The Origin of the Intrinsic Scatter in the Relation Between Black Hole Mass and Bulge Lumin

IN

TheAstrophysicalJournal.

APreprinttypesetusingLTEXstyleemulateapjv.26/01/00

THEORIGINOFTHEINTRINSICSCATTERINTHERELATIONBETWEENBLACKHOLEMASSAND

BULGELUMINOSITYFORNEARBYACTIVEGALAXIES1

MINJINKIM2,3,LUISC.HO2,CHIENY.PENG4,AARONJ.BARTH5,MYUNGSHINIM3,PAULMARTINI6,

7

ANDCHARLESH.NELSON

ToAppearinTheAstrophysicalJournal.

arXiv:0807.1337v1 [astro-ph] 8 Jul 2008ABSTRACT

Weinvestigatetheoriginoftheintrinsicscatterinthecorrelationbetweenblackholemass(MBH)andbulgeluminosity(Lbul)inasampleof45massive,local(z<∼0.35)type1activegalacticnuclei(AGNs).WederiveMBHfrompublishedopticalspectraassumingasphericalbroad-lineregion,andLbulfromdetailedtwo-dimensionaldecompositionofarchivalopticalHubbleSpaceTelescopeimages.AGNsfollowtheMBH−Lbulrelationofinactivegalaxies,butthezeropointisshiftedbyanaverageof∆logMBH≈−0.3dex.Weshowthatthemagnitudeofthezeropointoffset,whichisresponsiblefortheintrinsicscatterintheMBH−Lbulrelation,iscorrelatedwithseveralAGNandhostgalaxyproperties,allofwhichareultimatelyrelatedto,ordirectlyimpact,theBHmassaccretionrate.Atagivenbulgeluminosity,sourceswithhigherEddingtonratioshavelowerMBH.ThezeropointoffsetcanbeexplainedbyachangeinthenormalizationofthevirialproductusedtoestimateMBH,inconjunctionwithmodestBHgrowth(∼10%–40%)duringtheAGNphase.GalaxymergersandtidalinteractionsappeartoplayanimportantroleinregulatingAGNfuelinginlow-redshiftAGNs.

Subjectheadings:galaxies:active—galaxies:bulges—galaxies:fundamentalparameters—galaxies:

photometry—quasars:general

1.INTRODUCTION

Early-typegalaxiescommonlycontainacentralblackhole(BH)whosemassstronglycorrelateswiththebulgeluminos-ity(Kormendy&Richstone1995;Magorrianetal.1998)andstellarvelocitydispersion(Gebhardtetal.2000;Ferrarese&Merritt2000).Lower-massBHsfoundinlate-typespiralsandspheroidalgalaxiesfollowasimilarMBH−σ⋆relation(Barthetal.2005;Greene&Ho2006a)butapparentlyadifferentMBH−Lbulrelation(Greeneetal.2008).TheBH-hostscalingrelationssuggestthatBHsplayanimportantroleingalaxyfor-mationandevolution(e.g.,Granatoetal.2004;DiMatteoetal.2005;Robertsonetal.2006).Understandingthemecha-nismbywhichBHscoevolvewiththeirhostsimpactscurrentmodelsofcosmologicalstructureformation(e.g.,Boweretal.2006;Crotonetal.2006).

Akey,unansweredquestionishowtheBH-hostscalingre-lationsoriginated.ThisissuecanbeaddressedbyextendingtheBH-hostscalingrelationstoactivegalaxies—wherein,theBH,byselection,iscurrentlystillgrowing—andbytrackingthescalingrelationstohigherredshifttoseewhenandpossi-blyhowtheywereestablished.BHmassesintype1(broad-line,unobscured)activegalacticnuclei(AGNs)nowcanberoutinelyestimatedtoreasonableaccuracy(∼0.3−0.5dex),fromthe“virialmethod”usingsingle-epochultravioletorop-ticalspectra(e.g.,Kaspietal.2000;McLure&Dunlop2001;Vestergaard2002;Greene&Ho2005).Morechallengingtoobtainarereliablemeasurementsoftheunderlyinghostgalaxy,particularlyofthebulgecomponent,whichismaximallyaf-1Based

fectedbythebrightAGNcore(e.g.,McLureetal.1999;Floydetal.2004;Nelsonetal.2004;Greene&Ho2006b;Kimetal.2007),althoughsubstantialprogresshasbeenmade.

RecentstudiespresenttantalizingevidencethattheBH-hostscalingrelationsforactivegalaxiesevolvewithredshift,evenbyz≈0.4(Wooetal.2006;Treuetal.2007)andasfarbackasz≈4(Pengetal.2006a,2006b;Shieldsetal.2006;Ho2007).Comparedtolocal,inactivesystems,thesenseoftheevolutionisthatforagivenhostgalaxymassorgravitationalpotential,higherredshiftAGNshavealargerBHmassthansimilarsystemsatlowredshift.Takenatfacevalue,thissug-geststhatthegrowthoftheBHprecedes,oratleastoutpaces,thegrowthofthegalaxyathigherredshifts.Ontheotherhand,local(z≈0)AGNsseemtobehavequitedifferently.McLure&Dunlop(2002)studiedasampleof72nearbyAGNsandfindthattheyroughlyfollowthesameMBH−Lbulrelationdefinedbyinactivegalaxies,albeitwithasomewhatgreaterscatter.IntheMBH−σ⋆relationofAGNs,highlyaccretingAGNsseemtohaveadifferentnormalizationinthesensethattheytendtohavealowerMBHforagivenσ⋆(Greene&Ho2006b;Shenetal.2008).AsimilartrendisseenbyHoetal.(2008),us-ingHIlinewidthstoconstrainthegravitationalpotentialoftheunderlyinghostgalaxy.

Asaconcretesteptowardestablishingarobustz=0baselineforcomparisonwithhigh-zstudies,thisseriesofpapers(Kimetal.2007,2008;M.Kimetal.,inpreparation)focusesonquantifyingthelocalMBH−Lbul

onobservationsmadewiththeNASA/ESAHubbleSpaceTelescope,obtainedfromtheDataArchiveattheSpaceTelescopeScienceInstitute,whichis

operatedbytheAssociationofUniversitiesforResearchinAstronomy(AURA),Inc.,underNASAcontractNAS5-26555.TheseobservationsareassociatedwithprogramAR-10969andGO-9763.

2TheObservatoriesoftheCarnegieInstitutionofWashington,813SantaBarbaraStreet,Pasadena,CA91101;mjkim@ociw.edu,lho@ociw.edu.3DepartmentofPhysicsandAstronomy,FrontierPhysicsResearchDivision(FPRD),SeoulNationalUniversity,Seoul,Korea;mim@astro.snu.ac.kr.4NRCHerzbergInstituteofAstrophysics,5071WestSaanichRoad,Victoria,BritishColumbia,CanadaV9E2E7;cyp@nrc-cnrc.gc.ca.

5DepartmentofPhysicsandAstronomy,UniversityofCaliforniaatIrvine,4129FrederickReinesHall,Irvine,CA92697-4575;barth@uci.edu.

6CenterforCosmologyandAstroParticlePhysics,TheOhioStateUniversity,191WestWoodruffAvenue,OH43210;martini@astronomy.ohio-state.edu7PhysicsandAstronomyDepartment,DrakeUniversity,2507UniversityAvenue,DesMoines,IA50311;charles.nelson@drake.edu

1

2KIMetal.

1086N-1.5-2.0-2.57.58.08.5log (MBH / M )Ο9.09.5420

0.0

0.1

z

FIG.2.—Distributionofredshiftsforoursample.Theopenhistogramshows

thetotalsample;thehatchedhistogramshowstheobjectswithlowEddingtonratio(Lbol/LEdd≤0.1);thedashedhistogramshowstheradio-loudobjects.

0.0-0.5log (Lbol / LEdd)-1.00.20.3

FIG.1.—DistributionofMBHvs.Eddingtonratio.Radio-loudandradio-quietobjectsaredenotedbyfilledandopensymbols,respectively.

relationforactivegalaxies.Amongnearbyinactivegalaxies,BHmasscorrelatesonlymarginallylesstightlywithbulgelu-minosityormassthanwithbulgestellarvelocitydispersion(Marconi&Hunt2003;Häring&Rix2004;Novaketal.2006).Moreover,theMBH−Lbulrelationforactivegalax-iesshowsnolargesystematicdifferencesfromthatofinac-tivegalaxies(McLure&Dunlop2002).ThissuggeststhattheMBH−LbulrelationcanbeusedasausefulsubstitutefortheMBH−σ⋆relation,beinganespeciallyeffectiveobservationaltooltotrackthecosmologicalevolutionoftheBH-galaxycon-nection(Pengetal.2006a,2006b;Treuetal.2007).Whereasstellarvelocitydispersionsaredifficult,ifnotimpossible,tomeasurefordistantquasars,forexample,photometricmea-surementsofquasarhostscontinuetobefeasibleevenouttohighredshifts,eitherthroughdirectimaging(e.g.,Kukulaetal.2001;Ridgwayetal.2001)orthroughstronglensingmagni-fication(Pengetal.2006b).Inacompanionpaper,Kimetal.(2008)demonstratethatthebulgeluminosityoftype1AGNscanbemeasuredtoareasonableaccuracy(∼0.5mag)inHub-bleSpaceTelescope(HST)images,evenintheregimewhentheactivenucleusfaroutshinesthegalaxy.

InsteadofcharacterizingthefullMBH−LbulrelationforAGNs,thispaperrestrictsitselftoonlyoneimportantaspect:theoriginoftheintrinsicscatter.BychoosingasampleforwhichwecanestimatereliableBHmassesandforwhichwecanderiverobustmeasurementsofbulgeluminosityfromHSTimages,ourobjectiveistoquantifythetrueintrinsicscatteroftherelationandtocharacterizepossiblevariationsofthescat-terwithphysicalpropertiesoftheAGN,hostgalaxy,orenvi-ronment.ByelucidatingthephysicaldriversthatinfluencethescatteroftheMBH−Lbulrelation,wehopetogaininsightsonhowtheBH-hostgalaxyrelationswereestablished.

Thispaperisstructuredasfollows.Wedescribethesampleselectionin§2.Wepresenttheimage-fittingprocedureformea-suringbulgeluminositiesandourimagedecompositionresultsin§3.WeinvestigatetheMBH−Lbulrelationfortype1AGNsin§4.Finally,Section5discussestheoriginoftheintrinsic

8Our

scatterintheMBH−Lbulrelation,endingwithasummaryin§6.Throughoutweadoptthefollowingcosmologicalparame-ters:H0=100h=71kms−1Mpc−1,Ωm=0.27,andΩΛ=0.75(Spergeletal.2003).

2.SAMPLESELECTION

OurinitialselectionbeginswithallAGNsknowntopos-sessbroademissionlines(type1objects),andhavereasonablydeep,andnon-highlysaturatedopticalimagesintheHSTpub-licarchive.SinceweareinterestedinestablishingthelocalMBH−Lbulrelation,weonlyconsidersourceswithz<∼0.35.Next,wecarefullysearchtheliteratureforpublishedmeasure-mentsofbroademission-linewidths(eitherHαorHβ),whichareneededforcalculatingMBH.Althoughthisstepisneces-sarilysomewhatsubjective,wetrytobeconsistentinselect-ingonlyobjectsthathavelinewidthswithpublishederrorbars<∼10%.TheavailabilityofspectrophotometricmeasurementsisnotessentialforusbecauseourBHmassesultimatelymakeuseofthenuclearluminositiesfromourphotometricdecompo-sitionofthenucleus(§3.3).Theabovescreeningprocessyieldsapproximately200objects.

SincetheprincipalaimofthisworkistostudytheintrinsicscatteroftheMBH−Lbulrelation,itisimperativethatwechooseobjectsforwhichwecanobtainthemostreliableestimatesofthetwoprimaryquantitiesofinterest,MBHandLbul.Guidedbythisoverridinggoal,wepurposelyrestrictoursampletotheupperendoftheMBHdistribution.Allelsebeingequal,thisse-lectioncriterionbiasesthesampletowardmoreluminous,moremassive,earlier-typehostsforwhichwecanderivemorere-liablebulgeparametersbecausethestructuraldecompositionwillbelesscomplicatedthaninlater-typesystems.Anaddedbenefitofthismassselectionisthatoursamplewillconsistofcloseanalogsofhigher-redshiftquasars.Forconcreteness,wechoosesourceswithMBH>107.8M⊙.Thislimit8isadmittedlysomewhat

toMBH

BHmassesassumeasmallergeometricalfactorthanthatusedinOnkenetal.(2004),byafactorof1.8.Thus,MBH=107.8M⊙onourscaleisequivalent=108M⊙onthescaleofOnkenetal.Wenotethatthatourmainconclusionsdonotrelyonthegeometricalfactor.Thisissueisdiscussedin§5.2.

