INTEGRAL-FIELD STELLAR AND IONIZED GAS KINEMATICS OF PECULIAR VIRGO CLUSTER SPIRAL GALAXIES

It is well known that the environment affects the morphological types of galaxies in clusters. Many studies have shown that galaxies in clusters evolve morphologically with spirals becoming redder and in some cases lenticular as the result of environmental effects (Dressler 1980; Butcher & Oemler 1978, 1984; Dressler et al. 1997; Poggianti et al. 1999, 2009; Kormendy & Bender 2012) Several mechanisms have been proposed for driving galaxy evolution, including processes that affect the stars, gas, and dark matter, and processes that affect only the gas. In the first category, we include the following: (1) low-velocity tidal interactions and mergers (e.g., Toomre & Toomre 1972; Hernquist 1992), (2) high-velocity tidal interactions and collisions (e.g., Moore et al. 1996), and (3) tidal interaction between galaxies and the cluster as a whole or between galaxies and substructures within the cluster (Bekki 1999). In the second category, we include the following: (1) intracluster medium–interstellar medium (ICM-ISM) stripping (Gunn & Gott 1972; Nulsen 1982; Schulz & Struck 2001; Vollmer et al. 2001; van Gorkom 2004; Cen 2014), (2) gas accretion, which may occur in the outskirt of clusters, and (3) starvation or strangulation, where the galaxies could lose their gas reservoir, thus preventing their accretion onto the galaxy (Larson et al. 1980). While all of these processes probably do actually occur, it remains unclear which ones are dominant in driving the morphological evolution of cluster galaxies.

Detailed studies of stellar and ionized gas kinematics can help to discriminate between the different interaction processes. For example, gravitational interactions produce disturbed kinematics in both the stellar and gas components, whereas interactions of a hydrodynamic nature will directly affect only the gas. Recently, with the advent of Integral Field Units (IFUs) such as DensePak, SAURON, GMOS, SINFONI, and MUSE, these detailed studies have become possible. The observed velocity fields can be compared with those from simulations (e.g., Bendo & Barnes 2000; Jesseit et al. 2007; Kronberger et al. 2007, 2008), providing important clues about the physical processes that drive galaxy evolution.

The Virgo cluster is the nearest moderately rich cluster with a galaxy population spanning a large range of morphological types. The cluster has a moderately dense ICM and is dynamically young with on-going sub-cluster mergers and infalling galaxies, making it an ideal place for detailed studies of various environmental processes. Moreover, the Virgo cluster has a significant population of galaxies characterized by truncated star formation morphologies with no Hα in the outer disk but strong Hα in the inner region (Koopmann & Kenney, 2004), consistent with ICM-ISM stripping. However, some of them also have other peculiarities that are not presently well understood, presumably reflecting different types of interactions. These peculiar galaxies may be in the process of morphological transformation, and could be considered as "snapshots" in the evolutionary path from actively star-forming spiral galaxies to more passive spirals and lenticulars.

With these objectives in mind, in this work, we present a study of the stellar and ionized gas kinematics of 13 peculiar Virgo cluster galaxies using integral-field spectroscopy techniques. We profit from the ability of this technique to accurately map two-dimensional (2D) velocity fields for both the stars and the ionized gas in the centers of these galaxies. This data set has previously been used to estimate the three-dimensional (3D) cluster locations of all sample galaxies by using stellar kinematics to derive their distances (Cortés et al. 2008), and for a detailed investigation of the nature of two of the most peculiar galaxies of the present sample, NGC 4064 and NGC 4424 (e.g., Cortés et al. 2006).

The present paper is structured as follows. A brief description of the galaxy sample is given in Section 2. The observation and data reduction procedures are summarized in Section 3. A description of the observed stellar and ionized gas kinematics is presented in Section 4. In Section 5, we summarize the kinematical peculiarities observed in these galaxies. In Section 6, we discuss the velocity dispersion profiles and kinematical support of these galaxies. In Section 7, we compare our observations with simulations of merger remnants, ICM-ISM stripped galaxies, and tidal interactions. These results are discussed from the perspective of galaxy evolution in clusters in Section 8. We summarize our results and present our conclusions in Section 9. A discussion of individual galaxies is given in the Appendix.

The sample consists of 13 peculiar Virgo cluster spiral galaxies spanning a variety of optical morphologies (Figure 1, Table 1). Morphological selection was performed using the R and Hα atlas of Virgo cluster galaxies of Koopmann et al. (2001), whereas the kinematical selection made use of the published Hα kinematics of 89 Virgo cluster spirals by Rubin et al. (1999). While the sample selection is not uniform, it is designed to include bright Virgo spirals whose peculiarities are most poorly understood, and to include representatives of the different Hα types identified by Koopmann & Kenney (2004). In choosing sample galaxies within a given Hα type, we gave preference to those with kinematical peculiarities.

Zoom In Zoom Out Reset image size

Table 1. Galaxy Sample Properties

Name	R.A. (J2000)	Decl. (J200)	RSA/BST	RC3	C30	SFC	Hc	MH	R25	Inc.	PA	Vhelio	DM87	D
									(arcsec)	(deg)	(deg)	(km s−1)	(deg)	(Mpc)
(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)	(12)	(13)	(14)	(15)
NGC 4064	12 04 11.2	18 26 36	SBc(s):	SB(s)a:pec	0.43	T/C	8.78a	−22.5 $\pm\, ^{0.1}_{0.2}$	138	70	148	931 ±  12	8.8	18.0 $\pm\, ^{1.3}_{0.8}$
NGC 4293	12 21 12.8	18 22 57	Sa pec	(R)SB(s)0/a	0.40	T/A	7.35	−23.4 $\pm\, ^{0.3}_{0.2}$	216	67	62	930 ±  14	6.4	14.1 $\pm\, ^{1.0}_{1.5}$
NGC 4351	12 24 01.6	12 12 18	Sc(s) II.3	SB(rs)ab: pec	0.37	T/N[s]	10.44	−21.1 ± 0.1	92	47	65	2317 ±  16	1.7	20.3 $\pm\, ^{0.8}_{1.2}$
NGC 4424	12 27 11.5	09 25 15	Sa pec	SB(s)a:	0.42	T/C	9.17a	−21.7 ± 0.3	125	60	88	440 ± 6	3.1	15.2 ± 1.9
NGC 4429	12 27 26.4	11 06 29	S03(6)/Sa pec	SA(r)0+	0.54	T/A	6.76	−24.3 ± 0.1	220	60	89	1127 ± 31	1.5	16.3 ± 0.9
NGC 4450	12 28 29.3	17 05 07	Sab pec	SA(s)ab	0.49	T/A	6.90	−23.6 $\pm\, ^{0.1}_{0.2}$	205	46	179	1958 ± 6	4.7	12.6 $\pm\, ^{1.0}_{0.6}$
NGC 4457	12 28 59.3	03 34 16	RSb(rs) II	(R)SAB(s)0/A	0.66	T/N[s]	7.96	−23.1	160b	34 c	79d	881 ± 14	8.8	16e
NGC 4569	12 36 49.8	13 09 46	Sab(s) I-II	SAB(rs)ab	0.43	T/N[s]	6.77	−23.80 ± 0.01	266	64	26	−232 ± 22	1.7	13.0 ± 0.1f
NGC 4580	12 37 48.6	05 22 06	Sc/Sa	SAB(rs)a pec	0.41	T/N[s]	8.77	−22.95 ± 0.05	87	45	154	1036 ± 7	7.2	22.1 ± 0.5
NGC 4606	12 40 57.6	11 54 44	Sa pec	SB(s)a:	0.42	T/C	9.30	−22.20 $\pm\, ^{0.05}_{0.11}$	103	67	44	1655 ± 16	2.5	19.9 $\pm\, ^{1.0}_{0.5}$
NGC 4651	12 43 42.6	16 23 36	Sc(r) I-II	SA(rs)c	0.55	N	8.25	−23.90 $\pm\, ^{0.10}_{0.04}$	157	51	72	804 ± 10	5.1	26.9 $\pm\, ^{0.5}_{1.2}$
NGC 4694	12 48 15.1	10 59 00	Amorph	SB0 pec	0.62	T/C	9.03	−21.6 ± 0.2	120	42	146	1177 ± 11	4.5	13.4 $\pm\, ^{1.3}_{1.0}$
NGC 4698	12 48 23.0	08 29 14	Sa	SA(s)ab	0.51	A	7.43	−24.50 $\pm\, _{0.41}^{0.83}$	187	62	169	1005 ± 15	5.8	24.3 $\pm\, ^{5.1}_{4.2}$g

