant mid-IR emission, with less than 4, 6, 26 and39 µ Jy (3 σ ) at λ obs = 3.6, 4.5, 5.8 and 8.0 µ m. At λ obs = 24 µ mwe see emission along the orientation of the submm components,‘A’ and ‘B’, with little room for a significant contribution from theradio core (Fig. 6). (The MIPS 24- µ m filter transmits between 4.3and 5.4 µ m – rest frame, to half power – and so misses the strongestPAH features, e.g. Desai et al. 2007). We have extracted I and 3.6–8.0- µ m flux densities for the twocomponents, the radio core and ‘B’, that dominate the rest-frameoptical–IR emission from the 4C 60.07 system, and adopted othermeasurements from Chambers et al. (1996) and van Breugel et al.(1998). The different noise levels at 3.6–8.0 µ m in Fig. 6 make itdifficult to judge the spectral energy distribution (SED) shapes, so c (cid:13) , 000–000 nterferometric imaging of a forming, massive elliptical A B
WHT I [0.5][1.2] [1.7]IRAC 8.0IRAC 5.8 [5.0]MIPS 24IRAC 4.5[0.9][0.8]IRAC 3.6Keck K IRS 16[3.3] K =0.2][ λ rest Figure 6.
Greyscale representations of WHT I ( § K ′ (van Breugel et al. 1998), and Spitzer images of 4C 60.07 from 0.9 through to 24 µ m, usinga linear stretch from 0–10 σ , where σ is the pixel noise (0–20 σ at 24 µ m, where we have re-mapped the image into 0.86-arcsec pixels to match the IRACimages, 2.9 × smaller linearly than is necessary to fully sample the 24- µ m point spread function). The WHT and Keck images have been smoothed with a0.3-arcsec Gaussian. The Keck image has been rotated slightly to match the USNO-based astrometry of the WHT I -band image. In each image, two crossesshow the positions of the SMA submm peaks and the approximate rest-frame wavelength (in µ m) is shown in parentheses. The images are centred on theposition of the radio core, labelled with a square, while those of the steep-spectrum hot spots are labelled with diamonds. Each image is [25 ×
25] arcsec.
Table 3.
Mid-IR flux densities.Isophotal wavelength Flux density ( µ Jy)( µ m) Core/‘A’ ‘B’3.535 20.6 ± ± ± ± σ < ± σ < ± ± ± we plot both of the dominant components in Fig. 7 and list the fluxdensities in Table 3. The rest-frame UV– V emission from the coreis resolved. At rest-frame R and I ( λ obs = 3 . − . µ m) we seea blend of emission from the core and ‘K’; since our photometryis appropriate for point sources, the true flux densities are probably ∼
20 per cent higher than those shown. Component ‘B’ is barely re-solved (3.1 arcsec,
FWHM ), so its photometry ought to be accurate.The 24- µ m emission is a blend, with component ‘B’ contributingless than half. The total flux density at 24 µ m is . ± . mJy. Seymour et al. (2007) report different 8- and 24- µ m flux densities for4C 60.07. Their 7-arcsec aperture, centred on the 3.6- µ m emission, missesaround half the 8- µ m flux (which comes mostly from ‘B’). At 24 µ m,Seymour et al. used an aperture 26 arcsec in diameter (not 13 arcsec asstated in the paper; N. Seymour 2008, private communication) and thusinclude flux from the source to the south-east. Their 16- µ m aperture suffers AGN SEDs rise as a power law between 1 and 10 µ m(rest frame), while starbursts have flatter SEDs at rest-frame 1–3 µ m which rise at longer wavelengths (e.g. Ward et al. 1987;Donley et al. 2007). At z = 3 . , IRAC probes rest-frame 0.8–1.7 µ m, so S . µ m /S . µ m is a powerful AGN diagnostic, par-ticularly when plotted against a probe of the > µ m slope, suchas S µ m /S µ m (Ivison et al. 2004; Lacy et al. 2004; Magdis et al.2008).If the 3.6–5.8- µ m emission associated with the radio galaxycore is associated with starlight then we might expect that emis-sion to continue through to λ obs = 7.7 µ m – near the isophotalwavelength of the IRAC 8.0- µ m filter, which corresponds to the1.6- µ m stellar bump at z ∼ . . We do not see this, but usingthe IRAC photometry to constrain rest-frame H -band luminositiesand assuming L H /M ∼ , the stellar mass is poorly constrained: < M ⊙ , cf. (2 . ± . × M ⊙ in Seymour et al. (2007),who found the H -band light to be 95-per-cent stellar. We note thata much larger L/M (and hence lower stellar mass) is plausible, viafine-tuning of age and initial mass function, and there is probablya contribution to the rest-frame near-IR light from the AGN. Ourlimit is therefore likely to be very conservative, but is consistentwith a large mass of ∼ no such problem, although it was 12 arcsec in diameter rather than 6 arcsec,as stated, so we adopt their 16- µ m flux here.c (cid:13) , 000–000 Ivison et al.