BlackHoleMassvs.BulgeLuminosityRelation34KIMetal.BlackHoleMassvs.BulgeLuminosityRelation56KIMetal.BlackHoleMassvs.BulgeLuminosityRelation7

8KIMetal.

BlackHoleMassvs.BulgeLuminosityRelation9

10KIMetal.

0.50.4noitcar0.3F evitale0.2R0.10.0

-1.0-0.50.00.5

Mhost (out) - Mhost (in) [mag]

FIG.3.—Distributionofthedifferencesinhostgalaxymagnitudesfromthe2-Dimaging-fittingsimulationsofKimetal.(2008).ArtificialimagesofAGNhostgalaxieswitharangeofinputparametersweregenerated,andGAL-FITwasusedtorecovertheinputparameters.TheopenhistogramsdenotetheerrorsforidealizedconditionsinwhichthefittingwasdonewiththesamePSFasusedformakingtheinputimages.ThehatchedhistogramsgivetheerrorsforthemorerealisticsituationthataccountsforPSFmismatch(seeKimetal.2008fordetails).PSFmismatchcausessignificantsystematicerrors.

arbitrary,butityieldsasizablesampleof45objects.Ourob-jects(Table1)span∼2ordersofmagnitudeinaccretionrate(Eddingtonratio)overarelativelynarrowrange(∼1dex)inMBH(Fig.1)andredshift(Fig.2).Fromradiodataassembledfromtheliterature,33%(15/45)ofthesampleisradio-loud,de-finedbyR≥10,whereR≡fν(6cm)/f19).

ν(4400Å)(Kellermannetal.3.ANALYSIS

3.1.StructuralDecomposition

Ourcompanionpaper(Kimetal.2008)discussesindetailourtechniquefordecomposingtheHSTimagesofAGNhostgalaxies.Weperformedextensivesimulationstoquantifytheperformanceofthetwo-dimensional(2-D)image-fittingcodeGALFIT(Pengetal.2002)underconditionstypicallyencoun-teredinAGNhostgalaxyimagescontainedintheHSTarchive,similartothoseanalyzedinthisstudy.Wepaidparticularatten-tiontoquantifyingsystematicuncertaintiesinestimatingthephotometricparametersofthebulgecomponentanddevisedstrategiesformitigatingthem.

Intheregimewherethebright,unresolvedactivenucleusdominatesoverthelightofthehostgalaxy,ourabilitytoex-tractthebulgeluminositydependssensitivelyonthepropertiesofthepoint-spreadfunction(PSF).PSFmismatchsystemati-callybiasesthederivedbulgeluminositiestohighvalues,byasmuchas0.5–1mag(Fig.3).PSFmismatchoccursasaresultofvariationsintime,locationonthedetector,and,toalesserex-tent,differencesinthespectralenergydistributionbetweenthePSFstarandthesciencetarget.Thedominanteffect,however,comesfromthefactthatHSTPSFsareundersampled.BecausethePSFsarenotNyquist-sampled,itisimpossibletoshiftbyasub-pixelunittoperfectlyalignthePSFstarwiththeAGNcore.Kimetal.(2008)showthatthisfundamentalproblem

outweighsmostotherconcerns,includingthechoiceofactu-allyobserved(stellar)orsynthetic(TinyTim;Krist1995)PSFs.Theydemonstratethattheundersamplingproblemcanbesig-nificantlyalleviatedbybroadeningboththescienceimageandthePSFimagetocriticalsampling[fullwidthathalfmaximum(FWHM)∼2pixels].Thisisthestrategyweadopthere.Al-thoughthechoiceofrealversussyntheticPSFsissecondary,weuseobservedPSFstarswheneverpossible.Whenthesearenotavailable,weuseTinyTimPSFs.

Asinourcompanionpaper,weuseanupdatedversionofGALFIT(C.Y.Pengetal.,inpreparation)9.Thecodesimul-taneouslyfitsmultiplecomponentstomodelthehostgalaxy,withthefreedomtouseFouriermodestoaccommodatecom-plex,nonaxisymmetricfeaturessuchastidaldistortionsorevenspiralarms.Theseimprovementsallowustoobtainamoreac-curatedecompositionofthestructuralcomponentsofthehost,animportantconsiderationforouraimofderivingrobustbulgeluminosities.

WemodeltheactivenucleuswithasyntheticPSFandthehostgalaxywithellipsesrepresentedbyaSérsic(1968)func-tion:

I(r)=Ieexp󰀁−bn

󰀅

r

BlackHoleMassvs.BulgeLuminosityRelation11

FIG.4.—Examplesofnon-interactingandinteractingobjects,basedonthestrengthofthefirstFouriermode(a1)measuredwithGALFIT.Fromlefttoright,thesixobjectsarearrangedintheorderofincreasinga1.Thefirstthree,witha1<∼0.1,∼0.1,showessentiallynosignsofperturbation.Thelastthree,witha1>areincreasinglydisturbedmorphologically.Ineachcolumn,weshow,fromtoptobottom,theoriginalimage,model,andresidual.Theunitsoftheimagesareinarcseconds.

axialratiothanthebulge.Thediskcomponentaddsfouraddi-tionalfreeparameters,namelyposition,Re,µe,axisratio,andpositionangle.Figure24givesanexampleofabulge+diskfitforMS1059.0+7302.

Thechoicebetweenasingle-componentandadouble-componentfitforthehostgalaxyisnotalwaysclear.Addingextrafreeparametersobviouslyyieldsabetterfit.Thequestioniswhethertheextracomponentisclearlyrequiredandphysi-callymeaningful.Forexample,whenPSFmismatchispartic-ularlysevere,theextracomponentmightsimplybeattemptingtoaccountforthelargeresidualsfromthepoorPSFmodel.Insuchsituationsthe“extra”componenttendstohaveunusualpropertiessuchassuspiciouslytinyReorextremevaluesofn.Inpractice,complicatedfitsoftenrequiresomedegreeofjudg-mentcall,butwehavetriedtoerronthesideofcautionandgenerallyinvokeextracomponentsonlywhentheyareabso-lutelyneeded.Inmostcases,weadmitanadditionalcompo-nentonlyifitisclearlyvisibleintheoriginalimageorintheresidualimage.

Adiskcomponentmaybepresentbutundetectableinshal-lowimages(e.g.,Bennertetal.2008).Itisthusveryusefultoplaceanupperlimitonthediskcomponentevenifnodiskisre-quiredbythebest-fitmodeloftheHSTimage.Forobjectswithnodirectlydetecteddisks,wederiveupperlimitsforthediskcomponentbyplacingartificial(face-on)diskscoveringawiderangeofluminosityonthescienceimage,assumingthatthehostgalaxyfollowstherelationbetweenbulge-to-totallightra-tio(B/T)andtheratioofdisksizetobulgeeffectiveradiusde-rivedfromnearbyearly-typegalaxies(deJongetal.2004).Wethenfitthesimulatedimageswithtwo-component(bulge+disk)models.Thediskluminosityatwhichtheprogramfailstore-covertheinputvaluegivesanestimateoftheupperlimitforthediskcomponent.Wedidnotattempttoderivediskupperlim-itsforsystemsthatareexceptionallycomplicated(e.g.,highlydistorted,closecompanions,etc.).

TheupdatedversionofGALFITalsohastheabilitytomodelspiralarmsinthedisk.Thespiralstructureiscreatedbyahy-perbolictangentrotationfunctionwiththefollowingparame-ters:barlength,outerspiralradius,rotationrate,skyinclina-tion,andpositionangle.Thedetailsofthespiralstructuresarecreatedbyhigh-orderFouriermodes.Figure41showsthefitforPG1411+442,whosediskcomponentshowstwoprominentspiralarms.

Auniqueaspectofouranalysisisthatweattempttoquantita-tivelyestimatethedegreetowhichthehostgalaxyexhibitsnon-axisymmetricdistortions.Morphologicaldisturbancesmaybesignaturesofrecentmergersortidalinteractions,whichmighttriggerorenhanceAGNfueling.Whileavarietyoftechniqueshavebeendevisedtocharacterizemorphologicalasymmetryininactivegalaxies(e.g.,Conseliceetal.2000;Lotzetal.2004),theycannotbereadilyextendedtogalaxiescontainingbrightAGNsbecausethecentralpointsourcecandominatetheasym-metrysignal.ThelatestversionofGALFITimplementsasym-metryparametersasanintegralpartoftheimage-fittingpro-cess.Thisisaccomplishedbyintroducinghigher-orderFouriermodestochangetheshapeofthehostgalaxyfromaxisym-metricellipsesintomorecomplicatedshapes.AllthewhilethelightprofileofthehostgalaxymodelwouldstilldeclineasaSérsicprofileineverydirectionfromthepeak.Inthisscheme,thestrengthofanexternalperturbationonthehost

12KIMetal.

galaxywouldsensitivelyregisterashigh-amplitudeFouriermodes,withphaseanglesthatreflectthedirectionoftheper-turbation.

Iftheresidualimageshowssignificantnonaxisymmetry,weadoptedaFouriercomponenttofitit.TheFouriermodehasthefollowingform:

󰀁Nr(x,y)=r0(x,y)1+󰀂

amcos(m(θ+πm))󰀃.(3)

m=1

Inthisexpression,θ=arctan[(y−ycentroidoftheellipse,qisthec)/(x−xaxisratio,c)q],where(xrc,yc)

isthe0(x,y)isthegeneralizedellipse,athephaseangleformodemistheamplitudeformodem,andπm.TheFouriermodeisalwayscou-mispledwithageneralsingle(e.g.,Sérsic)componentandshowshowmuchacomponentisperturbedfromtheperfectellip-soid.Thus,theFouriermodeallowsustoquantifythede-greeofasymmetry.Inprinciple,wecanuseaninfinitenum-berofFouriermodes,butinpracticewefindthatfourmodes(m=1,3,4,5)areenoughtofittheasymmetricalstructuresen-counteredinoursample.Figure4illustratesaseriesofobjectswithincreasingstrengthsinanoobvioussignsofmorphological1.Sourceswithaperturbation,1<∼0.1showlit-tletowhereasthosewitha1>∼0.1appearincreasinglydisturbed.

DespitethesignificantnewfeaturesoftheupdatedversionofGALFIT,wenotethatthederivedvaluesofmanystandardparameters(e.g.,sizeandluminosity)arenotsubstantiallydif-ferentbetweentheoriginalandnewversionsofthecode.Insituationswheretherearedifferences,thenewfeaturesofthecodeallowforbetterconvergence,especiallywhenitcomestomulti-componentdecompositions.TheupdatedversionofGALFIThasbeentestedextensivelybyus,butforthesakeofbrevitywedeferafulldiscussionofthetechnicaldetailstoanupcomingpaper(C.Y.Pengetal.,inpreparation).

Table2summarizestheresultsofthestructuraldecomposi-tion.Foreachobject,welistthebest-fitnuclearandphotomet-ricparametersforthebulge,diskortidalfeature,andfortheoverallhostgalaxy;theparametera1isalsotabulated.

3.2.BulgeLuminosities

Theerrorbarsonthebulgemeasurementsareinfluencedbyanumberofsystematicuncertainties.UsingthesimulationsinKimetal.(2008)asaguide,thefinalerrorbudgetonthebulgeluminositywasestimatedasfollows.ForsourceswithL2,σ≈±0.3mag,whereasσ≈±0.4magifLbul/L≥0.nucbul/Lnuc<0.2.Ontopofthesevalues,additionaluncertaintiesareintroducedifbulge-to-diskdecompositionisrequired(∼0.1mag),ifsatu-rationoccurs(∼0.2mag),oriftheimagecontainssubstantialinnerfinestructure(∼0.3mag).Wealsodeterminetheerrorofthehostluminosity(mhost)fromthesimulationsatagivenLhost/Lnuc.AccordingtoKimetal.,theerroronthenucleusmagnitudeis∼±0.1mag.

BecauseofthecomplexityoftheGALFITdecomposition,itisworthwhiletocross-checkour2-Dparametricfitswithanon-parametricestimateofthehostluminosity(see,e.g.,Greeneetal.2008).Toperformthistest,weremovethenucleusfromeachsourcesimplybysubtractingashifted,scaledPSFmodelfromthepeakoftheAGNcore.Aftermaskingoutobviouscompanionsandforegroundstars,wesumuptheremainingfluxtoestimatethetotalhostgalaxymagnitude.Asaseparatetest,wecomputethehostgalaxyfluxaftersubtractingthePSFcomponentderivedfromthebest-fittingGALFITmodelforthe

entireimage.ThesetwotestsgivetherangeofvaluestabulatedasmaperinTable2.Comparisonoftheseestimateswiththehostmagnitudesobtainedfromtheparametricfits(mgoodagreementforthemajorityofthehost)showsrea-sonablysources.ThefewcasesinwhichmaperissubstantiallybrighterthanmfromPSFmismatchandcontam-hostcanbeattributedtolargeresidualsinationfromneighboringsources.