Notes. (1) Galaxy name; (2) right ascension in hours, minutes, and seconds; (3) declination in degrees, minutes, and seconds; (4) Hubble Types from Bingelli et al. (1987, hereafter BST), Sandage & Tammann (1987), or Sandage & Bedke (1994); (5) Hubble type from de Vaucouleurs et al. (1991, hereafter RC3); (6) central R light concentration parameter (Koopmann et al. 2001); (7) star formation class from Koopmann & Kenney (2004); (8) apparent magnitude in H band (Gavazzi et al. 1999); (9) absolute magnitude in H band (Cortés et al. 2008); (10) radius in units of arcseconds at the 25 R mag arcsec⁻² isophote (Koopmann et al. 2001); (11) inclination from Koopmann et al. (2001); (12) optical PA at R₂₅ (J. R. Cortés et al., in preparation); (13) heliocentric radial velocity from HyperLEDA; (14) the projected angular distance in degrees of the galaxy from M87; (15) line-of-sight distance (Cortés et al. 2008). ^a2MASS Galaxy Atlas, Jarrett et al. (2003). ^bExtrapolation from the R-band light profile. ^cHyperLEDA database. ^dMeasured at 100'' ∼ 0.6R₂₅. ^eVirgo Cluster distance. ^fStellar-kinematics-based distance considering dark matter halo from Cortés et al. (2008). ^gH i-based distance from Solanes et al. (2002).

Download table as:  ASCIITypeset image

In Virgo spirals, there is a poor correlation between the bulge-to-disk ratio and the normalized star formation rate so that the Hubble classification of spirals does not work well in the Virgo Cluster (Koopmann & Kenney 1998). The Hubble-type classification assigned to Virgo galaxies generally reflects the star formation rate rather than the bulge-to-disk ratio, so that Virgo spirals with reduced star formation rates are generally classified as early-type spirals, independent of their bulge-to-disk ratio. Our sample includes more early-type than late-type galaxies, since most of the strongly disturbed cluster galaxies have reduced star formation rates and so are classified as early types. Since the Hubble classifications of cluster spirals do not capture the intrinsic variation in galaxy morphologies, either the distributions of old stellar light or the young stars, we use instead the light concentration parameter C₃₀ (Abraham et al. 1994) as an objective measure of the bulge-to-disk ratio, and the Hα type (Koopmann & Kenney 2004) to describe the radial distribution of star formation.

Since Hα types seem to correlate with the type of interaction experienced by the galaxy (Koopmann & Kenney 2004), and also with some kinematical properties that we describe in this paper, here we provide brief definitions of these categories and indicate to which category our sample galaxies belong.

• 1.  
  Normal: NGC 4651. In Normal galaxies, the Hα radial distribution is close to the mean of isolated spirals, both in the shape of the radial distribution (which is close to the R light profile) and in amplitude.
• 2.  
  Truncated/Normal: NGC 4351, NGC 4457, NGC 4569, and NGC 4580. In Truncated/Normal galaxies, the Hα radial distribution is similar to that in a Normal galaxy out to a well-defined truncation radius, but there is virtually no Hα emission beyond.
• 3.  
  Truncated/Compact: NGC 4064, NGC 4424, NGC 4606, and NGC 4694. In Truncated/Compact galaxies, the Hα radial profile is much steeper than the R light profile at all radii, with a strong central peak and a very sharp drop with radius, such that there is virtually no emission beyond the central 1 kpc. In Koopmann & Kenney (2004), this category also included the provision that the central Hα intensity was much higher than Normal, but here we relax this provision. Thus we include NGC 4694⁷ in this category, which has an Hα radial profile like the other T/C galaxies but with a lower amplitude.
• 4.  
  Anemic or Truncated/Anemic galaxies: NGC 4293, NGC 4429, NGC 4698, and NGC 4450. In Anemic galaxies, the shape of the Hα radial distribution is similar to that of a Normal galaxy (and like the R light profile), but with a much lower amplitude. Anemic galaxies have weak but detectable emission over much of the stellar disk. Truncated/Anemic galaxies have an anemic distribution out to a well-defined truncation radius, but virtually no Hα emission beyond. Whereas most of our sample galaxies in these categories have Hα emission detected out to at least 0.3R₂₅, we also include in the Truncated/Anemic category the S0 galaxy NGC 4429, which has a very small and very weak Hα disk.

Distances to many individual early-type Virgo galaxies have now been measured to good accuracy, based on the surface brightness fluctuation method. The mean distance to early-type Virgo cluster galaxies is 16.5 ± 0.1 mpc (Mei et al. 2007). The distances to late-type and peculiar Virgo galaxies are less accurately known. The commonly used Tully–Fisher method based on H i line-widths works well for most late-type galaxies, but produces erroneous distances for a subset of cluster galaxies that are H i-poor and disturbed (Cortés et al. 2008). In Cortés et al. (2008), we derived distances to galaxies in the present Virgo sample based on a new approach using a stellar kinematics-based version of the Tully–Fisher relation. It is revealing to compare different distance estimates for perhaps the most disturbed galaxy in our sample, NGC 4424. The H i-based Tully–Fisher method provides a distance of 4.8 mpc (Solanes et al. 2002), our stellar-kinematics-based Tully–Fisher relation provides a distance of 15.2 mpc (Cortés et al. 2008), and a recent Type Ia supernova provides a distance of 15.5 mpc (Munari et al. 2013). While this is an extreme example, it illustrates that our stellar-kinematics-based Tully–Fisher distances may be relatively accurate. We adopt these distances in this paper and present them in Table 1.

3.1. Observations and Data Reduction

The galaxies were observed using the DensePak Integral-field unit (Barden et al. 1998) installed at the 3.5m WIYN telescope at Kitt Peak. The DensePak array consists of 90 fibers with a diameter of 35, and about 4'' separation with a covering area of 30'' ×45'', which corresponds to 2.3 × 3.5 kpc at the mean Virgo distance. The observations were taken in three observing runs during 1999 April 8–9th, 2001 May 24–25th 2001, and 2002 February 10–12th. The 860 line mm⁻¹, blaze angle 309 grating at 5000 Å was used to second order, covering the 4500–5500 Å wavelength range with a spectral dispersion of 0.48 Å per pixel and a spectral resolution of 2.02 Å. The DensePak array, in most cases, was oriented along the optical major axis of the galaxies (Table 2, Figure 2). Details about the exposure time used to observe the galaxies are given in Table 2. Radial velocity standards stars (Barbier–Brossat & Figon 2000) were observed as spectral templates with an exposure time of 180 s. Details about each star are presented in Table 3.

