A Search for "Dwarf" Seyfert Nuclei. VII. A Catalog of Central Stellar Velocity Dispersions of Nearby Galaxies
Luis C. Ho, Jenny E. Greene, Alexei V. Filippenko, Wallace L. W. Sargent
aa r X i v : . [ a s t r o - ph . GA ] J un T O APPEAR IN
The Astrophysical Journal Supplement Series .Preprint typeset using L A TEX style emulateapj v. 26/01/00
A SEARCH FOR “DWARF” SEYFERT NUCLEI. VII. A CATALOG OF CENTRAL STELLAR VELOCITYDISPERSIONS OF NEARBY GALAXIES L UIS
C. H O The Observatories of the Carnegie Institution of Washington, 813 Santa Barbara St., Pasadena, CA 91101 J ENNY
E. G
REENE Department of Astrophysical Sciences, Princeton University, Princeton, NJ A LEXEI
V. F
ILIPPENKO
Department of Astronomy, University of California, Berkeley, CA 94720-3411AND W ALLACE
L. W. S
ARGENT
Palomar Observatory, California Institute of Technology, MS 105-24, Pasadena, CA 91125
To appear in The Astrophysical Journal Supplement Series.
ABSTRACTWe present new central stellar velocity dispersion measurements for 428 galaxies in the Palomar spectroscopicsurvey of bright, northern galaxies. Of these, 142 have no previously published measurements, most being rela-tively late-type systems with low velocity dispersions ( ∼ <
100 km s - ). We provide updates to a number of literaturedispersions with large uncertainties. Our measurements are based on a direct pixel-fitting technique that can ac-commodate composite stellar populations by calculating an optimal linear combination of input stellar templates.The original Palomar survey data were taken under conditions that are not ideally suited for deriving stellar veloc-ity dispersions for galaxies with a wide range of Hubble types. We describe an effective strategy to circumvent thiscomplication and demonstrate that we can still obtain reliable velocity dispersions for this sample of well-studiednearby galaxies. Subject headings: galaxies: active — galaxies: kinematics and dynamics — galaxies: nuclei — galaxies: Seyfert— galaxies: starburst — surveys INTRODUCTION
The stellar velocity dispersion ( σ ⋆ ) of the central regions ofgalaxies is a parameter of considerable importance for a varietyof extragalactic investigations. Since the early pioneering workof Burbidge et al. (1961) and Minkowski (1962), many tech-niques have been developed for measuring σ ⋆ (e.g., Morton &Chevalier 1972; Richstone & Sargent 1972; Simkin 1974; Sar-gent et al. 1977; Tonry & Davis 1979; Bender 1990; Rix &White 1992; van der Marel & Franx 1993; Cappellari & Em-sellem 2004; Statler 1995; Barth et al. 2002). Given the exten-sive body of observational material on σ ⋆ for nearby galaxies,a number of catalogs have been compiled to consolidate thedata. The most widely used of these are the catalog of Whit-more et al. (1985), which was updated by McElroy (1995), andof Prugniel et al. (1998), which is continuously updated and isavailable through the electronic database HyperLeda (Paturel etal. 2003) .The vast majority of the published measurements of σ ⋆ per-tain to early-type galaxies, largely giant ellipticals and S0s. Sig-nificantly less data are available for galaxies along the spiral se-quence, and those that have been published often show markeddisagreement from study to study, as can be seen from perusalof the data tabulated in the above-mentioned catalogs. It is dis-concerting that many of the highly discrepant entries are, infact, associated with nearby, bright, well-studied galaxies. Thescatter in the published values of σ ⋆ can be blamed, at leastin part, on the inherent heterogeneity of combining many dis-parate sources, which often employ different telescopes, detec- tors, apertures, observing strategies, and analysis techniques.The above-cited catalogs attempt to homogenize the final com-pilations by scaling the individual literature sources to a set of“standard” galaxies measured through a roughly constant aper-ture size (2 ′′ × ′′ ).Notwithstanding these efforts, there is considerable motiva-tion for assembling an independent, homogeneous, internallyconsistent set of new measurements, especially if the data covera large sample of galaxies representing a wide range of Hubbletypes. A number of previous studies have been carried out withthis goal in mind, mostly focused on relatively early-type galax-ies (e.g., Davies et al. 1987; Bernardi et al. 2003). Our presentpaper adds to this effort using data taken as part of the Palo-mar spectroscopic survey of nearby galaxies. During the courseof an extensive investigation primarily aimed at characterizingthe nature of nuclear activity in nearby galaxies, we collectedhigh-quality, moderate-resolution, long-slit optical spectra ofthe central regions of 486 bright, northern galaxies. The sur-vey was conducted during the period 1984–1990; technical de-tails of the survey and presentation of various data products andscience results can be found in earlier papers in this series (Fil-ippenko & Sargent 1985; Ho et al. 1995, 1997a–1997e, 2003).This contribution focuses on central stellar velocity dispersionsextracted from the survey. THE SURVEY
A full description of the Palomar survey is given by Ho et al.(1995, 1997a). Here we mention only a few pertinent details.The survey covers a nearly complete, magnitude-limited Hubble Fellow, Princeton-Carnegie Fellow. http://leda.univ-lyon1.fr/ F IG . 1.— Sample blue ( left ) and red ( right ) spectra from the Palomar survey, adapted from Ho et al. (1995). The intensity of each spectrum has been scaled andarbitrarily shifted for clarity. The regions included in the fit are plotted in blue, while those masked from the fit are plotted as red dotted lines. Black dashed linesdenote regions outside of the fitting window. The stellar metal-line indices defined by Ho et al. (1997a) are labeled on the bottom of each panel. sample of 486 galaxies from the Revised Shapley-Ames cata-log (Sandage & Tammann 1981) that satisfy B T ≤ . δ > ◦ . The spectra were acquired using the Double Spec-trograph (Oke & Gunn 1982) mounted at the Cassegrain fo-cus of the Hale 5-m telescope at Palomar Observatory. A 2 ′′ -wide slit was used for most of the survey. The spectra simul-taneously cover the regions ∼ ∼ σ inst = 118 and42 km s - at 4500 Å and 6500 Å, respectively. (About 10%of the blue spectra were acquired in a slightly higher resolutionmode with σ inst = 74 km s - .) The spectra analyzed in this pa-per are the same as those reported in the spectral atlas of Ho etal. (1995); they were extracted from a rectangular aperture ofsize 2 ′′ × ′′ , which is roughly equivalent to linear dimensionsof 170 pc ×