AsignificantnumberofthesourcesinoursampleoverlapwiththosestudiedbyDunlopetal.(2003),affordinganinde-pendent,externalcheckofouranalysis.Dunlopetal.alsoper-formed2-Ddecompositionoftheirsample,buttheyfittedthehostgalaxieswithonlyasinglecomponent,modeledaseitheraclassicaldeVaucouleurs(n=4)bulgeoranexponential(n=1)disk.Afteraccountingfordifferencesintheadoptedcosmo-logicalparameters,wefind,notsurprisingly,thatforbulge-dominatedsourcesourbulgemagnitudesgenerallyagreewell(towithin0.1mag)withthosegivenbyDunlopetal.Theex-ceptionsareobjectswithlarge,nearbyneighborsandsourceswithmultiplecomponents.Whereasweperformasimultane-ous2-Dfitofallnearbysourcesthatcouldpotentiallyaffectourtargetofinterest,Dunlopetal.simplymaskedthemout.Thiscouldleadtosystematicerrorsinthederivedpropertiesoftheprimaryhost.

AparticularlystrikingexampleisPG1012+008,whichisanobviouslyinteractingsystemconsistingofthreegalaxies.Si-multaneouslyaccountingforthesubcomponents,includinganoff-centereddisk,ourbest-fitmodelyieldsabulgewithm17.1(F675W)andaneffectiveradiusofR′′

bul=

etal.obtainme=2.96,or9.3kpc.Bycontrast,Dunlop5′′bul=16.4magandRe=.75,whichcorrespondsto18kpcusingourassumeddistanceof6Mpc.Forsourcesthatclearlycontainbothabulgeandadisk(e.g.,PKS2349−01),ourtwo-componentfitsyieldmorerobustbulgeluminosities.Lastly,forcompleteness,wenotethatourfinalnuclearmagnitudes(Table3),convertedtotheRband,aresystematicallybrighterby0.4magcomparedtothosegiveninDunlopetal.Thisdifferencecanbetracedtothediffer-entassumptionsusedforcalculatingthek-correction.Dunlopetal.assumedthattheAGNspectrumcanberepresentedbyasinglepowerlaw(fν∝ν−2),whereasweusetheempiricalquasarcompositespectrumofVandenBerketal.(2001).

Fiveofourobjects(MS0754.6+3928,MS1059.0+7302,MS1545.3+0305,PG1416−129,andPG1426+015)overlapwiththesamplestudiedbySchadeetal.(2000),whoalsoperformed2-Dfitstoderivephotometricparametersforthehostgalaxies.ThetwostudiesshowsignificantdifferenceinthesensethatSchadeetal.tendtounderestimatethenuclearmagnitudesonaverageby0.2magandtooverestimatethebulgemagnitudesby∼0.7mag.Someobjectsshowparticularlystrikingdisagree-ment.Inouranalysis,thehostgalaxyofMS1059.0+7302iswelldescribedbyabulge+diskmodel.Ourbestfityieldsthe

mbul=17.02magandR=2.76fore=0′′

.38forthebulgeandmthedisk.Bycontrast,Schadedisk=15.

magandRe′′

etal.find

mbul=15.53magandR=1.28fore=3′′

.47forthebulgeandmthedisk.Weattributethediscrepancydisk=16.68

magandRbetweenoure′′

resultsandthoseofSchadeetal.toadifferenceinmethodology.Schadeetal.fittedtheHSTimagessimultane-ouslywithground-basedimages.Althoughtheground-basedimagesaredeeper,theyhaveamuchbroaderandlessstablePSFthantheHSTimages;itisdifficulttoknowhowthisef-fectimpactsthefittingresults.Otherdifferencesstemfromthemodeladoptedinthefit.Inourwork,PG1426+015isbestfit

BlackHoleMassvs.BulgeLuminosityRelation13

9) O• M / HBM8( gol7Reverberation mapSingle-epoch-20-21-22-23-24-25MR,bul (mag)

FIG.5.—CorrelationbetweenBHmassandabsoluteR-bandmagnitudeforthebulge.TheBHmassesarederivedfromreverberationmapping(redcircles)orfromsingle-epochspectra(bluestars),andthebulgeluminositiesarebasedontheGALFITdecomposition.ThevaluesofMasystematicuncertaintyBHestimatedfromsingle-epochspectraareassumedtohaveof0.2dex.Thebestfitisplottedasasolidline,andtheintrinsicscatterisdenotedbythedashedlines.TheMBH−Lbulrelationforinactivegalaxiesisshownbythethickhatchedline.

withatwo-componentbulge+diskmodelwithasignificantaFouriermode,whereasSchadeetal.employedonlyasingle-1componentbulgeforthehost.Ifweadoptasingle-componentmodel,ourresultsagreewellwiththoseofSchadeetal.

ThebulgemagnitudeslistedinTable2werederivedfromim-agestakenindifferentfilters.Foroursubsequentanalysis,weneedtoconvertthemagnitudestoasinglestandardbandpassatz=0.ForeaseofcomparisonwiththeMwechooseBH−Lbulrelationofinactivegalaxies(Bettonietal.2003),theRbandasthereference.WeperformthecolorconversionoftheobservedmagnitudeinthevariousHSTfilterstotheRbandandapplyk-correctionusinggalaxytemplatespectrafromCalzettietal.(1994)andKinneyetal.(1996).Weassumethatthebulgecomponenthasthespectrumofanellipticalgalaxyandthatthediskcomponentisapproximatedbyalate-type(Sc)galaxy.FortheimagestakenintheF814Wfilter,weemploythetemplatespectrumofastarburstgalaxyforthediskcomponentbecausethetemplatespectrumofalate-typegalaxyisunavailableinthiswavelengthregime.ThefinalR-bandabsolutemagnitudesaregiveninTable3.

3.3.BlackHoleMasses

TheBHmassesfortype21AGNscanbeestimatedfromthevirialproductMBH≈fRv/G,whereRistheradiusofthebroad-lineregion(BLR),visthelinewidthoftheBLRgasrepresentedbyFWHMHβ,theFWHMofthebroadHβline,andfisafactoroforderunitythatdependsonthestructure,dynamics,andinclinationangleoftheBLR.Directmeasure-mentsofRthroughreverberationmappingareavailableonlyforasmallnumberofsources(Petersonetal.2004),butfor-tunatelythisquantitycanbeestimatedfromthecorrelationbe-tweenRandluminosity(Kaspietal.2000,2005).Thevirialproductis,however,uncertainbythenormalizingfactorf.As-sumingthattheBLRissphericalandhasanisotropicvelocity

field,f=0.75,andthelatestradius-luminosityrelationfromBentzetal.(2006)yields,.0

MBH=5.5×106ML⊙

󰀅

λ5100

103kms−1

󰀆2,

(4)

whereλL5100isthecontinuumluminosityat5100Å(seeGreene&Ho2007bfordetails).

Thecontinuumluminositycanbeestimatedeitherthroughspectrophotometryorthroughourimageanalysis.Spectropho-tometryhastheadvantagethatthespecificcontinuumfluxatthedesiredwavelengthcanbedirectlymeasuredwithoutmak-ingassumptionsaboutthespectralshape.Ontheotherhand,accurateabsolutespectrophotometryisnontrivialtoachieveandisrarelyavailableformostobjectsintheliterature.More-over,ground-basedaperturesinvariablyblendthenucleuswithatleastpartofthehost.Bycontrast,ourcarefulimagede-compositionyieldsaclean,unambiguousmeasurementofthenuclearcontinuum.Ournuclearmagnitudeshaveatypicalun-certainty(dominatedbysystematiceffectsfromPSFmismatch)of∼0.1mag.Weneedtoassumeaspectrum(wechoosethequasartemplatefromVandenBerketal.2001)inordertoes-timatethecontinuumluminosityat5100Å,buttheamountofextrapolationforourfiltersisminimal.Amoresignificantun-certaintycomesfromtemporalvariabilitybetweenourphoto-metricmeasurementsandtheliterature-basedspectralobserva-tionsusedtoobtainFWHMminosityconsideredhereusuallyHβ.Nevertheless,AGNsofthelu-varybyonly∼0.13magonlongtimescales(e.g.,Giveonetal.1999).Thewidthsofbroademissionlinesintype1AGNstypicallyhaveanuncertaintyof∼10%(e.g.,Marzianietal.2003).Takingallofthesefactorsintoconsideration,weestimatethattheyintroduceameasure-mentuncertaintyofonly∼0.2dexinMforthesingle-epochmasses,BH.Thelargestsourceofuncertaintyhowever,probablycomesfromtheintrinsicscatteroftheradius-luminosityrela-tion,whichisestimatedtobe∼0.4dex(Bentzetal.2006).AccordingtoPetersonetal.(2004),thevaluesofMfromreverberationmappingareaccurateto∼30%,BHderivedor0.1dex.Wenotethatthereisanadditionaluncertaintyonthegeometri-calfactor(f).Forinstance,Collinetal.(2006)arguedthatfmightbedependentontheaccretionrate.Wevisitthisissuein§5.2.

4.THEMBH−LbulRELATIONFORTYPE1AGNS

Figure5showstheMtype1AGNs.ObjectswithBH−MLbulrelationforoursampleofBHderivedfromreverberationmappingareencodedseparatelyfromthosebasedonsingle-epochspectra.WeassumethattheMBH−Lbulrelationfollowsalog-logrelation

log(MBH/M⊙)=α+βMR,bul.

(5)

ThethickhatchedlinerepresentstheM(2003).BH−Lbulrelationforin-activegalaxiesfromBettonietal.Convertedtoourcosmology(seePengetal.2006a),thebest-fittingrelationforinactivegalaxiesisdescribedbyα=−2.6andβ=−0.5,withascatterof0.4dex.

Weestimateαandβfortheactivesamplebyminimizingχ2,definedas

󰀂Nχ2

≡

(yi−α−βxi)2i=1

14KIMetal.

9) O• M / HBM8( gol7low Lbol/LEddhigh Lbol/LEdd-20-21-22-23-24-25MR,bul (mag)

FIG.6.—DependenceoftheMBH−LbulrelationonEddingtonratio.Bluestarsandsolidlinerepresentthecorrelationandthebestfitforhigh-Eddingtonratioobjects(LEddingtonratiobol/LobjectsEdd≥0.1).Redcirclesanddashedlinerepresentlow-(Lfortheinactivegalaxies,bol/LwhoseEddrelation<0.1).Weisdenotedfixedthebyslopethethickto−0.hatched5,asderivedline.Measurementuncertaintiesof0.1and0.2dexareadoptedforMreverberationmappingandsingle-epochspectra,respectively.

BHestimatedfromwherey=log(MBH/M⊙),x=MR,bul,andǫsurementerrorsofyandx,respectively(Tremaineyiandǫxiarethemea-etal.2002).Thismethodtreatsbothxandyasindependentvariablesandaccountsforasymmetricuncertaintiesforeach.

TheestimationoftheMoferrorsforMBH−Lbulrelationdependsonǫyi,thechoiceBH.Ifweadoptanuncertaintyof0.4dexforthemassesbasedonsingle-epochspectra,theχ2valueispracticallydominatedbythereverberation-mappedobjectsbe-causetheiruncertaintiesareafactorof4smaller,resultinginaMBH−Lbulrelationstronglybiasedtowardthereverberation-mappedsubsample.Forconcreteness,weassumethatuncer-taintiesonthesingle-epochmassesare0.2dex,whichisatyp-icalmeasurementerror(§3.3).AsFigure5shows,oursampleofAGNsclusteraroundthefiducialMgalaxieswithsignificantscatter.BHThe−Lbulrelationofin-activeformalfitfortheAGNshasaslopeofβ=−0.26±0.05,flatterthanforinactivegalaxies(β=−0.5),butbecauseofthelimiteddynamicrangeinMBH,wedonotregardtheAGNfittoberobust.Amoremean-ingfulexerciseistofixtheslopeoftherelationtothevalueforinactivegalaxiesandthenexaminetheoffsetandscatteroftheAGNsample.Fixingβto−0.5,theAGNsamplehas∆α=−0.3andanrmsscatterof0.4dex.