Zoom In Zoom Out Reset image size

Table 2. DensePak Observation Logs

Name	Dates	PADensePak	Texp
		(°)	(s)
(1)	(2)	(3)	(4)
NGC 4604	1999 Apr 8	150	4 × 1800
NGC 4293	2002 Feb 11–12	72	1800, 2 × 1800
NGC 4351	2001 May 25	80	4 × 1800
NGC 4424	1999 Apr 9	90	6 × 1800
NGC 4429	2002 Feb 12	99	2 × 1800
NGC 4450	2002 Feb 12	175	4 × 1800
NGC 4457	2002 Feb 11	0	4 × 1800
NGC 4569	2001 May 24	0	4 × 1800
NGC 4580	2002 Feb 11	158	4 × 1800
NGC 4606	2002 Feb 12	33	4 × 1800
NGC 4651	2001 May 24	71	4 × 1800
NGC 4694	1999 Apr 9	140	3 × 1800
NGC 4698	2002 Feb 10	170	4 × 1800

Notes. (1) Galaxy name; (2) observation date; (3) PA of the DensePak array (taken from Koopmann et al. 2001); (4) exposure time.

Download table as:  ASCIITypeset image

Table 3. Characteristics of Template Stars

Name	Template Star	Spectral Type	(km s−1)
			(4)
(1)	(2)	(3)	Vhelio
NGC 4064	HD 90861	K2 III	37.13
NGC 4293	HD 69632	K0 III	−0.90
NGC 4351	HD 159479	K2 III	−24.6
NGC 4424	HD 86801	G0 V	−14.5
NGC 4429	HD 69632	K0 III	−0.90
NGC 4450	HD 69632	K0 III	−0.90
NGC 4457	HD 35005	G7 III	25.90
NGC 4569	HD 159479	K2 III	−24.6
NGC 4580	HD 35005	G7 III	25.90
NGC 4606	HD 35005	G7 III	25.90
NGC 4651	HD 159479	K2 III	−24.6
NGC 4694	HD 90861	K2 III	37.13
NGC 4698	HD 77823	K2 III	65.90

Notes. (1) Galaxy name; (2) template star name; (3) template star spectral type; (4) heliocentric radial velocity (Barbier–Brossat & Figon 2000).

Download table as:  ASCIITypeset image

The spectra were reduced in the usual manner with IRAF reduction for multi–fiber spectrographs. They were zero-subtracted and overscan-corrected with the standard IRAF tasks. The flat fielding correction, sky subtraction, fiber throughput correction, and wavelength calibration were carried out using the IRAF task DOHYDRA in the package HYDRA.⁸ Alignment between different exposures was checked by comparing the continuum maps between each exposure. The exposures were aligned using our own program written in IDL.⁹ After the alignment, the galaxy spectra were averaged to improve the signal-to-noise ratio and remove cosmic rays.

In two sample galaxies, NGC 4424 and NGC 4351, the spectra in many positions have a signal-to-noise ratio per pixel lower than the limit for obtaining reliable kinematics until the second moment (S/N ∼ 15). Therefore, spectra in these galaxies were binned using the adaptive scheme developed by Cappellari & Copin (2003), for achieving a signal-to-noise ratio good enough for obtaining reliable kinematics.

3.2. Stellar and Gas Kinematics

Stellar kinematics of the sample galaxies were derived by using the penalized pixel-fitting (pPXF) method developed by Cappellari & Emsellem (2004). This method allows the masking of emission lines, a more realistic estimation of errors than in other methods, and automation, allowing us to derive the kinematics of many spectra in a relatively short amount of time. Although the method can measure the Gauss-Hermite moments of the line-of-sight velocity distribution (LOSVD) up to h₆, the low signal-to-noise ratio of the spectra (S/N < 60) in most of our sample galaxies only allow us to obtain reliable measurements until the velocity dispersion (σ). We obtained Gauss-Hermite moments up to h₄ in only three sample galaxies: NGC 4429, NGC 4450, and NGC 4561. Template spectra were chosen from the spectra of observed radial velocity standard stars (typically G and K giants) that better fit the galaxy spectra. The pPXF method determines the broadening function (LOSVD) between the template star spectra and the galaxy spectra. If the instrumental dispersions between the template and the galaxy are different, then we need to convolve the template spectra by the square root of the quadratic difference between the instrumental dispersion of the galaxy and the template star. Since we observed the template stars and sample galaxies using the same observational setup during the same observing runs, and therefore the instrumental dispersions are the same for both, it was not necessary to convolve the template star and the derived velocity dispersion is already corrected by instrumental dispersion. A linear combination of different template star did not show any significant improvement in the fits. The heliocentric correction was applied to the derived line-of-sight velocities in each galaxy. Errors were estimated using a Monte-Carlo scheme. They were obtained as the standard deviation of the kinematical parameters (V and σ) from many realizations (N = 100) of the input spectra by adding Gaussian noise to a model galaxy spectrum. Comparison between the galaxy spectra in the central region and the model galaxy spectra obtained by pPXF from the template star spectra are shown in Figure 3.

Zoom In Zoom Out Reset image size

2D ionized gas kinematics were derived from the emission lines (Hβ and [O iii] λ5007) by fitting a Gaussian function to the emission lines in the continuum subtracted spectra, thereby obtaining V and σ for the ionized gas. Finally, we applied a heliocentric correction to the derived gas velocities. In this work, we focus only on the ionized gas line-of-sight velocity.

3.3. Kinematic Parameter Determination

Systemic velocities, kinematic position angles, and rotational velocities were obtained for each galaxy by fitting a pure circular tilted ring model (Begeman 1989) to the stellar and ionized gas velocity fields in cases where this was possible.

The method consists of dividing the galaxy into a set of concentric rings, each ring being characterized by a fixed value of the inclination i, the rotation velocity V_rot, and the kinematic position angle ϕ. We define the sky coordinates (x, y) as the DensePak array coordinates with x oriented over the position angle of the DensePak array ϕ_DensePak (see Table 2). For a given ring, the observed radial velocities recorded on a set of sky-coordinates are given by

$$\begin{equation} V_{\rm obs}(x,y)=V_{\rm sys}+V_{\rm rot}(R) \sin (i) \cos (\theta), \end{equation} \tag{ 1 }$$

where θ is the azimuthal angle in the plane of the galaxy, measured from the optical major axis of the galaxy. It can be shown that for any point (x_gal, y_gal) in the plane of the galaxy, θ is given by

$$\begin{equation} \cos(\theta)= \dsty\frac{x_{\rm gal}}{R} = \dsty\frac{(x -x_{0}) \cos (\phi - \phi _{\rm DensePak}) + (y-y_{0}) \sin (\phi - \phi _{\rm DensePak})}{R}, \end{equation} \tag{ 2 }$$

$$\begin{equation} \sin(\theta)= \dsty\frac{y_{\rm gal}}{R}= \dsty\frac{ (y-y_{0}) \cos (\phi - \phi _{\rm DensePak}) - (x-x_{0}) \sin (\phi - \phi _{\rm DensePak})}{R \cos (i)}, \end{equation} \tag{ 3 }$$

where R is the mean radius of the ring in the plane of the galaxy and (x₀, y₀) are the sky coordinates of the center of the ring. In this work, we define the position angle (either optical or kinematic) as the angle between the optical/kinematic major axis of the galaxy and the line from the galaxy center headed North, measured from north to east.