Figure 7.
SEDs of the two main components of 4C 60.07 from the opti-cal to the submm, with component ‘B’ represented by open circles andthe AGN core by filled circles. Beyond 8 µ m we show total flux densi-ties, as diamonds. The dotted line is the SED of the z = 1 . ERO,HR 10 (Elston et al. 1988; Stern et al. 2006), normalised in flux with re-spect to 4C 60.07 at rest-frame 185 µ m (only the well-sampled optical/IRSED is plotted); the dashed line is the SED of a model 10 -L ⊙ starburst(Lagache et al. 2003). λ obs ∼ . µ m. It is fit well by a S ν ∝ ν − . power law: anextremely red galaxy, akin to HR 10 (= ERO J164502+4626.4) at z ∼ . which has excellent coverage of its SED – see Fig. 7(Dey et al. 1999; Stern et al. 2006). The value of S . µ m /S . µ m seen for component ‘B’ ( . ± . , or [4.5] − [8.0] = 2.0) is un-usual: redder than any SMG in the SHADES sample (Ivison et al.2007), and ranking with the reddest of the 8- µ m-selected ob-jects explored by Magdis et al. (2008). These diagnostics are typ-ically used for objects at z ∼ , however, so we should explorewhether the colour remains extreme at z ∼ . Does it exhibit sim-ilar rest-frame optical/near-IR colours to HR 10? For z = 1 . , S . µ m /S . µ m is close to S . µ m /S . µ m at z = 3 . . HR 10has S . µ m /S . µ m ∼ . ± . , so although HR 10 is redderthan component ‘B’ at rest-frame < ∼ µ m, it is not nearly as red atrest-frame 1–10 µ m which means that component ‘B’ can be cate-gorised as an obscured AGN with little room for ambiguity. It is tempting to invoke a complex, many-component system andto speculate that this form of interaction may be related to radioloudness and the formation of the most massive galaxies, with im-plications for the form and role of feedback. However, we mustnot forget the origins of our target: 4C 60.07 was selected initiallyas a USS emitter and its radio activity is thus inescapably young,
10 Myr (Blundell & Rawlings 1999). We are therefore seeing asystem that is probably out of equilibrium.Can we marry together all the data for the 4C 60.07 system – inparticular the radio and submm interferometry, both continuum andspectral line (CO), and the mid-IR imaging? These characteristicsare presented schematically in Fig. 8.In our favoured model, 4C 60.07 comprises a far-IR-luminousstarburst triggered by a galaxy-galaxy interaction. That more than
V R
500 km/s1.2mm cont bridge?
Core
A B
Figure 8.