350 pc for a median distance of 17.9 Mpc (Ho etal. 1997a). VELOCITY DISPERSIONS
Method
Our velocity dispersion measurements are based on the directpixel-fitting method, which, as described by a number of au-thors (e.g., Rix & White 1992; van der Marel 1994; Kelson et al.2000; Barth et al. 2002), has many of advantages compared tomore traditional methods based on Fourier or cross-correlationtechniques. The Palomar survey has several characteristics thatpose special challenges for measuring accurate stellar veloc-ity dispersions. First, the majority of the survey galaxies con-tain emission lines from active galactic nuclei (AGNs), oftenstrong and of substantial velocity width, presenting a signifi- cant source of contamination for the stellar absorption features.Second, the spectral coverage of the survey was optimized forobtaining emission-line diagnostics and not for velocity disper-sion measurements. Finally, the survey covers a very broadrange of Hubble types, from dwarf irregulars to giant ellipticals.Galaxies with a wide range of stellar populations are especiallysusceptible to template mismatch. We use a modified version ofthe direct pixel-fitting code developed by Greene & Ho (2006).In brief, a nonlinear Levenberg-Marquardt minimization algo-rithm is used to compare the observed galaxy spectrum with amodel spectrum M ( λ ), which is assumed to be the convolutionof a stellar template spectrum, T ( λ ), and a line-of-sight velocitybroadening function approximated as a Gaussian, G ( λ ): M ( λ ) = P ( λ ) { [ T ( λ ) ⊗ G ( λ )] + C ( λ ) } . (1)Here, C ( λ ) is an additive term to dilute the stellar features. Itcan be a power-law function to represent an AGN continuum, ifpresent, or any other smooth component such as the featurelesscontinuum from hot stars. For many of our later-type galax-ies, adding a simple f λ = constant term effectively mimics thecontinuum dilution of the metal lines by intermediate-age (Aand early-F type) stars in the composite stellar population. Themultiplicative factor P ( λ ), typically chosen to be a third-orderLegendre polynomial, accounts for large-scale mismatches inthe continuum shapes of the galaxy and template star(s), whichcan arise from internal reddening in the galaxy, stellar popula-tion differences, and possible residual calibration errors.An important improvement over the original code of Greene& Ho is that T ( λ ), rather than being a single star, can be anoptimal linear combination of several stars determined througha nonlinear least-squares fit. In the case of later-type spirals,especially, this modification provides a much better fit for theircomposite stellar populations, as well as a more robustATALOG OF STELLAR VELOCITY DISPERSIONS 3 F IG . 2.— Sample fits for a representative set of galaxies. The top spectrum is that of the K0 III star HD 107328 from the Valdes et al. (2004) stellar library.For each galaxy, the original data are plotted as black histograms. The best-fitting model constructed from an optimal combination of broadened stellar templates isplotted as a thin blue curve. The regions excluded from the fit are marked as red dotted lines. The intensity of each spectrum has been scaled and arbitrarily shiftedfor clarity. determination of the final velocity dispersion of the galaxy be-cause the intrinsic widths of the template stars vary with spec-tral type. Our approach of using a mixture of template stars issimilar to those employed by several previous studies, includ-ing Rix & White (1992) and Cappellari & Emsellem (2004).3.2. Fitting Regions
The blue setup just misses Mg I λ b ”), the featuremost commonly used to derive velocity dispersions in the visi-ble part of the spectrum. Nevertheless, the blue spectra containa significant number of relatively strong metal-line features, in-cluding the G band at 4300 Å, a calcium feature at 4455 Å, andiron features at 4383, 4531, and 4668 Å (Fig. 1, left ; see Table 7in Ho et al. 1997a for definitions of these stellar absorption-lineindices). These metal-line features can be used to derive stellarvelocity dispersions, so long as they are strong enough in the in-tegrated spectrum. For the blue spectra we fit the region 4260–4950 Å; the blue end is chosen to include the G band, whilethe red end avoids the [O III ] λλ γ (4320–4370 Å) and H β (4830–4890 Å). In some strong emission-line objects, it is necessaryto mask a small region around He II λ σ inst ≈
120 km s - . For example, HO ET AL. F IG . 3.— Sample fits for NGC 3368 and NGC 4303 in the blue spectral region. The top spectrum shows the optimally weighted fit, followed by fits using singlestars of spectral type F6 III, G8 III, K0 III, and K3 III. The original data are plotted as black histograms, the fits are plotted as blue curves, and the regions excludedfrom the fit are plotted as red dotted lines. The intensity of each spectrum has been scaled and arbitrarily shifted for clarity. according to the literature NGC 221 and NGC 3489 have σ = 72and 112 km s - , respectively.The red spectra, with σ inst ≈
40 km s - , provide crucial reliefto the many galaxies in the survey that suffer from insufficientresolution in the blue. Unfortunately, very few strong, uncon-taminated stellar features exist in the spectral coverage of ourred setup, which is dominated almost entirely by strong emis-sion lines ([O I ] λλ II ] λλ α , and[S II ] λλ right ) illustrates, the Ca+Fe feature, lyingjust blueward of the H α +[N II ] complex, is fairly well iso-lated, even in objects with prominent, broad H α emission (e.g.,NGC 3031). Importantly, it is moderately strong in nearly allgalaxies, even those whose blue spectra are hopeless dilutedby A and F-type stars (e.g., NGC 3073). Using the measure-ments published by Ho et al. (1997a, Table 9), we find thatCa+Fe was reliably detected in 438 out of the 486 galaxies inthe Palomar survey (90%), with an average equivalent width of h W(Ca + Fe) i = 0 . - < T <
0; de Vaucouleurs et al. 1991), h W(Ca + Fe) i = 1 . h W(Ca + Fe) i = 0 . < T < right ).The blue limit provides as much leverage as possible to definethe continuum level without colliding with [O I ] λ II ] λ α emission, we had to curtail the red limit to 6510 Å; inthese cases, it was often also helpful to increase the order of thepolynomial factor (to ∼ -