The procedure is iterative, and described as follows. (1) We divide the galaxy into concentric rings with typical widths of one fiber on the major axis. (2) For each ring, we make a least-square fit of V_obs with x₀, y₀ and V_sys as free parameters, keeping V_rot, ϕ, and i fixed. Then, we took as center coordinates and systemic velocity the mean values over all of the rings. (3) After this, we kept x₀, y₀ and V_sys fixed, and we fitted the values V_rot and ϕ simultaneously. We did not attempt to fit the inclination i as it was difficult to disentangle the degeneracy between i and V_rot, so we kept this parameter fixed to its optical value as derived from Koopmann et al. (2001) for all the rings. (4) The improved values define a new systemic velocity V_sys and set of sky coordinates x₀, y₀. Repeat (2) and (3) with these parameters until convergence is obtained.

On average, we have six degrees of freedom in each annuli. We are aware that points close to the minor axis carry less information about the rotational velocity, so we used a weighting function of |cos (θ)| in order to give more weight to points close to the major axis.

Stellar rotation curves were obtained for all of the galaxies in the sample. Kinematic parameters derived from the stellar velocities are presented in Table 4. Stellar rotation curves are not corrected for asymmetric drift, since this is beyond of the scope of this paper. This correction requires solution of the collisionless Boltzmann equation. Stellar circular velocities for these galaxies are presented in Cortés et al. (2008). The case of NGC 4064 was complicated since this galaxy has a strong central bar, and a model based on a pure circular disk is not fully justified but this approach is useful for separating the circular streaming component from the radial streaming component.

Table 4. Parameters and Derived Quantities for the Stellar Velocity Fields

Name	Vsys	PAouter	PAinter	PAkin	Ψouter	Ψinter	Vrot	Rmax
	(km s−1)	(°)	(°)	(°)	(°)	(°)	(km s−1)	(arcsec)
(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)
NGC 4064	929 ± 3	148	152	129a ± 15	19 ± 15	23 ± 15	72 ± 7	25
NGC 4293	926 ± 4	62	72	42 ± 8	20 ± 8	30 ± 8	105 ± 10	27
NGC 4351	2310 ± 2	65	70	53 ± 8	12 ± 8	17 ± 8	41 ± 19	20
NGC 4424	442 ± 4	88	96	87 ± 8	1 ± 8	9 ± 8	31 ± 4	25
NGC 4429	1094 ± 6	89	97	87 ± 3	2 ± 3	10 ± 3	136 ± 9	25
NGC 4450	1953 ± 2	178	173	166b ± 5	12 ± 5	7 ± 5	114 ± 6	25
NGC 4457	881 ± 2	⋅⋅⋅	77	70 ± 3	⋅⋅⋅	7 ± 3	83 ± 5	20
NGC 4569	−222 ± 6	26	19	32 ± 6c	6 ± 6	13 ± 6	103 ± 3	25
NGC 4580	1031 ± 4	154	160	158 ± 4	4 ± 4	2 ± 4	103 ± 8	25
NGC 4606	1636 ± 5	44	39	39 ± 4	5 ± 4	0 ± 4	55 ± 4	25
NGC 4651	784 ± 2	72	79	82 ± 4	10 ± 4	3 ± 4	178 ± 8	28
NGC 4694	1174 ± 22	146	142	146 ± 6	0 ± 6	4 ± 6	63 ± 2	26
NGC 4698	1025 ± 6	169	167	168d ± 3	1 ± 3	1 ± 3	116 ± 10	25

Notes. (1) Galaxy name; (2) heliocentric systemic velocity; (3) photometric PA at outer radii (R₂₅); (4) photometric PA at intermediate radii (0.5 R₂₅); (5) mean kinematical PA; (6) kinematical misalignment with respect to PA_outer; (7) kinematical misalignment with respect to PA_inter; (8) stellar rotation velocity at R_max; (9) most external point with reliable stellar rotation velocities. ^aStrong non-circular motions due to central bar. ^bThe stellar kinematic PA in NGC 4450 changes from ∼175° in the central ∼10'' to ∼160° at r ∼ 15–25''. This may reflect a change from the inner bulge-dominated region to the bar-dominated region at larger radius. There is good agreement between the stellar kinematic PA for r <10'' and the photometric PA of the outer galaxy. ^cThe stellar velocity field of NGC 4569 is irregular. The SE side of the velocity map suggests a different PA than the NW side, perhaps due to dust extinction. The SE side suggests a kinematic PA closer to the photometric PA of the outer galaxy. ^dThe mean stellar kinematic PA given here is dominated by values for r >10'', and so is not significantly influenced by the small orthogonally rotating bulge.

Download table as:  ASCIITypeset image

We obtained ionized gas rotation curves for six galaxies in the sample. We were unable to fit a pure circular tilted ring model to the ionized gas velocity fields for some galaxies due to scarcity of emission (e.g., NGC 4424, NGC 4606, or NGC 4694) or large non-circular motions (e.g., NGC 4064, NGC 4569, and NGC 4457). However, systemic ionized gas velocities were estimated for all of the galaxies, either by taking the mean of the averaged LOS velocities at the same radius on opposite sides along the major axis, or by considering the LOS velocity at the position of the peak of the stellar continuum map. The kinematic parameters derived from ionized gas velocities are presented in Table 5.

Table 5. Parameters and Derived Quantities for the Ionized Gas Velocity Fields

Name	Vsys Hβ	Vsys [O iii]	PAkin Hβ	Ψ Hβ	PAkin [O iii]	Ψ [O iii]	Vrot Hβ	Rmax Hβ	Vrot [O iii]	Rmax [O iii]
	(km s−1)	(km s−1)	(°)	(°)	(°)	(°)	(km s−1)	(arcsec)	(km s−1)	(arcsec)
(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)
NGC 4064	943 ± 4	947a ± 2	150	⋅⋅⋅	⋅⋅⋅	⋅⋅⋅	31 ± 9	12	⋅⋅⋅	⋅⋅⋅
NGC 4351	2305 ± 3	2309 ± 2	62 ± 16	8 ± 16	62 ± 16	8 ± 16	66 ± 16	26	84 ± 6	28
NGC 4450	⋅⋅⋅	1960 ± 9	⋅⋅⋅	⋅⋅⋅	194 ± 8	19 ± 8	⋅⋅⋅	⋅⋅⋅	141 ± 23	20
NGC 4457	858 ± 6	854 ± 6	⋅⋅⋅	⋅⋅⋅	⋅⋅⋅	⋅⋅⋅	⋅⋅⋅	⋅⋅⋅	⋅⋅⋅	⋅⋅⋅
NGC 4569	−294a ± 3	−240 ± 4	⋅⋅⋅	⋅⋅⋅	⋅⋅⋅	⋅⋅⋅	⋅⋅⋅	⋅⋅⋅	⋅⋅⋅	⋅⋅⋅
NGC 4580	1029 ± 3	⋅⋅⋅	158 ± 4	0 ± 4	⋅⋅⋅	⋅⋅⋅	113 ± 10	24	⋅⋅⋅	⋅⋅⋅
NGC 4606	1647a ± 1	1650a ± 3	⋅⋅⋅	⋅⋅⋅	⋅⋅⋅	⋅⋅⋅	⋅⋅⋅	⋅⋅⋅	⋅⋅⋅	⋅⋅⋅
NGC 4651	778 ± 6	778 ± 4	79 ± 2	8 ± 2	76 ± 2	5 ± 2	213 ± 5	28	217 ± 10	28
NGC 4694	1152 ± 2	1161 ± 3	⋅⋅⋅	⋅⋅⋅	⋅⋅⋅	⋅⋅⋅	⋅⋅⋅	⋅⋅⋅	⋅⋅⋅	⋅⋅⋅
NGC 4698	⋅⋅⋅	978a ± 7	⋅⋅⋅	⋅⋅⋅	⋅⋅⋅	⋅⋅⋅	⋅⋅⋅	⋅⋅⋅	163 ± 12	26