Schematic of 4C 60.07. The horizontal axis is a spatial vector onthe sky, starting at the radio core and running through ‘A’ and ‘B’. The ver-tical axis shows relative velocity, where available. Black lines shown the ex-tent of CO(4–3) channel maps from Papadopoulos et al. (2000) – the dashedline represents the emission that spans both components (note the velocitygap between it and the 150-km s − component). Components ‘A’ and ‘B’are marked by circles at arbitrary velocities. The upper box represents theCO(1–0) map of the narrow component (Greve et al. 2004), while the lowerbox represents the 1.2-mm continuum, its velocity chosen to match that ofthe broad component. one component is involved is inescapable, given the morphologiesevident in the submm and mid-IR wavebands. Of the myriad com-ponents in the system, we identify the radio core as one with sig-nificant mass, since it appears to comprise a relatively large BH(capable of driving the radio lobes) and a host galaxy that is visi-ble in rest-frame UV, V , R and I . Since it lacks rest-frame far-IRemission, we suspect the radio galaxy host has exhausted (or drivenaway) its supply of molecular gas. Certainly, little H is evident viaits most sensitive available tracers, dust and CO. We identify com-ponent ‘B’ as the other high-mass component, as evident from itsvery broad CO(4–3) line emission – a dusty, gas-rich galaxy, as ev-ident from the submm emission, undergoing a huge burst of starformation. Component ‘B’ is very heavily enshrouded in the dustthat betrays its presence in the submm waveband, visible elsewherein its SED only longward of rest-frame R , probably via a buriedAGN with a power-law SED ( § J = 4 − and − .Papadopoulos et al. (2000) believed this emission was associatedwith the radio galaxy host but our high-resolution SMA imag-ing, together with the high-resolution CO(1–0) observations ofGreve et al. (2004), betray an association with component ‘A’. Webelieve that ‘A’ is a bridge of cold material tidally stripped from themassive, gas-rich starburst/AGN (‘B’) towards the host of the ra-dio core – an analogue of the scenario seen in some local ULIRGs,but on a more massive scale. This is in accord with its observedcharacteristics: the submm component has no counterpart at mid-IR wavelengths, is resolved on a scale of ∼ ∼
15 kpc) andits gas – consistent spatially for CO(1–0) and CO(4–3) – has a verylow velocity dispersion ( ∼
150 km s − ). c (cid:13) , 000–000 nterferometric imaging of a forming, massive elliptical New, high-resolution submm imaging of the HzRG, 4C 60.07, at z = 3 . , reveals two clumps of emission separated by 3.3 arcsec( ∼
25 kpc). Although one clump is resolved by our ∼ < ∼
15 kpc) and there is little evi-dence for a galaxy-wide starburst on larger scales from our obser-vations.Previously thought to comprise two gas-rich starbursts, onecentred on the radio galaxy, we discover an extra level of complex-ity due to the misalignment of the submm emission and the radiocore. Given the extreme youth of HzRGs, we interpret the 4C 60.07system as an early-stage merger, with its radio jets and a nearbygas-rich starburst/AGN both induced by the interaction. We inter-pret a second clump of dust, resolved by our submm imaging andcoincident with very narrow CO emission, as a plume of cold, dustygas in a tidal stream, seen in a non-equilibrium state. It is separatedspatially from the radio galaxy, which may have exhausted its sup-ply of fuel.Recent work on the z = 4 . quasar BR 1202 − ∼