6) to better trace the steeply risinggradient of the blue wing of the H α emission line.3.3. Template Stars
In addition to spectrophotometric standard stars, during thecourse of the survey we usually also took nightly observationsof at least one late-type giant star to be used as a velocity tem-plate. Velocity standards were not observed in a small numberof observing runs; this affected 50 galaxies, or roughly 10% ofthe survey. Because measuring velocity dispersions was not atop priority for the original survey, neither the number of starsnor their range of spectral types was chosen optimally. In someof the runs, only a single velocity template was observed, andat most there were two.The limitations of the Palomar template stars compel us toexplore an alternative calibration strategy. We use as our pri-mary source of templates the library of Coudé-feed stellar spec-tra published by Valdes et al. (2004). This tremendously usefuldatabase contains high-S/N spectra of 1273 stars of essentiallyall spectral types, covering 3460 to 9464 Å. The spectral reso-lution of the library, FWHM ≈ F IG . 4.— Sample fits for NGC 3631 and NGC 3115 in the red spectral region. The top spectrum shows the optimally weighted fit, followed by fits using singlestars of spectral type F6 III, G8 III, K0 III, and K3 III. The original data are plotted as black histograms, and the fits are plotted as blue curves. The intensity of eachspectrum has been scaled and arbitrarily shifted for clarity. between the two data sets.The Valdes library also gives us an extensive selection ofstars of different spectral types for our optimal fit. Throughexperimentation, we find that in general a set of four stars—spectral types F6 III, G8 III, K0 III, and K3 III—suffices toaccount for the stellar population mixture of almost all galaxiesin our sample. We give preference to stars of near-solar metal-licity to try to approximate the conditions in galactic bulges.Although type-A and early-F stars clearly exist in some galax-ies, in practice they do not need to be included because ourfitting regions deliberately avoid the Balmer absorption lines(Fig. 1) and the continuum dilution term [ C ( λ ) in Equation 1]effectively mimics the hot continuum of these stars.3.4. Fitting Results
Figure 2 gives examples of some typical fits. The top spec-trum is that of the red giant (K0 III) star HD 107328, shown tohelp guide the eye to identify the stellar features. Subsequentspectra illustrate galaxies with a wide range in emission-linestrengths and velocity dispersions. The original galaxy spec-trum is plotted as black histograms; the best-fitting, optimallyweighted, broadened velocity template is plotted as a thin blueline; and the masked regions are plotted as a red dotted line.Using a set of just four stars, we can usually achieve quite goodfits, with formal statistical errors on the velocity dispersions inthe range of 5%–10%. The results are also quite robust withrespect to the choice of template stars; interchanging differentstars of the same spectral type and similar metallicity affectsthe final dispersions at the level of 1% or less. In most objects,the largest fraction of the light comes, not surprisingly, fromK giants. The Fe λ λ σ c = 114 . ± . - (68 . ± . - for the higher resolution mode), whilethat for the red side is σ c = 37 . ± . - , where the er-ror bar represents the root-mean square (rms) scatter of thenight-to-night variations of the Palomar instrumental resolu-tion. The validity of this simple approach can be verified em-pirically by comparing the corrected dispersions with publishedvalues. Among the 223 galaxies with velocity dispersions de-rived in the blue, 189 have literature measurements; of the422 dispersions measured in the red, 283 have literature val-ues. As illustrated in Figure 5, the adopted resolution correc-tions yield reasonably satisfactory agreement between our dis-persion measurements and the literature values, particularly inthe regime when the dispersions are well resolved ( σ ∼ > σ inst ; solid points ). On the blue side (Fig. 5 a ), for σ ∼ > σ inst ≈ - , h σ blue - σ Literature i = 1 . - with an rms scatter of25.3 km s - . The red side delivers useful measurements downto σ ≈ σ inst ≈
20 km s - (Fig. 5 b ). Over the entire velocityrange, h σ red - σ Literature i = 3 . - with an rms scatter of 28.3km s - . There is no perceptible systematic bias, provided thatthe optimal fit excludes the K3 III star, as explained below.Our initial fits for the red-side spectra, which include the full HO ET AL. F IG . 5.— Comparison between velocity dispersions published in the literature with velocity dispersions derived using an optimal combination of Valdes templatestars for the ( a ) blue side and ( b ) red side, corrected for the relative resolution difference ( σ c ) between the Valdes and Palomar systems (see §3.3). Open symbolsmark objects that are poorly resolved. The dashed diagonal line denotes equality. complement of four template stars (F6 III to K3 III), revealeda puzzling systematic trend. Whereas the fits for low- σ galax-ies yield dispersions that, after resolution correction, agree rea-sonably well with literature values, objects with σ ∼ > - show a net systematic offset toward larger velocities, byroughly +
30 km s - . We believe that this effect arises from tem-plate mismatch. As shown in Figure 4, in small, low-luminositybulges, such as that in the Sc galaxy NGC 3631, the red absorp-tion features, especially Ca+Fe, are nearly equally well fit bytemplate stars of spectral type G8 III, K0 III, or K3 III. In starkcontrast, NGC 3115, a luminous S0 galaxy with a substantialbulge, clearly singles out the K3 III star as the preferred tem-plate, which then contributes most of the weight to the optimalfit. (We have verified that K1 III and K2 III templates givealmost equally good fits as the K3 III template.) Why? Thisis because the Ca+Fe feature is strongest in high- σ galaxiesand in late-type giants. Within the Palomar galaxy sample, thestrength of the Ca+Fe feature increases roughly with velocitydispersion, albeit with significant scatter. Dividing the sampleinto two, galaxies with σ <
150 km s - have h W(Ca + Fe) i =0 .
87 Å, to be compared with h W(Ca + Fe) i = 1 .