Notes. (1) Galaxy name; (2) Hβ systemic velocity; (3) [O iii] systemic velocity; (4) Hβ kinematic PA; (5) Hβ kinematic PA misalignment; (6) [O iii] kinematic PA; (7) [O iii] kinematic PA misalignment; (8) Hβ rotation velocity at last measurement; (9) most external point with reliable Hβ rotation velocities; (10) [O iii] rotation velocity at last measurement; (11) most external point with reliable [O iii] rotation velocities. ^aVelocity at central position of the array.

Download table as:  ASCIITypeset image

In some of our galaxies, the optical and/or kinematic position angles vary with radius, or there are misalignments between the optical and/or kinematic position angles. We discuss these variations in Section 5.1.

3.4. Comparison with the Literature

Eight galaxies in our sample have published stellar kinematic data from other groups. In general, the results are consistent given the difference in spatial resolution between our DensePak data (35) and that for the other groups (06–12).

Four galaxies have published long-slit stellar kinematics: NGC 4429 (Simien & Prugniel 1997), NGC 4450, NGC 4569 (Fillmore et al. 1986), and NGC 4698 (Héraudeau et al. 1999, Bertola et al. 1999). A comparison of our data with the data from these groups for the LOS velocity and velocity dispersion along the major axis is shown in Figure 4. In NGC 4429, our stellar velocity data show a good correspondence with the Simien & Prugniel data, although our velocity dispersions are slightly higher due to our lower spatial resolution. In NGC 4450, there are some disagreements between our data and the Fillmore et al. data which do not seem attributable to differences in spatial resolution. Since they did not publish any information about their errors, we do not know if the differences are significant. In the case of NGC 4569, the Fillmore et al. data exhibit a peak in the stellar velocity at 3'' that we do not resolve due to our lower spatial resolution. Both data sets seems consistent in their velocity dispersion measurements. In NGC 4698, our data seem consistent with the Héraudeau et al. data within the errors.

Zoom In Zoom Out Reset image size

Long-slit minor axis stellar kinematics published by Coccato et al. (2005) for NGC 4064 and NGC 4424 seem consistent with our stellar kinematic measurements. 2D stellar kinematics for NGC 4293 and NGC 4698 were obtained by the SAURON project (Falcón–Barroso et al. 2006). Our stellar velocity and velocity dispersion fields are similar to theirs, with differences inherent to the different spatial resolutions between SAURON and DensePak (094 for SAURON and 35 for DensePak). Mazzalay et al. (2014) presented a 3'' ×3'' Brγ stellar velocity field with SINFONI, with a high spatial resolution (∼025). This map exhibits a higher amplitude in velocity within the inner 3'' and a lower velocity dispersion, which is expected due to the huge difference in the spatial resolutions between the two maps.

Long-slit ionized gas kinematics (Hα) have been published by Rubin et al. (1999). In Figure 5, we present a comparison between Rubin's Hα long-slit kinematics and our ionized gas LOS velocities (Hβ, and [O iii] λ5007) along the major axis for nine sample galaxies. Our ionized gas velocities show good correspondence with Rubin's data, although in NGC 4651 our ionized gas velocities display an apparent lower rotation velocity. This could be due to the different spatial resolution between the long-slit spatial scale (2'') and DensePak (35), which will smear out DensePak velocities. There are also some systematic differences between [O iii] velocities and Hα velocities for NGC 4694 and NGC 4698. We do not know the cause of these differences. 2D ionized gas velocity fields have been obtained by Chemin et al. (2006) in Hα for NGC 4351, NGC 4450, NGC 4569, and NGC 4580. Hβ velocity fields look similar to the Hα velocity fields, but with differences inherent to the different spatial resolution (∼1'' for Chemin et al.) and signal-to-noise ratio since Hα is much more intense than Hβ emission. Falcón–Barroso et al. (2006) also presented Hβ and [O iii] ionized gas velocity fields for NGC 4293 and NGC 4698. We were not able to detect any gas emission in NGC 4293. In NGC 4698, both [O iii] velocity fields seem consistent, with differences inherent to the different spatial resolutions between SAURON and DensePak.

Zoom In Zoom Out Reset image size

The works of Chemin et al. (2006) and Falcón–Barroso et al. (2006) have determined kinematic PAs for some of our sample galaxies. Chemin et al. (2006) presented Hα kinematic PAs for NGC 4351, NGC 4450, NGC 4457, NGC 4569, and NGC 4580. For NGC 4351, we derived (see Section 5.1) PA_kin Hβ = 62° ± 16°, which is consistent with PA_kin Hα = 73° ± 5° from Chemin et al. (2006). Both gas kinematic PAs are different from the stellar kinematic PA_kin = 53° ± 8°. NGC 4351 is a lopsided galaxy that may be experiencing ram pressure stripping, which may account for the difference in kinematic PAs between stars and gas. For NGC 4450, Chemin et al. give a gas (Hα) kinematic PA = 171° ± 7° in agreement with our stellar kinematic PA. For NGC 4457 and NGC 4580, Chemin et al.'s kinematic PAs are in agreement with our stellar kinematic PA. In the case of NGC 4569, our stellar kinematic PA is 32° ± 6°, which contrasts with 23° ± 4° derived by Chemin et al. Our stellar velocity field of NGC 4569 is irregular. The SE side of the velocity field suggests a different PA than the NW side, perhaps due to dust extinction. The SE side suggests a kinematic PA closer to the photometric PA of the outer galaxy, and therefore closer to the Chemin et al. PA determination. The stellar kinematic PA derived in the near-infrared for NGC 4569 (Mazzalay et al. 2014) is about 25°. This seems to confirm that our derivation of the stellar kinematic PA is affected by dust extinction. Falcón–Barroso et al. (2006) provide the differences between the photometric and kinematic PAs for NGC 4293 and NGC 4698. Their values are 30° for NGC 4293 and 3° for NGC 4698, which are in agreement with our determination of Ψ_inter for both galaxies (see Table 4).

Figure 6 displays maps of the absorption and emission-line kinematics, as well as stellar continuum and emission line intensities for the 13 galaxies in our sample. The maps are displayed in order of increasing NGC number with the same scale and spatial resolution, and they are oriented with north up and east to the left. For each galaxy, we display (first row) the R-band image from the WIYN Telescope, maps of the distribution and kinematics of the Hβ and [O iii] λ5007 emission lines for all the galaxies where emission was detected (12/13), and ionized gas residual velocity maps where possible. The second row displays the continuum map (reconstructed by integrating the spectra within 5250 to 5450 Å), the mean stellar velocity field V, the stellar residual velocity map, the velocity dispersion field σ, and where possible the h₃ and h₄ moment maps. The open circles in both stellar and ionized gas maps represent broken fibers. The values at these positions were obtained by bilinear interpolation, but were used only for the purpose of display and were not used in the subsequent analysis. The crosses in the maps represent the positions of the peak in the continuum emission, and the straight line represents the optical PA of the galaxy obtained either by us or Koopmann et al. (2001). Also, in Figure 7, we present the stellar and ionized gas kinematics along the kinematic major and minor axes.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Detailed descriptions of the different galaxies are collected in the Appendix. Here, we concentrate on an overview of the general trends of the maps and the resulting velocity profiles.