10 per cent of the total – Wang et al. 2004). In4C 60.07, the galaxy nuclei correspond to the radio core and ‘B’,and the overlap region is ‘A’. There are also similarities between4C 60.07 and HE 0450 − z > ∼ , where quasars are becoming rarer, a binary quasarwith a small separation ( < < z < , tens are known – Mortlock et al. 1999; Kochanek et al. 1999;Hennawi et al. 2006). The 4C 60.07 system contains two AGN and,although not strictly speaking a ‘binary quasar’, it does supportthe argument made by Djorgovski (1991) that the high numberof binary quasars on small angular separations – orders of mag-nitude more than that predicted by an extrapolation of the quasarcorrelation function power law – is due to the triggering of AGNduring mergers (see also Smail et al. 2003; Alexander et al. 2003;Iono et al. 2006).We would not have been aware of the BH that presumablydrives the immense radio luminosity of 4C 60.07 were it not for the interaction that triggered both the FR II radio activity and thenearby starburst. Is there an unseen population of BHs at high red-shift – invisible unless actively fueled, yet sufficiently massive todrive a > -W radio source? Aside from its rare radio proper-ties, the SED of 4C 60.07’s host galaxy is not particularly unusual,and such objects would be impossible to sift from the ∼ × similarly bright IRAC sources in the sky.Reuland et al. (2007) claim that HzRGs and SMGs differ onlyin the activity of their AGNs. Why, then, are complex submm mor-phologies the exception rather than the rule in typical SMGs? Theanswer lies in the youth of the 4C 60.07 system, as set by its selec-tion as a USS radio emitter. We see 4C 60.07 in the first throes ofa violent interaction, probably within a few Myr of its latest burstof activity commencing, whereas analysis of the rest-frame opti-cal properties of SMGs show that they are typically > × olderthan HzRGs – ∼
100 Myr (Smail et al. 2004; Swinbank et al. 2006)– and are therefore more likely to display more relaxed morpholo-gies. This provides a ready explanation for the extended natureof the submm emission seen towards the Stevens et al. sample ofHzRGs, which is at odds with the compact emission seen for mostSMGs (Younger et al. 2007, 2008a; Biggs & Ivison 2008). Sincethese HzRGs will all be youthful, we expect non-equilibrium mor-phologies to be commonplace, though not necessarily the complexdual-AGN signature seen in 4C 60.07. It is not the activity of theirAGNs that mark HzRGs as different from SMGs, it is their youth.The possibility that HzRGs differ from SMGs has many implica-tions. It would be desirable to make similar observations of a largersample of HzRGs at a resolution similar or better than that em-ployed here.Recent studies of rest-frame far-IR emission, either directly(Tacconi et al. 2006; Younger et al. 2007, 2008a) or via the far-IR/radio correlation (Chapman et al. 2004; Biggs & Ivison 2008),have highlighted the importance of high spatial resolution. Single-dish observations – with SCUBA2 (Holland et al. 2006),
Herschel (Griffin et al. 2007) or
SPICA (Swinyard & Nakagawa 2008) – willoften provide an ambiguous or misleading picture. Our work re-inforces this view: we have spent a decade assuming that the radiocore of 4C 60.07 dominates the rest-frame far-IR emission when theemission is, in fact, associated with a gas-rich companion. We mustawait sensitive interferometers such as the Atacama Large Millime-tre Array (ALMA – Wootten 2003) to probe the interactions thatdrive many aspects of galaxy evolution, and design future missions,such as the
Far-Infrared Interferometer ( FIRI – Helmich & Ivison2008), with the ability to probe a large range of spatial scales.
Facilities:
SMA,
Spitzer , VLA, JCMT, WHT.
REFERENCES
Alexander D. M. et al., 2003, AJ, 125, 383Archibald E. N., Dunlop J. S., Hughes D. H., Rawlings S., EalesS. A., Ivison R. J., 2001, MNRAS, 323, 417Best P. N., Longair M. S., R¨ottgering H. J. A., 1998, MNRAS,295, 549Biggs A. D., Ivison R. J., 2006, MNRAS, 371, 963Biggs A. D., Ivison R. J., 2008, MNRAS, 385, 893Blandford R. D., Payne D. G., 1982, MNRAS, 199, 883Blundell K. M., Rawlings S., 1999, Nat, 399, 330Bornancini C. G., De Breuck C., de Vries W., Croft S., vanBreugel W., R¨ottgering H., Minniti D., 2007, MNRAS, 378, 551Canalizo G., Stockton A., 2001, ApJ, 555, 719 c (cid:13) , 000–000 Ivison et al.