23 Å for galax-ies with σ ≥
150 km s - . At the same time, the strength of theCa+Fe feature in stars increases toward later spectral types. Todemonstrate this, we measured the Ca+Fe feature for individualstars in the Valdes library, using the index definition given in Hoet al. (1997a). Choosing 15 stars of roughly similar metallic-ities for each spectral type, we find h W(Ca + Fe) i = 0 .
79, 0.99,1.17, 1.35, and 1.53 Å for G8 III, K0 III, K1 III, K2 III, andK3 III, respectively. Galaxies with σ ≥
150 km s - have Ca+Festrengths very similar to those of K1 III–K3 III stars, and thusit is not surprising that an optimal fit would give these starsgreatest weight. A bias in the derived velocity dispersion forhigh- σ galaxies arises if in these systems their Ca+Fe feature is boosted because of an abundance enhancement. We speculatethat the culprit is Ca. As an α element, Ca may be enhancedsimilarly as Mg in early-type galaxies (Prochaska et al. 2005;but see Graves et al. 2007). In such a situation, the apparentlygood match with the K1 III–K3 III templates is only an artifactof their mutually strong Ca+Fe feature. Since such late-type gi-ants have very narrow intrinsic line widths, the inferred velocitydispersion would be overestimated, thus leading to the observedbias. To bypass this complication, we removed the K3 III giantfrom the optimal fit of the red-side spectra.For each galaxy, we compute a final velocity dispersion asthe average of the blue-side and red-side dispersions, weightedby their respective error bars. The error bars reflect the quadra-ture sum of the formal statistical uncertainty from the optimal fitand the rms scatter of the resolution correction, which is domi-nated by the uncertainty in the original instrumental resolutionof the Palomar spectra. Among the 428 galaxies with new ve-locity dispersion measurements, 286 have published literaturevalues. Comparison between the objects in common (Fig. 6)show very good consistency. Over the entire range in veloci-ties, h σ final - σ Literature i = 3 . - . The scatter is still quitelarge (rms 28.3 km s - ), but its magnitude is consistent withthat found by Barth et al. (2002) based on a smaller sample of ∼
30 galaxies with high-quality velocity dispersion measure-ments.There are several notable outliers in Figure 6, for which theliterature values are larger than ours by more than ∼
80 km s - .The most extreme case is NGC 520, for which HyperLeda re-ports σ = 240 ±
25 km s - whereas we determine σ = 40 . ± . - . This is a complex, interacting galaxy (Arp 157), and theHyperLeda value of σ = 240 km s - pertains to the “southeast-northwest” component, not the primary nucleus of the “east-west” component (using the naming convention of Stanford &ATALOG OF STELLAR VELOCITY DISPERSIONS 7 F IG . 6.— Comparison of final velocity dispersions with literature values.The dashed diagonal line denotes equality. Several prominent outliers are la-beled (see §3.4). Balcells 1990a). The Palomar spectrum was centered on theposition of the primary nucleus. From visual inspection ofthe plots published by Stanford & Balcells (1990a, 1990b), itappears that the published dispersion of the primary nucleusshould be σ ≈ ±
25 km s - . (We thank the referee for mak-ing this estimate for us, with which we agree.) The velocitydispersion for NGC 2967 (prior to homogenization), 200 ± - , seems suspiciously high for an Sc galaxy; according toHyperLeda, it derives from an unpublished measurement by B.C. Whitmore & E. Malumuth (1984). The same applies to theSc galaxy NGC 4647, for which Hyperleda lists σ = 98 ±
39 kms - . Finally, we note that the literature values of both NGC 3628( σ = 171 ±
71 km s - ) and NGC 5364 ( σ = 91 ±
52 km s - ) haveexceptionally large error bars. If we exclude these five outliersfrom our sample, h σ final - σ Literature i = 2 . - , and the scatterreduces to 23.6 km s - . THE CATALOG
The final results are presented in Table 1. For each galaxy,we list the literature value of the central stellar velocity dis-persion, if available, followed by the dispersions derived fromthe blue ( σ blue ) and red ( σ red ) Palomar spectra, the final value( σ final ) obtained from the weighted average of σ blue and σ red ,and lastly the adopted value. Most of the literature values comefrom the HyperLeda database (Paturel et al. 2003), which, forany given galaxy, attempts to homogenize all published mea-surements into a single value by applying scaling factors de-termined from a set of “standard” galaxies measured througha roughly constant aperture size of 2 ′′ × ′′ . This aperture size,fortunately, exactly matches that employed in the Palomar sur-vey.For the final, adopted