4.1.DependenceonEddingtonRatio

TounderstandthephysicaloriginoftheintrinsicscatterintheMBH−Lbulrelation,wedividethesampleintotwobinsinEddingtonratio,atLasLbol/LEdd=0.1.TheEddingtonluminosityisdefinedEdd=1.26×1038(MBH/M⊙)ergss−1,andthebolo-metricluminosityisestimatedassumingLbol=9λL5100(Kaspietal.2000).Figure6showsaclearoffsetbetweenthetwosub-samples.AtagivenMtobehostedbymoreBH,objectswithhighEddingtonratiostendluminousbulges,or,alternatively,atagivenbulgeluminositytheytendtohavelessmassiveBHs.Inordertoquantifytheoffsetbetweenthetwosubsamples,wefixtheslopetothatoftheMBH−Lbulrelationforinactive

9) O• M / HBM8( gol7 B/T < 0.50.5 ≤ B/T < 1 B/T = 1-20-21-22-23-24-25MR,bul (mag)

FIG.7.—SimilartoFigure6,exceptthathereweshowthedependenceongalaxymorphology:ellipticalgalaxies(B/T=1;bluestarsandsolidline),bulge-dominatedsystems(0.5≤B/T<1;greensquaresanddash-dottedline),anddisk-dominatedsystems(B/T<0.5;redcirclesanddashedline).

galaxies.AtafixedMR,bul,theoffsetinM,theoffsetinMBHis∼−0.6dex;atafixedMBHR,bulis∼1.2mag.Wenotethattheseoffsetsaremuchlargerthanthemeasurementerrors.Perform-ingaKolmogorov-SmirnovtesttoevaluatehowMtwosubsamplesisdistributed,wefindthatthenullRhypothesis,bulinthethatthetwosubsamplesaredrawnfromthesameparentpopu-lationcanberejectedwithaprobabilityof97.3%.Asdiscussedin§5.2,thesegregationbetweenthetwosubsamplesreallydoreflectintrinsicdifferencesinEddingtonratiosratherthanun-certaintiesinthedeterminationofBHmass.

4.2.DependenceonMorphologicalType

Theavailabilityofrobuststructuraldecompositiongivesusanopportunitytoexaminepossibletrendswithmorphologicaltype.Usingthemeasuredvaluesofbulge-to-totalluminosityratio(B/T;Table2)andthecorrelationbetweenmorphologicaltypeandB/Tinnormal,inactivegalaxies(Simien&deVau-couleurs1986),wedividethesampleintothreesubgroups:B/T=1(ellipticals),0.5≤B/T<1(bulge-dominated),andB/T<0.5(disk-dominated).Figure7(seealsoTable4)showsthatthezeropointoftheMBH−Lbulrelation,andpossiblyscat-ter,maydependonB/T,althoughgiventhelimitedstatisticsweregardtheevidenceastentative.Ellipticalsandearly-type,bulge-dominatedgalaxiesappearvirtuallyindistinguishable,butlater-type,disk-dominatedsystems(B/T<0.5)appeardis-tinctlyoffsettolargerMoffsetBH(by∼0.4−0.6dex)atafixedMThemagnitudeoftheismuchlargerthanpossibleRsys-,bul.tematicbiasesinbulgeluminositiesresultingfromuncertaintiesinbulge-to-diskdecomposition(∼0.2mag;seeFig.14inKimetal.2008).Asdiscussedin§5.1,inmanyinstancesourtwo-componentfitsmaynotcorrespondstrictlytoabulge+diskde-compositionbutrathertoabulge+tidalfeaturedecomposition.

BlackHoleMassvs.BulgeLuminosityRelation15

9) O• M / HBM8PG 1700+518( goMRK 1048PKS 2349-01lPG 1613+658HE 0354-55007MS 1545.3+0305PG 1012+008InteractionNo Interaction-20-21-22-23-24-25MR,bul (mag)

FIG.8.—SimilartoFigure6,exceptthathereweshowthedependenceonthedegreeofinteraction.Non-interactingormildlyinteractingobjects(a1<0.3)aredenotedbybluestarsandthesolidline,whilestronglyinter-actingsystems(a1≥0.3),withtheirnameslabeled,aredenotedbyredcirclesandthedashedline.Foursourceswithprobableminorcompanionsaremarkedasfilledbluestars.TheMBH−Lbulrelationforinactivegalaxiesisshownbythethickhatchedline.

4.3.DependenceonTidalInteraction

Wemakeuseofthequantitativemeasureofgalaxyasymme-try,a1,tostudythepossibleeffectoftidalinteraction.Givenoursmallsamplesize,wesimplydivideitintotwobinsaccord-ingtothevalueofatoprovideausefulempirical1.AsshowninFigure4,aboundarybetweenobjects1≈0.1seemsthataredisturbedmorphologically(a1≥0.1)fromthosethatarenot(a1<0.1).Withthisthresholdforaobvioussegregation1,however,thetwopop-ulationsshownointheMButwiththeboundarysetatahigherthresholdofBHa−Lbulplane.illustratesthatfouroutofthesixobjectsin1=0.3,Fig-ure8oursamplewiththeclearestsignsofmorphologicaldisturbancedoprefer-entiallyseemtolieamongthemostextremenegativeoutliersintheMBH−Lbulrelation.

4.4.DependenceonRadioProperties

ThephysicaldriversresponsibleforthegenerationofjetsandradioemissioninAGNsarestillunclear.SuggestionshaveincludedBHmass(e.g.,Laor2000),accretionrate(e.g.,Ho2002),andhostgalaxymorphology,whichmightultimatelybelinkedtoBHspin(Sikoraetal.2007).Figure9showsthataclearseparationexistsbetweenradio-loudobjectsandradio-quietobjects.Radio-loudsourcesliepreferentiallybelowtheMBH−Lbulrelationofinactivegalaxiesandradio-quietsources.

4.5.DependenceonRedshift

Theredshiftrangeofourobjectsissmall(09) O• M / HBM8( gol7Radio-quiet Radio-loud-20-21-22-23-24-25MR,bul (mag)

FIG.9.—SimilartoFigure6,exceptthathereweshowthedependenceonradioemission:radio-loudsourcesaremarkedwithbluestarsandsolidline,andradio-quietsourcesaremarkedwithredcirclesanddottedline.TheMBH−Lbulrelationforinactivegalaxiesisshownbythethickhatchedline.

thesensethatlower-redshiftsourceshaveahigherMMThistrend,however,isprobablyaselectionBHatagivenR,bul.ef-fectbecauselow-Eddingtonratio,later-typegalaxiestendtobecloser.Indeed,thez≤0.15subsamplehas󰀁L󰀁B/T󰀂=0.71,tobecomparedwith󰀁Lbol/LEdd󰀂=0.11and󰀂=0.94forthez>0.15subsample.

bol/LEdd󰀂=0.24and󰀁B/T5.DISCUSSION

5.1.WhichisthePrimaryVariable?

Wehaveassembledasampleoflocalmassivetype1AGNswithreliablespectroscopicdataandhostgalaxyphotometricmeasurementstoinvestigatetheoriginof

9) O• M / HBM8( gol7 z ≤ 0.15 z > 0.15 -20-21-22-23-24-25MR,bul (mag)

FIG.10.—SimilartoFigure6,exceptthathereweshowthedependenceonredshift:low-redshiftsourcesaremarkedwithredcirclesanddashedline,andhigh-redshiftsourcesaremarkedwithbluestarsandsolidline.TheMforinactivegalaxiesisshownbythethickhatchedline.

BH−Lrelationbul16KIMetal.

ratio,however,remains(Fig.11a).

Theaboveinterpretationoffersaplausibleexplanationfortheapparentlinkbetweenmorphologicaltypeandaccretionrate,whichotherwiseissomewhatperplexing.Withinoursam-pleitistheapparentlyearlier-type,morebulge-dominatedsys-temsthatactuallyhavehigheraccretionrates.ThisrunscountertothetrendnormallyseeninnearbyAGNs(e.g.,Heckmanetal.2004;Greene&Ho2007a)andthegeneraltendencyforpresent-dayearly-typegalaxiestobemoregas-poorthanlate-typegalaxies.However,ifhigh-mass,luminousAGNsre-sultfromtheaftermathofgas-richgalaxy-galaxymergers(e.g.,Sandersetal.1988;Hopkinsetal.2006),ourresultsimplythatitisduringthemostadvancedstagesofthemergerthatac-cretiononthecentralBHattainsitsmaximumrate.Someofthemosthighlyaccretingobjects(thosewithlargeLbol/LEdd)inoursample,whichcoincideamongthosewiththelargestoffsetintheMBH−Lbulrelation,alsohappentobeamongtheonesthatshowthemostconspicuousmorphologicalsignaturesoftidalperturbations,asmeasuredbytheFourierparametera1(seeFig.8).TheseincludeHE0354−5500(Lbol/LEdd=0.57;a1=0.33),PG1012+008(Lbol/LEdd=0.32;a1=0.32),PG1613+658(Lbol/LEdd=0.44;a1=0.40),andPG1700+518(Lbol/LEdd=0.71;a1=0.39).However,noteveryobjectwithahighLbol/LEddhasalargevalueofa1.Thismayimplythatnotallquasarepisodesaretriggeredbymajorgalaxyinterac-tions,orthataccretioncanproceedatasubstantialrateevenafterthetidalfeatureshavedisappeared.Totheextentthatthepeakstarformationrateinthemergerhasalreadysubsideddur-ingthisphase,ourscenariooffersanadditionalexplanationforwhytype1AGNscontainintermediate-agestars(Kauffmannetal.2003)butgenerallynotmuchconcurrentstarformation(Ho2005;Kimetal.2006).

Withinoursample,fourAGNshavecompactsources—plausiblysmallaccretedcompanions—projectedclosetotheprimaryhostgalaxy.Thesemaybeexamplesofminormergers.Nonehasalargevalueofa1.TwoofthefourlieexactlyontheMBH−Lbulrelationofinactivegalaxies(Fig.8)andhaverela-tivelylowEddingtonratios(Lbol/LEdd=0.05forPG1426+015andLbol/LEdd=0.06forPKS2135−14).Theothertwolieoffsetbelowtherelation,butonlyPG1302−102hasalargeLbol/LEdd(0.4);PHL1093hasLbol/LEdd=0.03.Thus,withinourlim-itedstatistics,wehavenoevidencethatminormergersplayasignificantroleinelevatingtheaccretionrateinAGNs.

Withinthisbackdrop,wecanofferatentativeexplanationforthezeropointdifferencebetweenradio-loudandradio-quietsources(Fig.9),onethatultimatelylinksthegenerationofpowerfulradiojetstotheBHaccretionrateand/orhostgalaxymorphology.Butfirstweshouldclarifysometerminology.Therearetwopopulardefinitionsof“radio-loud”AGNsintheliterature,anditisimportantnottoconfusethem.Onecommonusageofthistermreferstosourcesthatareclassifiedsolelybytheirradio-to-opticalfluxratio(R)asdefinedbyKellermannetal.(19)10,regardlessoftheirradiopower.Onthisbasis,thevastmajorityofAGNsinthelocalUniverse(Ho2008),mostwithextremelylowluminosities,qualifyasbeingradio-loud,withthedegreeofradio-loudnessincreasingwithdecreasingLbol/LEdd(Ho2002;Terashima&Wilson2003;Greeneetal.2006).Theradioemissioninmostoftheselow-powersourcesislargelydominatedbyacompactcore,andanyjet-likefea-turesareconfinedtosub-galacticscales.Thehostgalaxiesen-compassallmorphologicaltypes,includingspiralgalaxies(Ho

theintrinsicscatterinthecorrelationbetweenBHmassandbulgeluminosity.Assumingageometricalfactoroff=0.75fortheBLR,wefindthattheAGNsinoursampleliesystem-aticallybelowtheMBH−Lbulrelationofinactivegalaxiesbyanaverageoffsetof∆α≈−0.3dex.Moreover,wehaveshownthatthemagnitudeoftheoffsetcorrelateswithsecondarypa-rametersconnectedwiththeAGN(Eddingtonratioanddegreeofradio-loudness)andhostgalaxy(redshift,bulge-to-diskra-tio,andsignsofmorphologicaldisturbance).

AmongtheseveralvariablesthatcorrelatewiththeoffsetintheMBH−Lbulrelation,theonlyonethatappearsunphysicalisthatrelatedtoredshift.Aswenotedin§4.5,theapparentde-pendenceonredshiftmostlikelyreflectstheselectioneffectthathigherredshiftsourcestendtobebiasedtowardhigherEdding-tonratios(e.g.,Boyleetal.2000)andmoreluminous,earlierHubbletypes.Ifwedividethesampleintotwoatz=0.15,theMBH−Lbulrelationforboththenearbyanddistanthalvescon-tinuetoexhibitthedependenceonLbol/LEddandB/Tthatweseeforthefullsample.

Still,amongtherestofthevariablesthatcorrelatewiththezeropointoffsetintheMBH−Lbulrelation,whichisthepri-maryone?GiventhatmanygalaxyandAGNparametersaremutuallycorrelated,thisisnotatrivialquestiontoanswer.Weproposethattheprimaryphysicaldriveristhemassaccretionrate,asreflectedintheEddingtonratio.WearguethatthehostmorphologyanddegreeoftidaldisturbancedirectlyaffecttheAGNaccretionrate,andthattheaccretionrate,inturn,islinkedtotheradio-loudnessparameter.