Chambers K. C., Miley G. K., van Breugel W. J. M., Huang J.-S.,1996, ApJS, 106, 215Chapman S. C., Smail I., Windhorst R., Muxlow T., Ivison R. J.,2004, ApJ, 611, 732Croft S. et al., 2006, ApJ, 647, 1040De Breuck C., Downes D., Neri R., van Breugel W., Reuland M.,Omont A., Ivison R., 2005, A&A, 430, L1De Breuck C. et al., 2003, A&A, 401, 911Desai V. et al., 2007, ApJ, 669, 810Dey A., Graham J. R., Ivison R. J., Smail I., Wright G. S., LiuM. C., 1999, ApJ, 519, 610Djorgovski S., 1991, in Astronomical Society of the Pacific Con-ference Series, Vol. 21, Crampton D., ed, The Space Distributionof Quasars, p. 349Donley J. L., Rieke G. H., P´erez-Gonz´alez P. G., Rigby J. R.,Alonso-Herrero A., 2007, ApJ, 660, 167Downes D., Solomon P. M., 1998, ApJ, 507, 615Elston R., Rieke G. H., Rieke M. J., 1988, ApJ, 331, L77Fanaroff B. L., Riley J. M., 1974, MNRAS, 167, 31PFazio G. G. et al., 2004, ApJS, 154, 10Frayer D. T., Ivison R. J., Smail I., Yun M. S., Armus L., 1999,AJ, 118, 139Greve T. R. et al., 2005, MNRAS, 359, 1165Greve T. R., Ivison R. J., Papadopoulos P. P., 2004, A&A, 419, 99Greve T. R., Stern D., Ivison R. J., De Breuck C., Kov´acs A.,Bertoldi F., 2007, MNRAS, 382, 48Griffin M. et al., 2007, Advances in Space Research, 40, 612Hatch N. A., Overzier R. A., R¨ottgering H. J. A., Kurk J. D., MileyG. K., 2008, MNRAS, 383, 931Helmich F., Ivison R. J., 2008, Experimental Astronomy, in pressHennawi J. F. et al., 2006, AJ, 131, 1Hippelein H., Meisenheimer K., 1993, Nat, 362, 224Ho P. T. P., Moran J. M., Lo K. Y., 2004, ApJ, 616, L1Holland W. et al., 2006, in Presented at the Society of Photo-Optical Instrumentation Engineers (SPIE) Conference, Vol.6275, Millimeter and Submillimeter Detectors and Instrumen-tation for Astronomy III. Edited by Zmuidzinas, Jonas; Holland,Wayne S.; Withington, Stafford; Duncan, William D.. Proceed-ings of the SPIE, Volume 6275, pp. 62751E (2006).Holland W. S. et al., 1999, MNRAS, 303, 659Houck J. R. et al., 2004, ApJS, 154, 18Hutchings J. B., Neff S. G., 1988, AJ, 96, 1575Iono D. et al., 2006, ApJ, 645, L97Ivison R. J., 2006, MNRAS, 370, 495Ivison R. J., Dunlop J. S., Smail I., Dey A., Liu M. C., GrahamJ. R., 2000, ApJ, 542, 27Ivison R. J. et al., 2007, MNRAS, 380, 199Ivison R. J. et al., 2004, ApJS, 154, 124Ivison R. J. et al., 2002, MNRAS, 337, 1Klamer I. J., Ekers R. D., Sadler E. M., Hunstead R. W., 2004,ApJ, 612, L97Klamer I. J., Ekers R. D., Sadler E. M., Weiss A., Hunstead R. W.,De Breuck C., 2005, ApJ, 621, L1Kochanek C. S., Falco E. E., Mu˜noz J. A., 1999, ApJ, 510, 590Lacy M. et al., 2004, ApJS, 154, 166Lagache G., Dole H., Puget J.-L., 2003, MNRAS, 338, 555Magdis G. E., Rigopoulou D., Huang J.