dispersion, there are strong reasons toprefer the Palomar measurements because of their homogene-ity. Although in many cases their error bars formally exceedthose of the literature sources, we believe that the error budgetfor the Palomar measurements is realistic, as evidenced, for ex-ample, from comparison with the high-accuracy measurementsfrom Barth et al. (2002) for galaxies in common. Nevertheless, for concreteness, the final column of Table 1 lists either the finalPalomar dispersion or the literature value, if available, based onwhichever has the smaller formal error bar.In total, our catalog gives new stellar velocity dispersionmeasurements for 428 galaxies, 88% of the parent survey. Ofthese, 142 (30%) have no previously published measurements.Not surprisingly, most of the new measurements are for late-type galaxies, systems where velocity dispersions are morechallenging to obtain because of their characteristically lowervalues ( ∼ <
100 km s - ) and complications due to their compositestellar populations and contamination by emission lines. Ournew measurements also provide updates to a number of liter-ature dispersions that previously had large uncertainties or, insome instances, were grossly in error.Stellar velocity dispersions could not be derived for 58 galax-ies, mostly because their stellar features are too weak. For thesake of completeness, for the 34 of these objects that have emis-sion lines, and for which no reliable dispersions exist in theliterature, we list an indirect estimate of their stellar velocitydispersion based on their observed gaseous velocity dispersionderived from the line profile of [N II ] λ σ g ≈ (0 . - . σ ∗ . In de-tail, the normalization of the σ g - σ ∗ relation shows a slight de-pendence on nuclear (H α ) luminosity and Eddington ratio, but only for sources spectroscopically classified as AGNs (LINERs,transition objects, and Seyferts). Those classified as H II (star-forming) nuclei obey σ g = 0 . σ ∗ with an rms scatter 0.19 dex.This is the relation that we use because all of the 34 emission-line sources with very weak stellar features are H II nuclei (Hoet al. 1997a) . The error bars in the adopted dispersions comefrom the quadrature sum of the uncertainties in the original[N II ] line widths (we conservatively assume 10%; Ho et al.1997a) and the 0.19 dex scatter in the σ g - σ ∗ relation. SUMMARY
The Palomar spectroscopic survey has furnished consider-able insights into the nature of nuclear activity in nearby galax-ies (see Ho 2008 for a review). Aside from some considerationsof the central stellar populations (Ho et al. 2003; Zhang et al.2008), however, comparatively little analysis has been done onthe absorption-line component of the spectra. This paper uti-lizes the survey spectra to derive a homogeneous set of newcentral stellar velocity dispersion measurements. A major ob-stacle is that the original survey data were not taken with thisapplication in mind. In particular, neither the number nor therange of calibration template stars is ideally suited for deriv-ing stellar velocity dispersions for galaxies with a wide rangeof composite stellar populations. The wavelength coverage ofthe blue-side and red-side spectra is nonstandard for velocitydispersion work and is rather sensitive to template mismatch.Moreover, the spectral resolution of the blue-side spectra is toocoarse to yield reliable dispersions for most of the later-typegalaxies in the sample.We describe an effective strategy to address these challenges.We use the extensive Coudé-feed spectral library of Valdes et al.(2004) as the primary source of stellar templates. Applying asimple correction for the nominal relative resolution differencebetween the Valdes and Palomar systems yields velocity dis-persions that show reasonably good agreement with literature In detail, Ho (2009) notes that σ g / σ ∗ for H II nuclei depends on σ ∗ , but for our present purposes we neglect this complication. HO ET AL.values. The direct-pixel fitting code of Greene & Ho (2006)was adapted to solve for an optimally weighted linear combi-nation of template stars, a crucial step to match the compos-ite stellar population typically found in later-type galaxies. Wedemonstrate that the Ca+Fe λ http://leda.univ-lyon1.fr ). The researchof L.C.H. was supported by the Carnegie Institution of Wash-ington. Support for J.E.G. was provided by NASA throughHubble Fellowship grant HF–01196, awarded by the SpaceTelescope Science Institute, which is operated by the Associa-tion of Universities for Research in Astronomy, Inc., for NASA,under contract NAS 5–26555. A.V.F. and W.L.W.S. are grate-ful for the financial support of the NSF, through