Althoughour2-DfitsindicatethatthehostsofmanyofourAGNshaveanon-zerodiskcomponentapartfromabulge,itisimportanttorecognizethat,withfewexceptions(Fairall9,HE0306−3301,MS1059.0+7302,PG2130+99),mostofthesourcesinoursampledonothaveregular,normaldisks.Thevastmajorityofthesample—byconstructionwhenweim-posedtheMBHcut—isdecisivelybulge-dominated.Manyofthefeaturesthatweattributetoa“disk,”infact,aresimplydif-fuse,extendedfeaturesaboveandbeyondthedominantbulgecomponent,whichwehaveparameterizedusingasingleSérsicfunction.Thereisnoapriorireasonwhythebulgeshouldbedefinedinsuchamanner,thatitcannothaveamorecomplexlightdistribution,especiallyatlargeradii.Inotherinstances,theextra-bulgecomponentishighlydisturbedandalmostcer-tainlyoftidalorigin.Withfewexceptions(Bennertetal.2008),thesefeatureshavegenerallyneverbeenmeasuredbeforequan-titativelyinAGNhostgalaxies.However,itisentirelydebat-ablewhetheranyofthesestructurestrulybelongstoorwilleversettleintoanormaldiskcomponent.Instead,wesurmisethatmanyofthetidaltailsandextended,distortedfeatures,infact,shouldbeconsideredaspartofthebulgeinformation.Theyarereminiscentofmorphologicalsignaturesattributedtothelate,advancedstagesofgas-richmergers(e.g.,Barnes&Hernquist1996;Lotzetal.2008)orpossiblyevengas-poor(“dry”)mergers(e.g.,vanDokkum2005;Naabetal.2006).Plausibleexamplesofthisphenomenoninoursampleinclude[HB]0316−346(Fig.15),HE1434−1600(Fig.19),andPG1012+008(Fig.32).Insupportofthishypothesis,Fig-ure11billustratesthatthemorphologicalsegregationseenintheMBH−Lbulrelation(Fig.7)essentiallydisappearswhenthebulgeluminosityisreplacedwiththetotalluminosityofthehost.Thescatteralsogoesdownslightly,from0.40dexto0.36dexintheMBH−Lhostrelation.ThedependenceonEddington

10Terashima

&Wilson(2003)advocatedacloselyrelatedradio-loudnessparameterRXbasedontheradio-to-X-rayfluxratio.

BlackHoleMassvs.BulgeLuminosityRelation17

(a)99(b)log (MBH / M O• )8log (MBH / M O• )87low Lbol/LEddhigh Lbol/LEdd-20-21-22-23MR,host (mag)-24-257 B/T < 0.50.5 ≤ B/T < 1 B/T = 1-20-21-22-23MR,host (mag)-24-25FIG.11.—CorrelationbetweenBHmassandabsoluteR-bandmagnitudeforthetotalemissionfromthehostgalaxy.(a)ThesymbolsarethesameasinFigure6;thedependenceonEddingtonratiostillremains.(b)ThesymbolsarethesameasinFigure7;thedependenceongalaxymorphologyismuchweaker.TherelationbetweenMBHandbulgeluminosityforinactivegalaxiesisdenotedbythethickhatchedline.

&Peng2001).TheMilkyWay’sSgrA⋆isafamiliarexample.Thesecondcommonlyuseddefinitionofradio-loudnessislessclear-cut,butitinvolvessomecombinationofrelative(R≫10)

23−24

WHz−1)measuresofradiopower.andabsolute(Prad>∼10

Whendetected,theradiojetshavesuper-galacticdimensionsandarehighlycollimated.Theradio-loudsourcesinthisstudy,

tot

withmedianR=820andP6cm=7×1025WHz−1,belongtothissecondcategory.Strongradiosourcesofthisvarietyin-variablyresideinearly-typegalaxies(e.g.,McLureetal.1999)andareassociatedwithhighaccretionrates(e.g.,Maccaroneetal.2003;Kördingetal.2006).Nevertheless,onlyaminorityofhighlyaccretingAGNsinmassive,early-typehostgalaxiesareradio-loud.Thereisnouniversallyacceptedexplanationforthislongstandingquandary.Onepossibilityisthatanecessaryingredientforthegenerationofpowerful,collimatedjetsistheexistenceofaBHwithalargespin(e.g.,Sikoraetal.2007),whichismoreeasilyattainedinmerger-drivenaccretioneventsduringtheformationofgiantellipticalgalaxiesthanindiskorspiralgalaxies(Volonterietal.2007).

5.2.PhysicalInterpretation

TheprincipalconclusionsofourstudyarethatthezeropointoftheMBH−LbulrelationforAGNsisoffsetfromthatofinac-tivegalaxiesandthatthemagnitudeoftheoffsetcorrelateswithseveralphysicalpropertiesoftheAGNandhostultimatelycon-nectedtotheaccretionrate.TheseeffectsaccountforthebulkoftheobservedintrinsicscatterintheAGNMBH−Lbulrelation.Interestingly,otherstudiesofnearbyAGNsusingdifferentprobesofthehostgalaxyhaveindependentlyarrivedatverysimilarconclusions.Onkenetal.(2004;seealsoNelsonetal.2004),analyzingasmallsampleoftype1AGNswithavailablestellarvelocitydispersionsandBHmassesdeterminedthroughreverberationmapping,concludedthattheactivesystemslieoffsetbelowtheMBH−σ⋆relationofinactivegalaxiesby∼0.2dexifoneassumesasphericalBLR.Greene&Ho(2006b)ob-

tainedanalmostidenticalresultusingamuchlargersampleoftype1AGNswithmeasuredσ⋆andMBHestimatedfromsingle-epochspectra.TheseconclusionshavebeenreaffirmedbyShenetal.(2008)intheiranalysisofcompositeAGNspectraderivedfromtheSloanDigitalSkySurvey.AdifferentapproachwastakenbyHoetal.(2008)through21cmHIobservations.Us-ingthemaximumrotationalvelocityandtotaldynamicalmassofthegalaxyasnewvariablestorepresentthegravitationalpo-tentialofthehost,theyfindthatbothquantitiesstronglycorre-latewithMBH.Inqualitativeagreementwiththeresultsfromthisstudy,Hoetal.findthatthezeropointinthescalingrela-tionsdependsprimarilyontheaccretionrateinthesensethat,atagivengalaxyrotationvelocityandespeciallydynamicalmass,AGNswithhigherEddingtonratioshavesystematicallylowerMBHthanthosewithlowerEddingtonratios.

Thesetrendscanbeinterpretedinoneoftwoways.Inthefirstinstance,wenotethatAGNs,byselection,haveactivelygrowingBHs.Wecanenvisionthat,atafixedbulgepoten-tial(velocitydispersion,luminosity,mass),activegalaxieshavelessmassiveBHsthaninactivegalaxiesifthebulkofthestarformationprecedes,andisnotwellsynchronizedwith,ama-joraccretionevent.Thisparticulartimesequence,whichisre-quiredinordertoimprintanetnegativeoffsetintheMBHversushostgalaxyrelations,seemstobesupportedbytheprevalenceofpost-starburstsignatures(Canalizo&Stockton2001;Kauff-mannetal.2003)aswellasthelowlevelsofongoingstarformation(Ho2005;Kimetal.2006)foundintype1AGNs.IftheAGNphaselasts,say,for∼108yrs,whichisneartheupperendofthecurrentlyestimatedlifetimes(Martini2004),thena108M⊙BHwouldincreaseitsmassbyafactorof2(∼0.3dex)ifitisradiatingatLbol/LEdd=0.5witharadiativeefficiencyof0.1.Thisexampleismerelyillustrative.Inreality,theAGNsinoursamplespanawiderrangeofLbol/LEddandtheAGNlife-timesmaybeshorter.Nevertheless,onaverageweexpectlumi-nousAGNstoliesystematicallybelowinactivegalaxiesonthe

18KIMetal.

MBH−Lbulrelation.Moreover,ourstudy,asdothoseofHoetal.(2008)andShenetal.(2008),furthershowsthatthemagni-tudeofthezeropointoffsetdependsontheEddingtonratio:theBHmassesofhigh-EddingtonratioAGNshavemorecatchinguptodothantheBHmassesoflower-EddingtonratioAGNs.Suchasystematictrendcanonlycomeaboutiftheaccretionratedirectlyrelatestotheevolutionaryphaseoftheaccretionevent.Thisseemsplausible,inlightoftheapparentassociationbetweenaccretionrateandthedegreeoftidalperturbation.InsteadoftheBHbeingundermassive,perhapsitisactuallythebulgethatisoverluminous,by∆MthepossibilitythattheR,luminositybul≈0.5−0.6mag.First,wedismissenhance-mentcouldbeduetocontaminationfromnebularemissionfromthenarrow-lineregion.Althoughthespatialextentofthenarrow-lineregioninquasarscanreachseveralkpc(Ben-nertetal.2002),substantiallyoverlappingwiththebulge,thetypical[OIII]luminositiesinoursample(󰀁Lergss−1)contributelessthan3%totheluminosity[OIII]󰀂=1042ofthebulge(󰀁MR,bul󰀂=−22.75mag).Giventheevidenceoutlinedin§5.1thatthemostextremeoutliersintheMarecentmergerortidalinteraction,BH−Lbulseemtohaveundergoneamorelikelypossibilityisthatthebulgeluminositymaybemoderatelyen-hancedbythelatestepisodeofcentralstarformation.Indeed,forasmallsampleofreverberation-mappedSeyfert1galaxieswithstellarvelocitydispersionandbulgeluminositymeasure-ments,Nelsonetal.(2004)haveshownthattheseobjectsaresomewhatbrighter(∼0.4mag)thaninactivegalaxiesatagivenvelocitydispersion.TheyattributedthisoffsetintheFaber-Jackson(1976)relationtoyoungerstellarpopulationsinAGNs.BoostingtheR-bandluminosityby0.5magrequires∼15%ofthestellarmasstocomefroma1Gyrpopulationwithsolarmetallicity(Nelsonetal.2004).WhilethisoffersaplausibleexplanationfortheoffsetintheMthefactBH−thatLbulrelationseeninoursample,itcannotaccountforthemostrecentandlargestsamplesofnearbytype1AGNs,statisticallyatleast,shownegativeoffsets(withrespecttoinactivegalaxies)whenMBHiscomparedtoallbulgeorhostgalaxyparameters(stellarvelocitydispersion,bulgeluminosity,totalgalaxydynamicalmass).Thedirectionoftheoffsetisthesame(atagivengalaxyparameter,MBHislower),andthemagnitudeoftheoffsetisalsosimilar(∼0.2−0.3dex).Althoughthisclearlyneedstobeverifiedwithalargesamplethathasreliablemeasurementsofbothvelocitydispersionandbulgeluminosities,themostrecentstudiessuggest,contrarytoNelsonetal.(2004),thatlocaltype1AGNsactuallydonotdepartfromthestandardFaber-Jackson(1976)relation.Furthermore,foralargesampleofsourceswithmeasurementsofrotationalvelocityandtotalgalaxyluminos-ity,Hoetal.(2008)showthattype1AGNsshownoobvi-ousdeviationsfromtheTully-Fisher(1977)relationofinactivegalaxies.Inlightoftheseconsiderations,wefavortheviewthatthenegativeoffsetintheManexcessBHin−LLbulrelationrepresentsadeficitinMBHratherthanbul.