4.1. Stellar and Ionized Gas Velocity Fields

4.2. Stellar and Ionized Gas Rotation Curves, Stellar Velocity Dispersion, and V/σ Radial Profiles

Stellar and ionized gas rotation curves, as well as radial profiles of stellar velocity dispersion and V/σ are shown in Figure 8. Residual maps between the observations and pure circular models are displayed in Figure 6. The velocity dispersion profiles were calculated by folding the stellar velocity dispersion along the kinematic major axis (given by PA_kin in Table 4). Errors were estimated as the difference between the measured values and the mean of the velocity dispersions on the two sides of the galaxy. Velocity dispersion profiles are not corrected for inclination effects, as this requires a detailed model of the velocity ellipsoid which is beyond the scope of this work. The V/σ radial profile was calculated as the ratio between the stellar rotation curve (corrected by inclination) and the velocity dispersion profile at the same galactocentric radius.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Rotation curves, stellar velocity dispersion, and V/σ profiles for all of the sample galaxies are overplotted in Figure 9. The stellar rotation curves span a range of amplitudes and a variety of shapes. There are galaxies with high stellar rotation velocities (e.g., NGC 4651 or NGC 4429) and those with extremely low stellar rotation velocities (e.g., NGC 4424, NGC 4351, or NGC 4606), and this difference reflects the difference in mass between the galaxies. However, there is also variety in the shapes of the stellar rotation curves for a given galaxy mass. Whereas some galaxies, such as NGC 4064, NGC 4580, and NGC 4651, display a monotonically rising stellar rotation curve over the entire array, others, such as NGC 4429, NGC 4450, and NGC 4457, exhibit a maximum within the inner 10'' indicating high central mass concentrations.

Zoom In Zoom Out Reset image size

Stellar velocity dispersion profiles display behaviors as diverse as the stellar rotation curves. Some galaxies have high amplitudes in their central velocity dispersion, e.g., NGC 4429 with σ ∼ 200 km s⁻¹. Others, such as NGC 4424 and NGC 4694, exhibit very small amplitudes (σ ∼ 50 km s⁻¹). These differences are mostly due to galaxy mass. Galaxies such as NGC 4569, NGC 4651, and NGC 4450 exhibit centrally peaked σ profiles that decrease outward, but other galaxies, such as NGC 4698, NGC 4694, and NGC 4606, have essentially flat velocity dispersion profiles.

Figure 10 shows the correlation between the absolute H magnitude, derived using apparent H magnitudes (Gavazzi et al. 1999) and our stellar-kinematics-based distances (Cortés et al. 2008), and the maximum stellar velocity rotation, and the central stellar velocity dispersion. As expected from the Tully–Fisher and Faber-Jackson relations, the variations in V_max and σ among the sample galaxies are due predominantly to galaxy mass.

Zoom In Zoom Out Reset image size

Finally, the V/σ ratios show that some galaxies (e.g., NGC 4651, NGC 4569, or NGC 4580) are largely supported by rotation (V/σ ⩾ 1 at R ⩾ 0.10 R₂₅, where R₂₅ is the radius where the surface brightness has a value of 25 mag arcsec⁻¹), but others such as NGC 4424, NGC 4429, and NGC 4698 have V/σ < 1 over the entire array, and are therefore supported by random motions as far as we measure.

While V and σ are fundamental measures of galaxy support, there are important kinematic features of galaxies that are not captured by the radial variations of V and σ. These include differences between photometric and kinematic axes, changes in kinematic position angle with radius (Figure 11), differences in stellar and gas kinematics, and kinematically distinct components. For example, the presence of kinematic misalignments in galaxies are often used to assess the frequency of triaxiality by measuring differences between the photometric and the stellar kinematics major axes, or to determine the occurrence of accretion events by measuring the misalignment between the kinematics of the stellar and gaseous components (e.g., Falcón–Barroso et al. 2006). Changes in position angle may indicate a bar (Cortés et al. 2006).

Zoom In Zoom Out Reset image size

Many things can cause differences in stellar and gas kinematics. Some of these are normal features of galaxies that do not directly indicate any galaxy interaction, such as nuclear outflows, bar or spiral arm streaming motions, and a different balance of support between rotational and random motions. In this paper, we are more interested in those things which indicate some type of interaction, such as gas lying in a tilted plane with respect to the stars due to accretion events, or ram pressure disturbing the gas but not the stars.

Kinematically distinct components are of interest since they can indicate a past merger or accretion event. Four of our sample galaxies display signatures of kinematically distinct components, with the most remarkable cases in NGC 4429 and NGC 4698.

5.1. Stellar Kinematics Misalignments

Both photometric and kinematic position angles can vary with radius. These radial changes provide important information concerning the structure and history of the galaxy and need to be considered when comparing photometric and kinematic position angles. In our sample, there are several galaxies that exhibit significant radial changes in photometric or kinematic position angles. For example, in NGC 4293 and NGC 4569, there are photometric position angle changes from the inner to the outer galaxy associated with apparent warping of the outer disks. To handle this complexity, we will compare the stellar kinematic position angles with photometric PAs measured at more than one radius. In NGC 4064 and NGC 4450, the stellar kinematic PA varies with radius (Figure 11) due to the effect of stellar bars whose relative importance varies with radius. We will discuss these cases in the text.

We define the inner galaxy stellar kinematics position angle PA_kin as the mean of the kinematic position angles ϕ (see Section 3.3), within the inner 25'', i.e.,

$$\begin{equation} {\rm PA_{\rm kin}} =\left< \phi \right>. \end{equation} \tag{ 4 }$$

To help deal with the fact that in some of our galaxies the outer galaxy is disturbed or tilted with respect to the inner galaxy, we compare the inner galaxy stellar kinematics PA (SKPA) with the photometric PA measured at two different radii, an outer radius of 1.0R₂₅, PA_outer and an intermediate radius of 0.5R₂₅, PA_inter. For the photometric position angle determinations, we use photometry from our own optical images (J. R. Cortés et al. in preparation).

The kinematical misalignments Ψ_outer and Ψ_inter are defined as (Cappellari et al. 2007)

$$\begin{equation} \Psi _{\rm outer} = |{\rm PA}_{\rm kin} - {\rm PA}_{\rm outer}|, \end{equation} \tag{ 5 }$$

$$\begin{equation} \Psi _{\rm inter} = |{\rm PA}_{\rm kin} - {\rm PA}_{\rm inter}|, \end{equation} \tag{ 6 }$$

and are given for all of the galaxies in our sample in Table 4.

For six sample galaxies, both Ψ_outer and Ψ_inter ⩽ 7° and the degree of misalignment is small. For these galaxies the inner stellar kinematics exhibit rotational motions consistent with stellar bodies flattened by rotation. This set includes several galaxies that are nonetheless clearly disturbed: NGC 4424, NGC 4606, and NGC 4694.