grants AST–0607485 and AST–0606868, respectively. REFERENCESBarth, A. J., Ho, L. C., & Sargent, W. L. W. 2002, AJ, 124, 2607Batcheldor, D., et al. 2005, ApJS, 160, 76Bender, R. 1990, A&A, 229, 441Bernardi, M., et al. 2003, AJ, 125, 1817Burbidge, E. M., Burbidge, G. R., & Fish, R. A. 1961, ApJ, 133, 393Cappellari, M., & Emsellem, E. 2004, PASP, 116, 138Corsini, E. M., Pizzella, A., Coccato, L., & Bertola, F. 2003, A&A, 408, 873Davies, R. L., Burstein, D., Dressler, A., Faber, S. M., Lynden-Bell, D.,Terlevich, R. J., & Wegner, G. 1987, ApJS, 64, 581de Vaucouleurs, G., de Vaucouleurs, A., Corwin Jr., H. G., Buta, R. J., Paturel,G., & Fouqué, R. 1991, Third Reference Catalogue of Bright Galaxies (NewYork: Springer)Filippenko, A. V., & Ho, L. C. 2003, ApJ, 588, L13Filippenko, A. V., Matheson, T, & Barth, A. J. 1995, ApJ, 455, L139Filippenko, A. V., Matheson, T., Leonard, D. C., Barth, A. J., & Van Dyk, S. D.1997, PASP, 109, 461Filippenko, A. V., & Sargent, W. L. W. 1985, ApJS, 57, 503Funes, J. G., Corsini, E. M., Cappellari, M., Pizzella, A., Vega Beltrán, J. C., &Bertola, F. 2002, A&A, 388, 50Ganda, K., Falcón-Barroso, J., Peletier, R. F., Cappellari, M., Emsellem, E.,McDermid, R. M., de Zeeuw, P. T., & Carollo, C. M. 2006, MNRAS, 367,46Graves, G. J., Faber, S. M., Schiavon, R. P., & Yan, R. 2007, ApJ, 671, 243Greene, J. E., & Ho, L. C. 2006, ApJ, 641, 117Héraudeau, Ph., & Simien, F. 1998, A&AS, 133, 317Ho, L. C. 2008, ARA&A, 46, 475——. 2009, ApJ, in pressHo, L. C., Filippenko, A. V., & Sargent, W. L. W. 1995, ApJS, 98, 477——. 1997a, ApJS, 112, 315——. 1997b, ApJ, 487, 568——. 1997c, ApJ, 487, 579——. 1997d, ApJ, 487, 591——. 2003, ApJ, 583, 159Ho, L. C., Filippenko, A. V., Sargent, W. L. W., & Peng, C. Y. 1997e, ApJS,112, 391 Kelson, D. D., Illingworth, G. D., van Dokkum, P. G., & Franx, M. 2000, ApJ,531, 159Kormendy, J., & McClure, R. D. 1993, AJ, 105, 1793Marconi, A., et al. 2003, ApJ, 586, 868McElroy, D. B. 1995, ApJS, 100, 105Minkowski, R. 1962, in IAU Symp. 15, Problems of Extra-Galactic Research,ed. G. C. McVittie (New York: Macmillan), 112Morton, D. C., & Chevalier, R. 1972, ApJ, 174, 489Nelson, C. H., Green, R. F., Bower, G., Gebhardt, K., & Weistrop, D. 2004,ApJ, 615, 652Oke, J. B., & Gunn, J. E. 1982, PASP, 94, 586Paturel, G., Petit, C., Prugniel, Ph., Theureau, G., Rousseau, J., Brouty, M.,Dubois, P., & Cambrésy, L. 2003, A&A, 412, 45Prochaska, L. C., Rose, J. A., & Schiavon, R. P. 2005, AJ, 130, 2666Prugniel, Ph., Zasov, A., Busarello, G., & Simien, F. 1998, A&AS, 127, 117Richstone, D. O., & Sargent, W. L. W. 1972, ApJ, 176, 91Rix, H.-W., & White, S. D. M. 1992, MNRAS, 254, 389Sandage, A. R., & Tammann, G. A. 1981, A Revised Shapley-Ames Catalog ofBright Galaxies (Washington, DC: Carnegie Inst. of Washington)Sargent, W. L. W., Schechter, P. L., Boksenberg, A., & Shortridge, K. 1977,ApJ, 212, 326Sarzi, M., et al. 2002, ApJ, 567, 237Simkin, S. M. 1974, A&A, 31, 129Stanford, S. A., & Balcells, M. 1990a, in IAU Colloq. 124, Paired andInteracting Galaxies, ed. J. W. Sulentic, W. C. Keel & C. M. Telesco(Washington, D. C.: NASA Conference Publications No. 3098), 347——. 1990b, ApJ, 355, 59Statler, T. S. 1995, AJ, 109, 1371Tonry, J., & Davis, M. 1979, AJ, 84, 1511Valdes, F., Gupta, R., Rose, J. A., Singh, H. P., & Bell, D. J. 2004, ApJS, 152,251van der Marel, R. P. 1994, ApJ, 270, 271van der Marel, R. P., & Franx, M. 1993, ApJ, 407, 525Wegner, G., et al. 2003, AJ, 126, 2268Whitmore, B. C., McElroy, D. B., & Tonry, J. L. 1985, ApJS, 59, 1Zhang, Y., Gu, Q.-S., & Ho, L. C. 2008, A&A, 487, 177
ATALOG OF STELLAR VELOCITY DISPERSIONS 9
TABLE 1CATALOG OF STELLAR VELOCITY DISPERSIONS
Galaxy Literature Palomar Adopted σ Reference σ blue σ red σ final Notes σ ( km s − ) ( km s − ) ( km s − ) ( km s − ) ( km s − )IC 10 · · · · · · · · · · · · · · · ± · · · · · · ± · · · · · · ± ± · · · · · · · · · ± ± ± ± ± ± · · · · · · · · · ± ± ± · · · · · · ± ± ± ± · · · · · · · · · ± ± ± · · · · · · · · · · · · · · · ± · · · · · · · · · ± ± ± ± · · · · · · · · · ± ± · · · · · · · · · ± ± · · · · · · · · · ± ± · · · ± ± ± ± ± ± ± ± · · · · · · ± ± ± ± · · · · · · · · · ± ± ± ± ± ± ± ± ± · · · · · · · · · ± ± ± ± ± ± · · · · · · · · · ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± · · · · · · · · · ± ± ± ±
25 1 · · · ± ± ± · · · · · · ± ± ± ± ± ± ± ± ± ± · · · ± ± ± ± · · · ± ± ± ± · · · ± ± ± · · · · · · · · · · · · · · · < · · · · · · ± ± ± ± · · · · · · · · · ± ± ± ±
20 1 · · · ± ± ± ± · · · ± ± ± ± ± ± ± ± · · · · · · · · · ± ± ± · · · · · · · · · · · · · · · ± · · · · · · · · · ± ± ± · · · · · · ± ± ± ± ± ± ± ± ± · · · · · · ± ± ± ± ± · · · ± ± ± · · · · · · · · · ± ± ± ± ± ± ± ± ± · · · · · · · · · ± · · · · · · · · · · · · · · · < · · · · · · · · · · · · · · · ± · · · · · · · · · ± ± ± TABLE 1—