Alternatively,wecanassertthatbothactiveandinactivegalaxiesintrinsicallyshouldobeythesameBHmassversushostscalingrelations.Fromthisstandpoint,thezeropointoff-setbetweenAGNsandinactivegalaxies,aswellasthevaria-tionsoftheoffsetwithLbol/LEdd,canbeviewedasasystem-aticunderestimateofthetruevalueofMarebasedonavirialproductassumingBH.RecallthatourBHmassesasphericaldistributionofBLRcloudswithisotropicvelocities,forwhichthegeometricfactorisf=0.75.If,forexample,theBLR(oratleasttheportionofitthatpredominantlyemitstheBalmer

lines)hasaflattened,disk-likegeometrywithkinematicsdom-inatedbyrotation,andonthescaleoftheBLRtype1sourceshappentobepreferentiallymoreface-ontoourlineofsight,thenwesystematicallyunderestimatethedeprojectedrotationvelocityandhenceMtointerpretBH.Wu&Han(2001)invokedthislineofreasoningtheoffsetoftype1AGNsontheMrelationandconcludedthatonaveragetheirBLRsareinclinedBH−σ⋆by󰀁i󰀂≈36◦.Thesameargumentcan,inprinciple,beappliedtotheobservedoffsetintheMfectofinclinationprobablyentersBH−atLsomebulrelation.Whiletheef-level(seealsoCollinetal.2006),itcannotaccountforthefactthatthemagnitudeoftheoffsetdependsonLbeingbol/LunderestimatedEdd.Thelatterisnotatrivialconse-quenceofthemassbecausesourceswithhighLbol/LEddtrulydoexhibitcharacteristicallydistinctX-ray,optical,andradioproperties(Bolleretal.1996;Boroson2002;Greeneetal.2006).Whateverthephysicaloriginoftheoffset(Onkenetal.2004;Collinetal.2006;Marconietal.2008),wecanempiricallyadjustthenormalizationfactorofthevirialproductbyforcingtheAGNsampletoagreewiththefiducialreferenceofinactivegalaxies.Toremovethezeropointoffsetof∆α≈−0.2to−0.3dex,then,thenormalizationshouldbein-creasedbyafactorof∼1.6−2,fromf=0.75tof≈1.2−1.5.Forthemostextremeoffsetsof∆α≈−0.6,fincreasesto∼3.Withoutadditionalinformation,unfortunately,theabovetwoalternativeexplanations—anundermassiveMffactor—aredegenerate.ItisBHversusanun-derestimatedeasytoimaginethatbotheffectsmustoperatejointly.Ontheonehand,theBHsinAGNsare,afterall,gainingmass.Ontheotherhand,asdiscussedinCollinetal.(2006),therearemultiplereasonstobelievethattheBLRhasanonsphericalgeometryandthattheEddingtonratiomayinfluenceitsstructureanddynamics.TheonlywaytoresolvethisdegeneracyistoobtainindependentestimatesofMBHforAGNsthatdonotrelyontheBLRvirialtechnique.Todate,effortstoapplyresolvedstellardynamicaltechniquestoreverberation-mappedAGNshaveyieldedveryroughestimatesofMetal.2006;BHforonlyacoupleofsources(NGC3227:DaviesHicks&Malkan2008;NGC4151:Onkenetal.2007),andthusattemptingtocross-calibratethetwotechniquesisstillfartoopremature.BHmassestimatorsbasedonX-rayvariabilityseemmorepromising.Gierli´nskietal.(2008;seealsoHayashidaetal.1998)findthat,foraccret-ingBHsintheirhardspectralstate,theamplitudeoftheirhigh-frequencyX-rayvariabilityscalesinverselywithMForasmallsubsetofBHoveraverywiderangeofmasses.nearbytype1AGNs,theX-ray-derivedmassesshowroughagreementwithMf≈1.2.BHobtainedthroughreverberationmappingassumingAverysimilarconclusionwasreachedbyNikołajuketal.(2006).FromcomparisonofMBHforreverberation-mappedsourceswithmassesobtainedusingtheX-rayexcessvariancemethod,theseauthorsestimatef=1.06±0.26,which,inter-estingly,liesinbetweenthevaluesoftheffactorforthetwoextremealternativesdiscussedabove.

AssumingthattheX-ray-derivednormalizationfactortrulydoesrepresentthecorrectnormalizationfactorforthevirialmasses,thentheinferredgrowthratesforMBHinluminousAGNsaremuchmoremodest,fromtypicallyaslittleas10%–40%(0.05–0.15dex)toatmost280%(0.45dex).

Still,wenotethatthetendencyforBHsinlocalAGNstobelessmassivethantheBHsininactivegalaxiesofsimilartyperunscountertothetrendobservedathigherredshift.Alreadybyz=0.36,type1AGNsbegintodepartfromthelocalMandM−LBH−σ⋆BHbulrelations(Wooetal.2006;Treuetal.2007),

BlackHoleMassvs.BulgeLuminosityRelation

butintheoppositedirectionasthatseenatlowerredshifts:atagivenσ⋆orLbul,AGNsareoffsetcomparedtolocalinactivesystemsby∆logMBH≈+0.5dex.Thistrendhasnowbeenex-tendedbyWooetal.(2008)outtoz=0.57usingstellarvelocitydispersionmeasurements.Atevenhigherredshifts,directσ⋆measurementsarenolongerfeasible,butsurrogateestimatesofσ⋆usingnarrowemissionlines(Salvianderetal.2007)aswellasprobesofthehostgalaxyusingimaging(Pengetal.2006a,b)andCOemissionlines(Shieldsetal.2006;Ho2007)sup-portthenotionthatthegrowthoftheBHsinAGNshavebeendecoupledfrom,andoutpaced,theunderlyinghost.

6.SUMMARY

19

Weperformedtwo-dimensionalstructuraldecompositionofasampleof45nearby(z<∼0.35)type1AGNswithavailablearchivalopticalHSTimagesandpublishedspectroscopicdata.WecalculatedvirialBHmassesassumingasphericalBLRwithisotropicvelocities.UsinganewversionoftheversatilecodeGALFIT,wederiveddetailedfitstothestructuralcomponentsofthehostgalaxies,yieldingnotonlyrobustmeasurementsofthebulgeluminositieswithrealisticerrorbarsbutalso,forthefirsttime,quantitativeestimatesofnonaxisymmetricfeaturessuchasextendeddisksandtidalarms.

OurprincipalaimistounderstandtheoriginoftheintrinsicscatterintheMBH−LbulofactivegalaxiesoverarestrictedrangeofBHmasses(MBH≈108.5±0.5M⊙).WhileAGNscloselyfol-lowtheMBH−Lbulrelationofinactivegalaxies,wefindthattheintrinsicscatterissubstantial(0.40dex)andthatthezeropointoftherelationisshiftedbyanaverageof∆logMBH≈−0.3dex.ThemagnitudeofthezeropointoffsetintheMBH−Lbul

relationdependsonpropertiesoftheAGN(Eddingtonratioandradio-loudnessparameter)andthehostgalaxy(morphologicaltypeanddegreeoftidalperturbation).Wearguethattheprin-cipalphysicalparameterresponsibleforthevariationinzeropointistheBHaccretionrate,asreflectedintheEddingtonra-tio.WesuggestthatgalaxymergersandtidalinteractionsplayasubstantialroleinboostingtheaccretionrateinthissampleofAGNs.AsignificantfractionofthezeropointoffsetintheMBH−LbulrelationcanbeexplainedifthevirialBHmasseshavebeenunderestimated,asindicatedfromcomparisonwithindependentmassesderivedfromX-rayvariabilitytechniques.AfteraccountingforthischangeinthenormalizationofthevirialBHmassscale,weestimatethatBHsduringtheAGNphaseexperienceamodestgrowthof∼10%–40%inmass.Wearegratefultotherefereeforprovidingatimelyandhelp-fulreview.WethankJamesDunlopandRossMcLureforusefuldiscussions.ThisworkwassupportedbytheCarnegieInsti-tutionofWashingtonandbyNASAthroughgrantsHST-AR-10969andHST-GO-9763fromtheSpaceTelescopeScienceInstitute,whichisoperatedbytheAssociationofUniversitiesforResearchinAstronomy,Inc.,underNASAcontractNAS5-26555).M.K.andM.I.acknowledgethesupportoftheKoreaScienceandEngineeringFoundation(KOSEF)throughtheAs-trophysicalResearchCenterfortheStructureandEvolutionoftheCosmos(ARCSEC).C.Y.P.isgratefulforsupportthroughthePlaskettFellowshipprogramofHerzbergInstituteofAstro-physicsandtheSTScIInstituteFellowshipprogram.ResearchbyA.J.B.wasalsosupportedbybyNSFgrantAST-0548198.

REFERENCES

Barnes,J.E.,&Hernquist,L.E.1996,ApJ,471,115

Barth,A.J.,Greene,J.E.,&Ho,L.C.2005,ApJ,619,L151Barvainis,R.,&Antonucci,R.19,ApJS,70,257

Becker,R.H.,White,R.L.,&Edwards,A.L.1991,ApJS,75,1Becker,R.H.,White,R.L.,&Helfand,D.J.1995,ApJ,450,559

Bennert,N.,Canalizo,G.,Jungwiert,B.,Stockton,A.,Schweizer,F.,Peng,C.Y.,&Lacy,M.2008,ApJ,677,846

Bennert,N.,Falcke,H.,Schultz,H.,Wilson,A.S.,&Wills,B.J.2002,ApJ,574,L105

Bennett,C.L.,Lawrence,C.R.,Burke,B.F.,Hewitt,J.N.,&Mahoney,J.1986,ApJS,61,1

Bentz,M.C.,Peterson,B.M.,Pogge,R.W.,Vestergaard,M.,&Onken,C.A.2006,ApJ,4,133

Bettoni,D.,Falomo,R.,Fasano,G.,&Govoni,F.2003,A&A,399,869Boller,T.,Brandt,W.N.,&Fink,H.1996,A&A,305,53Boroson,T.A.2002,ApJ,565,78——.2005,AJ,130,381

Bower,R.G.,etal.2006,MNRAS,370,5

Boyle,B.J.,Shanks,T.,Croom,S.M.,Smith,R.J.,Miller,L.,Loaring,N.,&Heymans,C.2000,MNRAS,317,1014

Calzetti,D.,Kinney,A.L.&Storchi-Bergmann,T.1994,ApJ429,582Canalizo,G.,&Stockton,A.2001,ApJ,555,719

Collin,S.,Kawaguchi,T.,Peterson,B.M.,&Vestergaard,M.2006,A&A,456,75

Condon,J.J.,Cotton,W.D.,Greisen,E.W.,Yin,Q.F.,Perley,R.A.,Taylor,G.B.,&Broderick,J.J.1998,AJ,115,1693

Conselice,C.J.,Bershady,M.A.,&Jangren,A.2000,ApJ,529,886Croton,D.J.,etal.2006,MNRAS,365,11Davies,R.I.,etal.2006,ApJ,6,754

deJong,R.S.,Simard,L.,Davies,R.L.,Saglia,R.P.,Burstein,D.,Colless,M.,McMahon,R.,&Wegner,G.2004,MNRAS,355,1155deVaucouleurs,G.1948,Ann.d’Astrophys.,11,247

DiMatteo,T.,Springel,V.,&Hernquist,L.2005,Nature,433,604

Dunlop,J.S.,McLure,R.J.,Kukula,M.J.,Baum,S.A.,O’Dea,C.P.,&Hughes,D.H.2003,MNRAS,340,1095

Eracleous,M.,&Halpern,J.P.1994,ApJS,90,1Faber,S.M.,&Jackson,R.E.1976,ApJ,204,668Ferrarese,L.,&Merritt,D.2000,ApJ,539,L9

Floyd,D.J.E.,Kukula,M.J.,Dunlop,J.S.,McLure,R.J.,Miller,L.,Percival,W.J.,Baum,S.A.,&O’Dea,C.P.2004,MNRAS,355,196Gebhardt,K.,etal.2000,ApJ,539,L13

Gierli´nski,M.,Nikołajuk,M.,&Czerny,B.2008,MNRAS,383,741

Giveon,U.,Maoz,D.,Kaspi,S.,Netzer,H.,&Smith,P.S.1999,MNRAS,306,637

Granato,G.L.,DeZotti,G.,Silva,L.,Bressan,A.,&Danese,L.2004,ApJ,600,580

Greene,J.E.,&Ho,L.C.2005,ApJ,630,122——.2006a,ApJ,1,L21——.2006b,ApJ,1,117——.2007a,ApJ,667,131——.2007b,ApJ,670,92

Greene,J.E.,Ho,L.C.,&Barth,A.J.2008,ApJ,inpressGreene,J.E.,Ho,L.C.,&Ulvestad,J.S.2006,ApJ,636,56

Griffith,M.R.,Wright,A.E.,Burke,B.F.,&Ekers,R.D.1994,ApJS,90,179——.1995,ApJS,97,347

Grupe,D.,Wills,B.J.,Leighly,K.M.,&Meusinger,H.2004,AJ,127,156Häring,N.,&Rix,H.-W.2004,ApJ,604,L

Hayashida,K.,Miyamoto,S.,Kitamoto,S.,Negoro,H.,&Inoue,H.1998,ApJ,500,2

Heckman,T.M.,Kauffmann,G.,Brinchmann,J.,Charlot,S.,Tremonti,C.,&White,S.D.M.2004,ApJ,613,109

Hicks,E.K.S.,&Malkan,M.A.2008,ApJS,174,31Ho,L.C.2002,ApJ,5,120——.2005,ApJ,629,680——.2007,ApJ,669,821

——.2008,ARA&A,inpress(astro-ph/0803.2268)

Ho,L.C.,Darling,J.,&Greene,J.E.2008,ApJ,inpress(arXiv:0803.1952)Ho,L.C.,&Peng,C.Y.2001,ApJ,555,650

Hopkins,P.,Hernquist,L,Cox,T.J.,DiMatteo,T.,Robertson,B.,&Springel,V.2006,ApJS,163,1

Kaspi,S.,Maoz,D.,Netzer,H.,Peterson,B.M.,Vestergaard,M.,&Jannuzi,B.T.2005,ApJ,629,61

Kaspi,S.,Smith,P.S.,Netzer,H.,Maoz,D.,Jannuzi,B.T.,&Giveon,U.2000,ApJ,533,631

Kauffmann,G.,etal.2003,MNRAS,346,1055

Kellermann,K.I.,&Pauliny-Toth,I.I.K.1973,AJ,78,828

Kellermann,K.I.,Sramek,R.,Schmidt,M.,Shaffer,D.B.,&Green,R.19,AJ,98,1195

Kim,M.,Ho,L.C.,&Im,M.2006,ApJ,2,702

Kim,M.,Ho,L.C.,Peng,C.Y.,Barth,A.J.,&Im,M.2008,ApJS,inpressKim,M.,Ho,L.C.,Peng,C.Y.,&Im,M.2007,ApJ,658,107

20KIMetal.