The other seven sample galaxies have PA differences of ⩾10° for Ψ_outer, Ψ_inter, or both. These galaxies can be grouped into those with bars or suspected bars (NGC 4064, NGC 4450, and NGC 4293), those with irregular or disturbed stellar velocity fields (NGC 4351 and NGC 4569), and those with photometric features not aligned with most of the galaxy (NGC 4429 and NGC 4651).

In both NGC 4064 and NGC 4450, the stellar kinematic PA, ϕ, varies with radius (Figure 11), which we think is due to strong bar streaming motions whose magnitude changes with radius. NGC 4064 exhibits a clear difference between the kinematic PA and optical PA angle of about 50° in the inner 10'', a difference which is reduced to 10° at the end of the array (r = 22''). This produces an S-type shape in the isovelocity contours. This galaxy exhibits photometric and kinematical evidence of a bar (Cortés et al. 2006) that skews the isovelocity contours parallel to the major axis of the bar (Athanassoula 1992; Vauterin & Dejonghe 1997). Thus, it seems clear that the presence of the bar is the cause of the large differences between the kinematical and optical PAs in NGC 4064.

NGC 4450 seems similar but less extreme, probably because it has a larger bulge. While Ψ_outer and Ψ_inter differ significantly from zero, this is because the stellar kinematic position angle varies with radius and Ψ values are calculated from the mean inner stellar kinematic PA. In the central 10'', it agrees well with the photometric PA of 175° in the outer galaxy. However, at r  ∼  15–25'', the stellar kinematic PA is closer to 160°. This probably reflects a change from the inner bulge-dominated region to the bar-dominated region at larger radius.

In NGC 4293, the mean stellar kinematic PA is very different from the photometric PAs, and the reason is not clear. The outer stellar disk is tilted with respect to the inner disk, indicating a gravitational interaction, but the stellar kinematic position angle shows a very large difference of Ψ_inter = 35° with respect to the photometric PA at both intermediate and small radii (see the Appendix). There is no clear photometric evidence of a bar, although we cannot rule out a bar extended largely along the line of sight. Moreover, the stellar kinematic major and minor axes are largely perpendicular, and the velocity field does not exhibit the characteristic S-type shape which is expected in a barred galaxy, so there is also no clear kinematic evidence of a bar. A bar that is extended largely along the line of sight may be the reason for the large PA differences in NGC 4293, although other explanations are possible.

In NGC 4651, Ψ_outer = 10° but Ψ_inter is small because the outer disk is clearly disturbed, whereas the inner galaxy is much more regular and appears as a normal spiral galaxy. The outer galaxy contains a stellar tail and shells, and a disturbed outer H i distribution that appears warped (Chung et al. 2009), probably due to a minor merger. In NGC 4429, the stellar kinematic PA agrees with the photometric PA in both the inner galaxy (r ⩽ 25'') and the outer galaxy, but differs by 10° with respect to the photometric PA at intermediate radii. This is near the stellar ring at r ∼80'', which seems tilted with respect to the rest of the galaxy.

In NGC 4569, Ψ_inter = 13° ± 6°, which is moderately large. Whereas Ψ_outer is smaller, this is due to the outer disk being tilted with respect to the inner disk, and it does not explain the apparent difference between stellar and photometric PAs in the inner galaxy. The inner stellar velocity field is somewhat irregular, possibly due to dust. The velocity field on the SE side of the major axis suggests a stellar kinematic PA closer to the photometric PA of the inner galaxy, whereas the NW side, which appears dustier, shows a greater difference.

In NGC 4351, the stellar velocity field is irregular and there are large values of Ψ. Both features may reflect disturbed central stellar kinematics. The outer isovelocity contours are curved toward the SW, suggesting disturbed stellar motions. While the stellar kinematic data on this fainter galaxy are somewhat noisy, similar features are observed in the gas isovelocity contours, suggesting that they may be real.

5.2. Discrepancies between Stellar and Ionized Gas Kinematics

5.3. Kinematically Distinct Components

6.1. Velocity Dispersion Profiles

The stellar velocity dispersion profiles in stellar disks typically have exponential profiles with scale lengths about twice the scale length of the light from the disks (Bottema 1993). In our case, fitting an exponential profile could be misleading since our data are restricted to a region where the bulge is important. Instead, we fit a linear profile to our velocity dispersion profiles σ(R) (Table 6; Figure 8) defined as

$$\begin{equation} \sigma (R) = \sigma _{0} + \alpha R, \end{equation} \tag{ 7 }$$

where σ₀ is the central velocity dispersion and α is the slope of the velocity dispersion profile. We found that our galaxies display profiles with slopes ranging from roughly -36 km s⁻¹ kpc⁻¹ to 6 km s⁻¹ kpc⁻¹. The galaxies NGC 4293, 4450, 4569, and 4651 have σ profiles with slopes steeper than −16 km s⁻¹ kpc⁻¹, but others have very small slopes with essentially flat profiles; these galaxies are NGC 4064, NGC 4351, NGC 4424, NGC 4606, NGC 4694 and NGC 4698. In Figure 12 (top panel), we show the concentration parameter C₃₀ (related to the bulge-disk ratio) with α. We do not see a clear correlation with galaxies having steep or flat velocity dispersion profiles independent of how conspicuous their bulge are.

Zoom In Zoom Out Reset image size

Table 6. Parameters Derived from the Stellar Velocity Dispersion Radial Profiles

Name	σ0	α
	(km s−1)	(km s−1 kpc−1)
(1)	(2)	(3)
NGC 4064	61 ± 1	−4.6 ± 2.3
NGC 4293	94 ± 4	−16.0 ± 5.8
NGC 4351	24 ± 2	−1.0 ± 1.0
NGC 4424	47 ± 3	−2.7 ± 2.7
NGC 4429	203 ± 4	−12.7 ± 5.1
NGC 4450	139 ± 2	−36.0 ± 4.9
NGC 4457	99 ± 2	−10.3 ± 2.6
NGC 4569	118 ± 2	−25.4 ± 2.9
NGC 4580	58 ± 3	−5.6 ± 2.8
NGC 4606	47 ± 3	−3.1 ± 3.1
NGC 4651	112 ± 2	−22.3 ± 1.5
NGC 4694	42 ± 2	6.2 ± 3.1
NGC 4698	143 ± 3	−3.4 ± 3.4

Notes. (1) Galaxy name; (2) central velocity dispersion from a linear model for the dispersion profile; (3) velocity dispersion slope in km s⁻¹ kpc⁻¹.

Download table as:  ASCIITypeset image

In the group of low α galaxies, NGC 4064, NGC 4424, NGC 4606, and NGC 4694 belong to the group of galaxies with "truncated/compact" profilesof Hα (Koopmann & Kenney 2004; Section 2). NGC 4351 has been classified as "truncated/normal" but the low amplitude and flat dispersion profile could be due to the fact that the signal-to-noise ratio of the spectra in this galaxy is not enough to obtain reliable velocity dispersion measurements. Finally, NGC 4698 is a big anemic Sa galaxy with a clear secondary kinematic component. These objects also have the smallest V/σ values within the central 20'' within the sample (Figure 9). What is the cause of these flat stellar velocity dispersion profiles?