Continued
Galaxy Literature Palomar Adopted σ Reference σ blue σ red σ final Notes σ ( km s − ) ( km s − ) ( km s − ) ( km s − ) ( km s − )NGC 1003 · · · · · · · · · · · · · · · · · · NGC 1023 204.5 ± ± ± ± ± ± ± ± ± ± ±
15 1 · · · · · · · · · ± ± · · · ± ± ± ± ± ± ± ± · · · · · · · · · ± ± ± · · · · · · · · · · · · · · · ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± · · · ± ± ± ± · · · · · · · · · ± ± ± ± ± ± · · · · · · · · · · · · · · · ± · · · · · · · · · · · · · · · ± ± ± ± ± ± ± ± ± ± ± ± · · · ± ± ± · · · · · · ± ± ± ± ± ± ± ± ± · · · · · · · · · ± ± ± ± ± ± ± ± ± ± ± ± ± · · · · · · · · · ± ± ± · · · · · · · · · ± ± ± · · · · · · · · · · · · · · · · · · NGC 2403 · · · · · · · · · · · · · · · ± · · · · · · · · · · · · · · · ± · · · · · · · · · ± ± ± ±
10 1 · · · · · · · · · ± · · · · · · ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± · · · · · · · · · ± ± ± ± · · · · · · · · · ± ± · · · ± ± ± · · · · · · · · · ± ± ± ± ± ± ± ± · · · · · · · · · ± ± ± ± ± ± ± ± ± · · · ± ± ± ± ± ± ± ± ± ± ± ± ± ATALOG OF STELLAR VELOCITY DISPERSIONS 11
TABLE 1—
Continued
Galaxy Literature Palomar Adopted σ Reference σ blue σ red σ final Notes σ ( km s − ) ( km s − ) ( km s − ) ( km s − ) ( km s − )NGC 2832 334.0 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± · · · ± ± ± ± ± ± ± ± ± ± ± ± ± ±
12 6 · · · ± ± ± ± · · · ± ± ± · · · · · · · · · · · · · · · ± · · · · · · · · · ± ± ± ± ± ± ± ± · · · · · · · · · · · · · · · ± · · · · · · · · · ± ± ± ± ± ± ± ± ± · · · · · · · · · ± ±
30 1 · · · ± ± ± · · · · · · · · · ± ± ± ± · · · ± ± ± · · · · · · · · · · · · · · · ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± · · · · · · · · · ± ± ± ± · · · ± ± ± ±
11 7 167.3 ± ± ± ± ± ± ± ± ± ±
10 1 · · · ± ± ± ± ± ± ± ± ± · · · · · · · · · ± ± ± ± ± ± ± ± ± ± ± ± · · · ± ± ± ± ± ± ± ± ± · · · ± ± ± · · · · · · · · · ± ± ± ± · · · ± ± ± · · · · · · · · · ± ± ± ± · · · ± ± ± ± · · · ± ± ± ± · · · ± ± ± · · · · · · · · · ± ± ± · · · · · · · · · ± ± ± ± · · · ± ± ± ± · · · ± ± ± ± ± ± ± ± ± ± ± ± ± TABLE 1—
Continued
Galaxy Literature Palomar Adopted σ Reference σ blue σ red σ final Notes σ ( km s − ) ( km s − ) ( km s − ) ( km s − ) ( km s − )NGC 3384 148.4 ± ± ± ± ± · · · · · · ± · · · ± ± · · · · · · ± · · · ± ± ± · · · ± ± ± ± ± ± ± ± ± · · · ± ± ± · · · · · · · · · ± ± ± · · · · · · · · · · · · · · · ± · · · · · · · · · ± ± ± · · · · · · · · · · · · · · · ± ± · · · ± ± ± ± · · · ± ± ± · · · · · · · · · ± ± ± ± · · · ± ± ± · · · · · · · · · ± ± ± ± ± · · · ± ± ±
39 9 131.2 ± ± ± ± · · · · · · · · · ± ± ± ± · · · ± ± ± ± · · · ± ± ± · · · · · · · · · ± ± ± · · · · · · · · · ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± · · · ± ± ± · · · · · · · · · ± ± ± ± ± ± ± ± ± ± ± ± ± · · · · · · ± ± ± ± · · · · · · · · · ± ± ± · · · · · · · · · ± ± ± ± ± ± ± ± · · · · · · · · · ± ± ± ± · · · ± ± ± · · · · · · · · · ± ± ± · · · · · · · · · ± ± ± · · · · · · · · · ± ± ± · · · · · · · · · ± ± ± · · · · · · · · · ± ± ± ± · · · ± ± ± ± · · · ± ± ± ± · · · ± ± ± · · · · · · · · · ± ± ± · · · · · · ± ± ± ± ATALOG OF STELLAR VELOCITY DISPERSIONS 13
TABLE 1—
Continued
Galaxy Literature Palomar Adopted σ Reference σ blue σ red σ final Notes σ ( km s − ) ( km s − ) ( km s − ) ( km s − ) ( km s − )NGC 3738 · · · · · · · · · ± ± ± · · · · · · · · · ± ± ± · · · · · · · · · ± ± ± ± · · · ± ± ± · · · · · · · · · ± ± ± ± ± ± ± ± · · · · · · · · · ± ± ± · · · · · · ± ± ± ± ± · · · ± ± ± ± ± ± ± ± · · · · · · ± ± ± ± · · · · · · · · · ± ± ± ± · · · ± ± ± ± ± ± ± ± ± ± ± ± ± ± · · · ± ± ± ± · · · ± ± ± · · · · · · · · · ± ± ± · · · · · · ± ± ± ± ± · · · ± ± ± ±
20 10 186.3 ± ± ± ± ±
10 1 311.1 ± · · · ± ± · · · · · · · · · ± ± ± ± ± ± ± ± ± ± ± ± ± ± · · · ± ± ± ± · · · ± ± ± ± · · · ± ± ± · · · · · · · · · ± ± ± ± · · · ± ± ± · · · · · · · · · ± ± ± · · · · · · · · · ± ± ± ± ± ± ± ± ± ± ± ± ± · · · · · · · · · ± ± ± ± · · · ± ± ± ± ± ± ± ± · · · · · · · · · ± ± ± ± ± ± ± ± ± ± ± ± ± · · · · · · · · · · · · · · · < · · · · · · · · · · · · · · · · · · NGC 4150 87 ± · · · ± ± ± ± · · · · · · · · · ± · · · · · · · · · ± ± ± ± · · · ± ± ± · · · · · · · · · ± ± ± ± ± ± ± ± ± ± ± ± ± TABLE 1—
Continued
Galaxy Literature Palomar Adopted σ Reference σ blue σ red σ final Notes σ ( km s − ) ( km s − ) ( km s − ) ( km s − ) ( km s − )NGC 4178 · · · · · · · · · ± ± ± ± ± ± ± ± · · · · · · · · · · · · · · · ± ± ± ± ± ± ± ± ± ± ± ± · · · ± ± ± · · · · · · · · · · · · · · · ± ± ± ± ± ± ± · · · ± ± ± ± ± ± ± ± ±