Onken,C.A.,Ferrarese,L.,Merritt,D.,Peterson,B.M.,Pogge,R.W.,Vestergaard,M.,&Wandel,A.2004,ApJ,615,5

Peng,C.Y.,Ho,L.C.,Impey,C.D.,&Rix,H.-W.2002,AJ,124,266

Peng,C.Y.,Impey,C.D.,Ho,L.C.,Barton,E.J.,&Rix,H.-W.2006a,ApJ,0,114

Peng,C.Y.,Impey,C.D.,Rix,H.-W.,Kochanek,C.S.,Keeton,C.R.,Falco,E.E.,Lehár,J.,&McLeod,B.A.2006b,ApJ,9,616Peterson,B.M.,etal.2004,ApJ,613,682

Ridgway,S.E.,Heckman,T.M.,Calzetti,D.,&Lehnert,M.2001,ApJ,550,122

Robertson,B.,Hernquist,L.,Cox,T.J.,DiMatteo,T.,Hopkins,P.F.,Martini,P.,&Springel,V.2006,ApJ,1,90

Salviander,S.,Shields,G.A.,Gebhardt,K.,&Bonning,E.W.2007,ApJ,662,131

Sanders,D.B.,Soifer,B.T.,Elias,J.H.,Madore,B.F.,Matthews,K.,Neugebauer,G.,&Scoville,N.Z.1988,ApJ,325,74

Schade,D.J.,Boyle,B.J.,&Letawsky,M.2000,MNRAS,315,498

Sérsic,J.L.1968,AtlasdeGalaxiasAustrales(Córdoba:Obs.Astron.,Univ.Nac.Córdoba)

Shen,J.,VandenBerk,D.E.,Schneider,D.P.,&Hall,P.B.2008,AJ,135,928Shields,G.A.,Menezes,K.L.,Massart,C.A.,&VandenBout,P.2006,ApJ,1,683

Sikora,M.,Stawarz,L.,&Lasota,J.-P.2007,ApJ,658,815Simien,F.,&deVaucouleurs,G.1986,ApJ,302,5Spergel,D.N.,etal.2003,ApJS,148,175

Terashima,Y.,&Wilson,A.S.2003,ApJ,583,145Tremaine,S.,etal.2002,ApJ,574,740

Treu,T.,Woo.,J.-H.,Malkan,M.A.,&Blandford,R.D.2007,ApJ,667,117Tully,R.B.,&Fisher,J.R.1977,A&A,54,661VandenBerk,D.E.,etal.2001,AJ,122,549vanDokkum,P.G.2005,AJ,130,27Vestergaard,M.2002,ApJ,571,733

Vestergaard,M.,&Peterson,B.M.2006,ApJ,1,6

Volonteri,M.,Sikora,M.,&Lasota,J.-P.2007,ApJ,667,704

Woo,J.-H.,Treu,T.,Malkan,M.A.,&Blandford,R.D.2006,ApJ,5,900——.2008,ApJ,inpress(astro-ph/0804.0235)Wu,X.-B.,&Han,J.L.2001,ApJ,561,L59

Zheng,X.Z.,Xia,X.Y.,Mao,S.,Wu,H.,&Deng,Z.G.2002,AJ,124,18

Kinney,A.L.,Calzetti,D.,Bohlin,R.C.,McQuade,K.,Storchi-Bergmann,T.&Schmitt,H.R.1996,ApJ,467,38

Körding,E.G.,Jester,S.,&Fender,R.P.2006,MNRAS,372,1366Kormendy,J.,&Kennicutt,R.C.,Jr.2004,ARA&A,42,603Kormendy,J.,&Richstone,D.O.1995,ARA&A,33,581

Krist,J.1995,inAstronomicalDataAnalysisSoftwareandSystemsIV,ed.R.A.Shaw,H.E.Payne,&J.J.E.Hayes(SanFrancisco:ASP),349

Kukula,M.,Dunlop,J.S.,McLure,R.J.,Miller,L.,Percival,W.J.,Baum,S.A.,&O’Dea,C.P.2001,MNRAS,326,1533Laor,A.2000,ApJ,543,L111

Letawe,G.,Courbin,F.,Magain,P.,Hilker,M.,Jablonka,P.,Jahnke,K.,&Wisotzki,L.2004,A&A,424,455

Letawe,G.,Magain,P.,&Courbin,F.2008,A&A,480,69

Lotz,J.M.,Jonsson,P.,Cox,T.J.,&Primack,J.R.2008,MNRAS,submitted(astro-ph/0805.1246)

Lotz,J.M.,Primack,J.,&Madau,P.2004,AJ,128,163

Maccarone,T.J.,Gallo,E.,&Fender,R.2003,MNRAS,345,L19Magorrian,J.,etal.1998,AJ,115,2285

Marconi,A.,Axon,D.J.,Maiolino,R.,Nagao,T.,Pastorini,G.,Pietrini,P.,Robinson,A.,&Torricelli,G.2008,ApJ,678,693Marconi,A.,&Hunt,K.L.2003,ApJ,5,L21

Martini,P.2004,inCarnegieObservatoriesAstrophysicsSeries,Vol.1:CoevolutionofBlackHolesandGalaxies,ed.L.C.Ho(Cambridge:CambridgeUniv.Press),170

Marziani,P.,Sulentic,J.W.,Zamanov,R.,Calvani,M.,Dultzin-Hacyan,D.,Bachev,R.,&Zwitter,T.2003,ApJS,145,199

Mauch,T.,Murphy,T.,Buttery,H.J.,Curran,J.,Hunstead,R.W.,Piestrzynski,B.,Robertson,J.G.,&SadlerE.M.2003,MNRAS,342,1117McLure,R.J.,&Dunlop,J.S.2001,MNRAS,327,199——.2002,MNRAS,331,795

McLure,R.J.,Kukula,M.J.,Dunlop,J.S.,Baum,S.A.,O’Dea,C.P.,&Hughes,D.H.1999,MNRAS,308,377

Naab,T.,Khochfar,S.,&Burkert,A.2006,ApJ,636,L81

Nelson,C.H.,Green,R.F.,Bower,G.,Gebhardt,K.,&Weistrop,D.2004,ApJ,615,652

Nikołajuk,M.,Czerny,B.,Ziolkowski,J.,&Gierli´nski,M.2006,MNRAS,370,1534

Novak,G.S.,Faber,S.M.,&Dekel,A.2006,ApJ,637,96Onken,C.A.,etal.2007,ApJ,670,105

APPENDIX

NOTESONINDIVIDUALOBJECTS

Commentsonthefittingresultsforindividualobjectsaregivenhere.

3C59(Fig.12)—ThehostcanbefitwithclassicalbulgerepresentedbyadeVaucouleurs(n=4)profile.

E1821+3(Fig.13)—Thebestfitshowsthatthebulgeisslightlydisturbed(a1=0.17).Thereisnoevidenceofadisk.

Fairall9(Fig.14)—Thehostrequiresabulgeandadisk.Aring-likestructure,whichwedonotmodel,isalsoseenintheresiduals.Thebulgeappearsbequitecompact,withaneffectiveradiusof1.6pixels,althoughitisnotwelldecomposedfromthenucleus.Thus,thebulgeluminositymightbehighlyuncertain.

[HB]0316−346(Fig.15)—Highlydisturbedobjectwithprominenttidalfeatures.Wefitthehostwithann=4bulgeandtwodiskcomponentswithFouriermodes.Oneofthediskcomponentsisnotcenteredonthenucleus.

[HB]2201+315(Fig.16)—Thisobjectwasobservedintwodifferentfilters(F555WandF702W),butbothimagesaresaturatedinthecore.Thebulgeluminositiesderivedfromthetwodifferentimages(correctedtotheRband)differby0.7mag.Wefitthehostwithasinglen=4bulgeinbothimages.Theresidualimagefromthelong(2700s)exposureshowspossiblesignsofanextendeddisk.

HE0306−3301(Fig.17)—Thebest-fitresultshowsanelongatedbulgecomponent(b/a≈0.45)andadiskwithaspiralarm.Thesignal-to-noiseratio(S/N)oftheobservedPSFstarislowerthanthatofthescienceimage.

HE0354−5500(Fig.18)—Thisappearstobeamergingsystem.Wefitthehostgalaxywithabulge(n=4)andanoff-centeredtidal-likefeaturewithahigh-amplitudeFouriermode(a1=0.33).TheS/NoftheobservedPSFstarislowerthanthatofthescienceimage.

HE1434−1600(Fig.19)—Wefitthehostwithonlyasinglebulgecomponent(mbul≈17.5mag),althoughtheresidualimageshowsevidenceforarcsandripples,whichmaybeevidenceofarecentcollision(Letaweetal.2004).Alternatively,ifthehostisfitwithtwocomponents,thebulgemagnitudebecomes18.6mag.TheS/NoftheobservedPSFstarislowerthanthatofthescienceimage.

MC1635+119(Fig.20)—Wefitthehostwithasinglebulgecomponent(n=4).

MRK1048(Fig.21)—Thisobjecthasaclosecompanionandanextremelylargebutfaint,off-centertidalfeature.However,neitherofthesehasalargeeffectonthefittingresult.Wefitthehostwithaclassicalbulge(n=4)andatidaltail.

MS0244.8+1928(Fig.22)—Thehostisreasonablywellrepresentedbyann=4bulge,althoughtheresidualsindicatethattheremightbeanadditionalfaint,outerenvelope.

MS0754.6+3928(Fig.23)—Thehostisreasonablywellrepresentedbyann=4bulge,althoughtheresidualsindicatethattheremightbeanadditionalfaint,outerenvelope.

BlackHoleMassvs.BulgeLuminosityRelation21

MS1059.0+7302(Fig.24)—Thebest-fittingmodelrequiresabulgeandadisk.Thereisaring-likestructureintheresidualimage.

MS1545.3+0305(Fig.25)—Thebestfitrequiresabulge(n=4)andadisturbedexponentialdisk.Aring-likestructureisprominentintheoriginalandresidualimages.

OX169(Fig.26)—Therearesignificantresidualsaftersubtractingthebest-fitobservedPSF,whichmightbeduetoPSFmismatch.ThefitwiththeTinyTimPSFresultsinabrighterAGNby0.1mag.Thehostcontainsabulge(n=3.9)andaprominent,extended,asymmetricstructurethatisfitwithaSérsiccomponentwithasmalln(∼0.2).

PG0052+251(Fig.27)—Itappearstohaveatidallydisturbedspiralarm,whichmayberelatedtothetwosmallgalaxiesinitsvicinity.Wefitthisobjectwithaclassicalbulge(n=4)andatruncateddiskwithasmalln(∼0.2).

PG0804+761(Fig.28)—Thereappearstobeafaintcentralfeaturethatresemblesabarorhighlyinclineddisk-likestructure.Thebulgeluminosity,however,ishardlyaffectedbythebarcomponent.

PG0923+201(Fig.29)—Thehostisfitwithasinglebulgecomponent(n=4).Therearethreenearbycompanionsthatwerefitsimultaneously.

PG0953+414(Fig.30)—Thehostisfitwithasinglebulgecomponent(n=4),whichappearsslightlydisturbedbasedontheamplitudeoftheFouriermode(a1=0.12).

PG1004+130(Fig.31)—Thehostisfitwithasinglebulgecomponent(n=4).TheresidualimageshowssignificantPSFmismatch,whichmightaffectthefittingresult.