-S., Fazio G. G., WillnerS. P., Ashby M. L. N., 2008, MNRAS, 386, 11Matthews K., Soifer B. T., 1994, Experimental Astronomy, 3, 77Matthews T. A., Morgan W. W., Schmidt M., 1964, ApJ, 140, 35McCarthy P. J., 1993, ARA&A, 31, 639 Mortlock D. J., Webster R. L., Francis P. J., 1999, MNRAS, 309,836Neri R. et al., 2003, ApJ, 597, L113Papadopoulos P. P., Feain I. J., Wagg J., Wilner D. J., 2008, ArXive-prints, 801Papadopoulos P. P., R¨ottgering H. J. A., van der Werf P. P., Guil-loteau S., Omont A., van Breugel W. J. M., Tilanus R. P. J., 2000,ApJ, 528, 626Rees M. J., 1978, Nat, 275, 516Reuland M., R¨ottgering H., van Breugel W., De Breuck C., 2004,MNRAS, 353, 377Reuland M. et al., 2007, AJ, 133, 2607Reuland M. et al., 2003, ApJ, 592, 755Rieke G. H. et al., 2004, ApJS, 154, 25R¨ottgering H. J. A., van Ojik R., Miley G. K., Chambers K. C.,van Breugel W. J. M., de Koff S., 1997, A&A, 326, 505Sanders D. B., Soifer B. T., Elias J. H., Neugebauer G., MatthewsK., 1988, ApJ, 328, L35Schuster M. T., Marengo M., Patten B. M., 2006, in Presented atthe Society of Photo-Optical Instrumentation Engineers (SPIE)Conference, Vol. 6270, Observatory Operations: Strategies, Pro-cesses, and Systems. Edited by Silva, David R.; Doxsey, RodgerE.. Proceedings of the SPIE, Volume 6270, pp. 627020 (2006).Seymour N. et al., 2007, ApJS, 171, 353Smail I., Chapman S. C., Blain A. W., Ivison R. J., 2004, ApJ,616, 71Smail I., Scharf C. A., Ivison R. J., Stevens J. A., Bower R. G.,Dunlop J. S., 2003, ApJ, 599, 86Stern D., Chary R.-R., Eisenhardt P. R. M., Moustakas L. A.,2006, AJ, 132, 1405Stevens J. A. et al., 2003, Nat, 425, 264Swinbank A. M., Chapman S. C., Smail I., Lindner C., Borys C.,Blain A. W., Ivison R. J., Lewis G. F., 2006, MNRAS, 371, 465Swinyard B. M., Nakagawa T., 2008, Experimental Astronomy, inpressTacconi L. J. et al., 2008, ApJ, 680, 246Tacconi L. J. et al., 2006, ApJ, 640, 228Tielens A. G. G. M., Miley G. K., Willis A. G., 1979, A&AS, 35,153Tulloch S., 2000, The Newsletter of the Isaac Newton Group ofTelescopes (ING Newsl.), issue no. 2, p. 26-27, 2, 26van Breugel W. J. M., Stanford S. A., Spinrad H., Stern D., Gra-ham J. R., 1998, ApJ, 502, 614V´azquez G. A., Leitherer C., 2005, ApJ, 621, 695Villar-Mart´ın M. et al., 2006, MNRAS, 366, L1Walter F. et al., 2003, Nat, 424, 406Walter F., Carilli C., Bertoldi F., Menten K., Cox P., Lo K. Y., FanX., Strauss M. A., 2004, ApJ, 615, L17Wang Z. et al., 2004, ApJS, 154, 193Ward M., Elvis M., Fabbiano G., Carleton N. P., Willner S. P.,Lawrence A., 1987, ApJ, 315, 74Werner M. W. et al., 2004, ApJS, 154, 1Wootten A., 2003, in Presented at the Society of Photo-OpticalInstrumentation Engineers (SPIE) Conference, Vol. 4837, Os-chmann J. M., Stepp L. M., eds, Large Ground-based Telescopes.Edited by Oschmann, Jacobus M.; Stepp, Larry M. Proceedingsof the SPIE, Volume 4837, pp. 110-118 (2003)., p. 110Younger J. D. et al., 2008a, MNRAS, 387, 707Younger J. D. et al., 2007, ApJ, 671, 1531Younger J. D. et al., 2008b, ArXiv e-prints, 807 c (cid:13) , 000–000 nterferometric imaging of a forming, massive elliptical This paper has been typeset from a TEX/ L A TEX file prepared by theauthor. c (cid:13)000