10 12 134.1 ± · · · ± ± · · · · · · · · · · · · · · · < · · · · · · · · · · · · · · · · · · NGC 4244 · · · · · · · · · ± ± ± · · · · · · · · · ± ± ± ± ± ± ± ± ± · · · ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± · · · · · · · · · ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± · · · ± ± ± · · · · · · · · · ± ± ± ± · · · ± ± ± ± ± ± ± ± ± · · · ± ± ± ± · · · ± ± ± ± · · · ± ± ± ± · · · ± ± ± · · · · · · ± ± ± ± ± ± ± ± ± ± ± ± ± ± · · · · · · · · · ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± · · · ± ± ± ± · · · ± ± ± ± · · · · · · · · · ± NGC 4405 · · · · · · · · · · · · · · · ± ± ± ± ± ± ATALOG OF STELLAR VELOCITY DISPERSIONS 15
TABLE 1—
Continued
Galaxy Literature Palomar Adopted σ Reference σ blue σ red σ final Notes σ ( km s − ) ( km s − ) ( km s − ) ( km s − ) ( km s − )NGC 4414 117 ± · · · ± ± ± ± ± ± ± ± ± · · · ± ± ± ± · · · ± ± ± · · · · · · · · · ± ± ± ± ± ± ± ± ± ± ± ± ± · · · · · · ± ± ± ± ± ± ± ± ± ± · · · ± ± ± · · · · · · · · · ± ± ± ± ± ± ± ± ± · · · ± ± ± ± ± ± ± ± · · · · · · · · · ± ± ± ± ± ± ± ± · · · · · · ± ± ± ± · · · · · · · · · ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± · · · · · · · · · · · · · · · ± ± ± ± ± ± · · · · · · · · · ± ± ± ± ± ± ± ± · · · · · · · · · · · · · · · ± · · · · · · ± ± ± ± ± ± ± ± ± ± ± ± ± ± · · · · · · · · · ± ± ± ± ± ± ± ± ± ± ± ± ± · · · · · · · · · · · · · · · < · · · · · · · · · ± ± ± ± ± ± ± ± ±
15 1 · · · ± ± ± ± · · · ± ± ± ± ± ± ± ± · · · · · · · · · ± ± ± ± ± ± ± ± ± ± ± ± ± · · · · · · · · · ± ± ± · · · · · · · · · ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± TABLE 1—
Continued
Galaxy Literature Palomar Adopted σ Reference σ blue σ red σ final Notes σ ( km s − ) ( km s − ) ( km s − ) ( km s − ) ( km s − )NGC 4594 241.1 ± ± ± ± ± ± ± ± ± ± · · · · · · · · · ± ± ± ± ± ± ± ± ± · · · ± ± ± · · · · · · · · · · · · · · · < ± ± ± ± ± · · · · · · · · · · · · · · · < ± ± ± ± ± ± · · · ± ± ± ± · · · ± ± ± ± ± ± ± ± ± · · · ± ± ± ± ± ± ± ± ± ± ± ± ± · · · · · · · · · ± ± ± · · · · · · · · · ± ± ± · · · · · · · · · · · · · · · ± ± ± ± ± ± · · · · · · ± ± ± ± · · · · · · · · · ± ± ± · · · · · · · · · ± ± ± ± · · · · · · · · · ± ± ± ± ± ± ± · · · ± ± ± · · · · · · · · · ± ± ± ± ± ± ± ± ± ± ± ± ± · · · · · · ± ± ± ± ± ± ± ± ± ± ± ± ± ± · · · · · · ± ± ± ± · · · · · · · · · ± ± ± ± ± ± ± ± ± ± ± ± ± · · · · · · · · · ± ± ± ± ± ± ± ± · · · · · · · · · ± ± ± ± ± ± ± ± ±
10 6 178.9 ± ± ± ± · · · · · · · · · ± ± ± ± ± ± ± ± ± · · · ± ± ± ± ± ± ± ± · · · · · · · · · · · · · · · < · · · · · · · · · ± ± ± ± · · · ± ± ± ± ± ± ± ± · · · · · · · · · · · · · · · ± ATALOG OF STELLAR VELOCITY DISPERSIONS 17
TABLE 1—
Continued
Galaxy Literature Palomar Adopted σ Reference σ blue σ red σ final Notes σ ( km s − ) ( km s − ) ( km s − ) ( km s − ) ( km s − )NGC 5248 · · · · · · · · · ± ± ± ± · · · ± ± ± ± ± ± ± ± · · · · · · · · · ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± · · · ± ± ± · · · · · · ± ± ± ± · · · · · · ± ± ± ± · · · · · · · · · ± ± ± · · · · · · ± ± ± ± · · · · · · ± ± ± ± ± · · · ± ± ± ± ± ± ± ± · · · · · · · · · ± ± ± ± ± ± ± ± · · · · · · · · · ± ± ± ±
12 8 · · · · · · · · · ± ± ± ± ± ± · · · · · · ± ± ± ± ± ± ± ± ± ± · · · · · · · · · ± NGC 5631 168.2 ± ± ± ± ± ± ± ± ± ± · · · · · · · · · ± ± ± · · · · · · · · · ± ± ± ± · · · · · · · · · ± NGC 5669 · · · · · · · · · · · · · · · ± ± · · · ± ± ± ± · · · ± ± ± · · · · · · · · · · · · · · · < ± ± ± ± ± ± ± ± ± ± · · · · · · ± ± ± ± · · · · · · · · · ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± · · · · · · ± ± ± ± ± ± ± ± ± ± · · · ± ± ± · · · · · · ± ± ± ± · · · · · · ± ± ± ± · · · · · · · · · ± ± ± · · · · · · · · · ± ± ± TABLE 1—
Continued
Galaxy Literature Palomar Adopted σ Reference σ blue σ red σ final Notes σ ( km s − ) ( km s − ) ( km s − ) ( km s − ) ( km s − )NGC 5970 · · · · · · · · · ± ± ± ± ± ± ± ± · · · · · · ± ± ± ± · · · · · · · · · ± ± ± · · · · · · · · · ± ± ± ± · · · ± ± ± · · · · · · · · · ± ± ± · · · · · · · · · ± ± ± ± · · · ± ± ± · · · · · · · · · · · · · · · ± ± ± ± ± ± ± ± ± ± ± ± · · · ± ± ± ± · · · ± ± ± ± ± ± ± ± ± · · · ± ± ± ± · · · · · · · · · ± ± · · · ± ± ± · · · · · · ± ± ± ± ±
11 1 · · · · · · · · · ± ± · · · ± ± ± ± ± ± ± ± · · · · · · · · · ± ± ± ± · · · ± ± ± · · · · · · · · · ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± · · · ± ± ± ± · · · ± ± ± ±
11 1 151.6 ± ± ± ± ± ± ± ± ± ± ± ± ± ± · · · · · · · · · · · · · · · ± · · · · · · · · · ± ± ± ±
11 1 · · · ± ± ± ± · · · ± ± ± · · · · · · · · · ± ± ± ± ± ± ± ± · · · · · · · · · ± ± ± · · · · · · · · · ± ± ± ± · · · ± ± ± · · · · · · · · · ± ± ± · · · · · · · · · ± ± ± OTE .—Notes: (1) The adopted velocity dispersion was estimated from the [N II] λ σ = 240km s − pertains to the “southeast-northwest” component of this merging galaxy; the correct dispersion for the primary nucleus is σ = 100 ±