PG1012+008(Fig.32)—Theimageofthehostshowsclearevidenceofinteractionwithaspiralgalaxyandasmaller,compactneighbor.Theprimaryhostcanbefitwithasinglebulgecomponent(n=4),buttheresidualimageshowssignificantstructure.PG1116+215(Fig.33)—TheresidualimageindicatesthatthecentralcoreisslightlyaffectedbyPSFmismatch.Thehostgalaxy,however,iswell-representedbyasingle-componentbulgewithnfixedto4.Ifweallowntobefree,thebest-fittingvalueofn=1.84yieldsabulgeluminositythatis0.3magfainter.

PG1202+281(Fig.34)—Thissourcehasanumberofnearbygalaxies,thebrightestandnearestofwhichwefitsimultaneously.Theprimaryhostiswelldescribedbyasingle-componentbulge(n=4).

PG1211+143(Fig.35)—LikePG0804+761,thereappearstobeafaintcentralfeaturethatresemblesabarorhighlyinclineddisk-likestructure.ItisunclearifthisisanartifactduetoPSFmismatch.Dependingonwhetherthisextracomponentisincluded,thebulgeluminosityrangesbetween16.7and17.2mag,withabest-fitvalueof16.9mag.

PG1226+023(Fig.36)—Becauseofthestrongbleedingregionsfromthesaturatedcore,wedidnotperformnonparametricaperturephotometry.Thehostisreasonablywellfitwithasinglebulgecomponent(n=4).Theresidualimageshowsthewell-knownjetofthissource(3C273),aswellassomediffuse,extendedemission.Thebox-likeimprintintheresidualimageresultsfromthePSFimagebeingsmallerthanthescienceimage.

PG1302−102(Fig.37)—Thisobjectwasobservedintwodifferentfilters(F606WandF702W),butbothimagesaresaturatedinthecore.Theimagecontainstwocompactsourcessuperposedonthemainhost,whichwefitsimultaneously.Thehostcanbefitwithasinglebulgecomponent,althoughsignificantstructureonlargescalesremainintheresidualimage.Thefitsfrombothfiltersareingoodagreementtowithintheuncertainty.

PG1307+085(Fig.38)—Althoughthecoreoftheimageissaturated,thehostiswell-describedbyasinglebulgecomponent(n=4).

PG1309+355(Fig.39)—Thecoreoftheimageissaturated.Wefitthehostwithasinglebulgecomponent(n=4).Theresidualimageshowsevidenceofspiral-likesubstructure,butwedidnotattempttomodelit.

PG1351+0(Fig.40)—Thisisanalmostface-onsystemwithspiralarms.Thefitisdonewithaclassicalbulge(n=3.8)andanexponentialdisk.

PG1411+442(Fig.41)—Thisisanextremelydisturbedobjectthatappearstohaveanearbycompanion.Wefitthisobjectwithaclassicalbulge(n=4)andaspiraldiskwithFouriermodes.

PG1416−129(Fig.42)—Thehostiswell-describedbyasinglebulgecomponent(n=4).

PG1426+015(Fig.43)—Thisobjecthasatidaltailandasmallcompanion.Thebest-fittingmodelforthehostconsistsofapseudobulge(n=2.1)andadiskwithFouriermodes.Modelingthehostwithonlyasinglecomponentyieldsamuchbrighterbulge(14.2vs.16.08mag),buttheresidualsofthefitaresignificantlyworsethanthoseofthetwo-componentfit.

PG1444+407(Fig.44)—Thehostisfitwithabulge(n=4)andasomewhatdisturbeddiskcomponent(n=0.31,a1=0.10).Theresidualssuggestthataring-likecomponentmightbepresent.

PG1613+658(Fig.45)—Thisisahighlydisturbedobjectinamergingsystem.Thefitisambiguous.Thebest-fittingmodelforthehostconsistsofaclassicalbulge(n=4)andadisturbeddisk(n=1)withFouriermodes.Thebulgeis∼0.7magfainterthanthebest-fittingcaseifthefitisdonewithasinglebulgecomponent(n=4),buttheresidualsofthefitaresignificantlyworsethanthoseofthetwo-componentfit.

PG1617+175(Fig.46)—Althoughthebest-fittingmodelforthehostcontainsabulgeandadisk,asingle-componentmodelalsoworksreasonablywell.Inthetwo-componentfits,thebulgeluminositiesrangefrom17.7mag(n=4)to17.9mag(n=1).

PG1700+518(Fig.47)—Thehostiswell-representedbyasinglebulgecomponent(n=4)plusatidaltail,whichisfitwithFouriermodes.

PG2130+099(Fig.48)—ThisobjectisverysimilartoPG0052+251;bothhavering-likespiraldisk.Sincethecentralfewpixelsaresaturated,thefitisslightlyuncertain.Themodelforthehostconsistsofapseudobulge(n=0.44)andadisk(n=0.33)withFouriermodes.

PHL909(Fig.49)—Thehostisfitwithaclassicalbulge(n=4),buttheresidualimageshowsafaintcentralfeaturethatresemblesatinybarorhighlyinclineddisk-likestructure.Thehostgalaxyisslightlydisturbed.

22KIMetal.

PHL1093(Fig.50)—Thehostiswellfitwithasinglebulgecomponent(n=4).Threesmallnearbycompanionsareincludedinthefitsimultaneously.

PKS0736+01(Fig.51)—Thehostiswellfitwithasinglebulgecomponent(n=4).Thereareseveralfaintblobsnearby,butitisunclearwhethertheseareassociatedwiththeprimaryhost.

PKS1020−103(Fig.52)—Asinglebulgecomponent(n=4)isenoughtodescribethehost.PKS1217+02(Fig.53)—Thehostiswellfitwithasinglebulgecomponent(n=4).

PKS2135−14(Fig.54)—Wefitthehostwithasinglebulgecomponent(n=4).Twonearbyobjectsareincludedsimultaneouslyinthefit.

PKS2349−01(Fig.55)—Therearetwocurvedtidaltails.Thebest-fitmodelforthehostconsistsofaclassicalbulge(n=4)andahighlydistorteddiskwithn≈0.88.

PKS2355−082(Fig.56)—Thehostiswellfitwithasinglebulgecomponent(n=4).

BlackHoleMassvs.BulgeLuminosityRelation23

FIG.12.—GALFITdecompositionfor3C59.(a)Azimuthallyaveragedprofile,showingtheoriginaldata(opencircles),thebestfit(solidline),andthesub-components(PSFandbulge;dashedlines).Theresidualsareplottedonthebottom.Wepresentthe2-Dimageoftheoriginaldata(b),thebest-fitmodelforthehost(theAGNisexcludedtobetterhighlightthehost),withtheamplitudeofthefirstFouriermode(a1)labeled(c),andtheresiduals(d).Theunitsoftheimagesareinarcseconds.Allimagesareonanasinhstretch.

24KIMetal.

FIG.13.—GALFITdecompositionforE1821+3;symbolsandconventionsasinFigure12.

BlackHoleMassvs.BulgeLuminosityRelation25

FIG.14.—GALFITdecompositionforFairall9;symbolsandconventionsasinFigure12.

26KIMetal.

FIG.15.—GALFITdecompositionfor[HB]0316−346;symbolsandconventionsasinFigure12.

BlackHoleMassvs.BulgeLuminosityRelation27

FIG.16.—GALFITdecompositionfor[HB]2201+315(PC/F555W);symbolsandconventionsasinFigure12.

28KIMetal.

FIG.17.—GALFITdecompositionforHE0306−3301;symbolsandconventionsasinFigure12.

BlackHoleMassvs.BulgeLuminosityRelation29

FIG.18.—GALFITdecompositionforHE0354−5500;symbolsandconventionsasinFigure12.

30KIMetal.

FIG.19.—GALFITdecompositionforHE1434−1600;symbolsandconventionsasinFigure12.

BlackHoleMassvs.BulgeLuminosityRelation31

FIG.20.—GALFITdecompositionforMC1635+119;symbolsandconventionsasinFigure12.

32KIMetal.

FIG.21.—GALFITdecompositionforMRK1048;symbolsandconventionsasinFigure12.

BlackHoleMassvs.BulgeLuminosityRelation33

FIG.22.—GALFITdecompositionforMS0244.8+1928;symbolsandconventionsasinFigure12.

34KIMetal.

FIG.23.—GALFITdecompositionforMS0754.6+3928;symbolsandconventionsasinFigure12.

BlackHoleMassvs.BulgeLuminosityRelation35

FIG.24.—GALFITdecompositionforMS1059.0+7302;symbolsandconventionsasinFigure12.

36KIMetal.

FIG.25.—GALFITdecompositionforMS1545.3+0305;symbolsandconventionsasinFigure12.

BlackHoleMassvs.BulgeLuminosityRelation37

FIG.26.—GALFITdecompositionforOX169;symbolsandconventionsasinFigure12.

38KIMetal.

FIG.27.—GALFITdecompositionforPG0052+251;symbolsandconventionsasinFigure12.

BlackHoleMassvs.BulgeLuminosityRelation39

FIG.28.—GALFITdecompositionforPG0804+761;symbolsandconventionsasinFigure12.

40KIMetal.

FIG.29.—GALFITdecompositionforPG0923+201;symbolsandconventionsasinFigure12.

BlackHoleMassvs.BulgeLuminosityRelation41

FIG.30.—GALFITdecompositionforPG0953+414;symbolsandconventionsasinFigure12.

42KIMetal.

FIG.31.—GALFITdecompositionforPG1004+130;symbolsandconventionsasinFigure12.

BlackHoleMassvs.BulgeLuminosityRelation43

FIG.32.—GALFITdecompositionforPG1012+008;symbolsandconventionsasinFigure12.

44KIMetal.

FIG.33.—GALFITdecompositionforPG1116+215;symbolsandconventionsasinFigure12.

BlackHoleMassvs.BulgeLuminosityRelation45

FIG.34.—GALFITdecompositionforPG1202+281;symbolsandconventionsasinFigure12.

46KIMetal.

FIG.35.—GALFITdecompositionforPG1211+143;symbolsandconventionsasinFigure12.

BlackHoleMassvs.BulgeLuminosityRelation47

FIG.36.—GALFITdecompositionforPG1226+023;symbolsandconventionsasinFigure12.

48KIMetal.

FIG.37.—GALFITdecompositionforPG1302−102(PC/F702W);symbolsandconventionsasinFigure12.

BlackHoleMassvs.BulgeLuminosityRelation49

FIG.38.—GALFITdecompositionforPG1307+085;symbolsandconventionsasinFigure12.

50KIMetal.

FIG.39.—GALFITdecompositionforPG1309+355;symbolsandconventionsasinFigure12.

BlackHoleMassvs.BulgeLuminosityRelation51

FIG.40.—GALFITdecompositionforPG1351+0;symbolsandconventionsasinFigure12.

52KIMetal.

FIG.41.—GALFITdecompositionforPG1411+442;symbolsandconventionsasinFigure12.

BlackHoleMassvs.BulgeLuminosityRelation53

FIG.42.—GALFITdecompositionforPG1416−129;symbolsandconventionsasinFigure12.

54KIMetal.

FIG.43.—GALFITdecompositionforPG1426+015;symbolsandconventionsasinFigure12.

BlackHoleMassvs.BulgeLuminosityRelation55

FIG.44.—GALFITdecompositionforPG1444+407;symbolsandconventionsasinFigure12.

56KIMetal.

FIG.45.—GALFITdecompositionforPG1613+658;symbolsandconventionsasinFigure12.

BlackHoleMassvs.BulgeLuminosityRelation57

FIG.46.—GALFITdecompositionforPG1617+175;symbolsandconventionsasinFigure12.

58KIMetal.

FIG.47.—GALFITdecompositionforPG1700+518;symbolsandconventionsasinFigure12.

BlackHoleMassvs.BulgeLuminosityRelation59

FIG.48.—ExampleofGALFITdecompositionforPG2130+099;symbolsandconventionsasinFigure12.

60KIMetal.

FIG.49.—GALFITdecompositionforPHL909;symbolsandconventionsasinFigure12.

BlackHoleMassvs.BulgeLuminosityRelation61

FIG.50.—GALFITdecompositionforPHL1093;symbolsandconventionsasinFigure12.

62KIMetal.

FIG.51.—GALFITdecompositionforPKS0736+01;symbolsandconventionsasinFigure12.

BlackHoleMassvs.BulgeLuminosityRelation63

FIG.52.—GALFITdecompositionforPKS1020−103;symbolsandconventionsasinFigure12.

KIMetal.

FIG.53.—GALFITdecompositionforPKS1217+02;symbolsandconventionsasinFigure12.

BlackHoleMassvs.BulgeLuminosityRelation65

FIG.54.—GALFITdecompositionforPKS2135−14;symbolsandconventionsasinFigure12.

66KIMetal.

FIG.55.—GALFITdecompositionforPKS2349−01;symbolsandconventionsasinFigure12.

BlackHoleMassvs.BulgeLuminosityRelation67

FIG.56.—GALFITdecompositionforPKS2355−082;symbolsandconventionsasinFigure12.