One possible cause could be that we are measuring σ, even for the outermost observed position within the bulge-dominated region. We checked this using the available near-infrared photometry and light profile decomposition for NGC 4606, NGC 4694, and NGC 4698 (Gavazzi et al. 1996), and our own R-band light profile decomposition for NGC 4064 and NGC 4424 (Cortés et al. 2006). We found that for NGC 4064, NGC 4424, and NGC 4606, the bulge is much smaller than the size of the array (12'', 5'', and 9'', respectively), so the flat σ extends farther than the bulge-dominated region. These are galaxies with less conspicuous bulges (low C₃₀, see Figure 12) and flat σ profiles. In the cases of NGC 4694 and NGC 4698, the extension of the bulge is 16'' and 18'', respectively, so it could be that our measurements are still in the bulge-dominated region in these galaxies. In fact, these galaxies have C₃₀ > 0.5, so their bulge is conspicuous.

Another possible explanation for the flat stellar velocity dispersion profiles and low V/σ values is some kind of gravitational interaction. Mergers and tidal interactions tend to heat galaxy disks and increase the velocity dispersion (e.g., Bendo & Barnes 2000). Mergers simulations (Bendo & Barnes 2000) show that direct encounters of disk galaxies with mass ratios of 3:1 can produce remnants with flat velocity dispersion profiles, slowly rising stellar rotation curves, and V/σ < 1 in the inner 20% of the optical radii. NGC 4424 and NGC 4698 have remarkably low V/σ ratios at radii where the disk starts to be important (0.3 and 0.5 respectively). In the case of NGC 4698, the existence of an orthogonally rotating core clearly indicates that this galaxy is the result of a merger. The optical morphology of NGC 4424 has features that suggest an intermediate-mass ratio merger (Kenney et al. 1996, Cortés et al. 2006). The fact that V/σ is remarkably low and flat is consistent with the merger scenario. With its small bulge, disk-like morphology, and elliptical-like kinematics, NGC 4424 is likely an intermediate mass-ratio merger (Bournaud et al. 2004; Cortés et al. 2006).

NGC 4606 and NGC 4694 exhibit non-elliptical isophotes (J. R. Cortés et al., in preparation), slowly rising rotation curves, and low V/σ ratios that resemble the kinematics found in the Bendo & Barnes prolate 3:1 merger remnants. On the other hand, NGC 4606 has a close companion NGC 4607 (d ∼ 17 kpc, and Δv = 593 km s⁻¹), which could disturb this galaxy and produce a flat σ profile. The two galaxies have similar luminosities and masses. NGC 4607 looks less disturbed than NGC 4606, which might be inconsistent with a strong tidal interaction between these two galaxies, although if the interaction is prograde for one galaxy (e.g., NGC 4606) and retrograde for the other (e.g., NGC 4607), then it is possible to have one companion significantly more disturbed than the other. In any case, NGC 4606 has clearly experienced a recent gravitational disturbance.

Finally, the case of NGC 4694 is less compelling since the bulge is about the size of the DensePak array, and the V/σ ratio becomes ∼1 at end of the bulge-dominated region. This galaxy has a disturbed H i tail connecting with a close dwarf companion VCC 2062, indicating a clear ongoing gravitational interaction (van Driel & van Woerden 1989; Chung et al. 2007; Chung et al. 2009). The velocity dispersion actually increases slightly with increasing radius, a trend not observed in any other galaxy. We speculate that this could be the result of recent star formation in the center of the galaxy, which creates a low-dispersion circumnuclear disk.

6.2. Kinematical Support

In order to understand the kinematical support of sample galaxies, we made use of the λ_R parameter (Emsellem et al. 2007), which is a practical way to quantify the global velocity structure of galaxies using the 2D spatial information provided by the integral-field units and for quantifying the specific angular momentum. This dimensionless parameter can be measured via 2D spectroscopy (Emsellem et al. 2007) as

$$\begin{equation} \lambda _{R} = \frac{\sum _{i=1}^{N} F_{i}R_{i}|V_{i}|}{\sum _{i=1}^{N}F_{i}R_{i}\sqrt{V_{i}^{2} + \sigma _{i}^{2}}}, \end{equation} \tag{ 8 }$$

where N is the number of fibers in the DensePak array, F_i is the flux inside the ith DensePak fiber, R_i its distance to the center, and V_i and σ_i are the corresponding mean stellar velocity and velocity dispersion. The parameter λ_R is close to unity when the mean rotation (V) dominates. On the contrary, λ_R goes to zero if the mean velocity dispersion (σ) dominates.

The values of λ_R for sample galaxies are displayed in Table 7. Following the kinematical classification introduced by Emsellem et al. (2007) and refined by the ATLAS^3D project (Emsellem et al. 2011), all of our galaxies can be classified as fast rotators with their location in the λ_R— diagram (Figures 13 and 14), above the so-called "green line" (Figure 14) defined as λ_R = 0.31$\sqrt{\epsilon }$ (Emsellem et al. 2011), as expected for disk galaxies. In fact, all of our galaxies show nearly regular and symmetric velocity fields consistent with being fast rotators. NGC 4698 has the smallest λ_R = 0.26, and the highest corresponds to NGC 4651 with λ_R = 0.71. Also, most of our sample galaxies are located below the location expected for an isotropic rotator (Figure 14, black line), so it is clear that they cannot be considered simple isotropic oblate rotators as expected.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Table 7. Measured and Modeled Parameters: $\sqrt{\langle V^{2}/\sigma ^{2}\rangle }$, λ_R, and Ellipticity

Name	$\sqrt{\langle V^{2}/\sigma ^{2}\rangle }$	λR		R1/2	λR(iso)	$\lambda _{R_{\rm 1/2}}{ \rm (iso)}$	k	$\lambda _{R_{\rm 1/2}}$	〈〉1/2
(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)
NGC 4064	0.61 ± 0.05	0.56	0.69	37	0.69	0.75	0.81	0.61	0.63
NGC 4293	0.47 ± 0.08	0.44	0.39	67	0.76	0.91	0.58	0.53	0.66
NGC 4424	0.37 ± 0.09	0.36	0.61	39	0.79	0.82	0.46	0.37	0.66
NGC 4429	0.41 ± 0.02	0.38	0.41	61	0.63	0.85	0.60	0.51	0.56
NGC 4450	0.35 ± 0.03	0.40	0.26	53	0.72	0.82	0.56	0.46	0.35
NGC 4457	0.31 ± 0.08	0.28	0.20	26	0.61	0.62	0.46	0.29	0.16
NGC 4569	0.43 ± 0.03	0.44	0.36	95	0.66	0.89	0.66	0.59	0.59
NGC 4580	0.76 ± 0.11	0.62	0.29	29	0.70	0.73	0.89	0.65	0.30
NGC 4606	0.74 ± 0.08	0.52	0.48	34	0.63	0.67	0.83	0.56	0.57
NGC 4651	0.94 ± 0.04	0.71	0.34	36	0.60	0.68	1.18	0.80	0.36
NGC 4694	0.68 ± 0.08	0.49	0.54	27	0.90	0.90	0.55	0.49	0.48
NGC 4698	0.28 ± 0.03	0.26	0.07	⋅⋅⋅	⋅⋅⋅	⋅⋅⋅	⋅⋅⋅	⋅⋅⋅	⋅⋅⋅

Notes. (1) Galaxy name; (2) luminosity-weighted V/σ; (3) λ_R measured over the entire array (25''); (4) ellipticity estimated within the inner 25'' from 2MASS H-band images (Jarrett et al. 2003); (5) half-light radius in arcseconds; (6) λ_Riso over the inner 25'' derived from isotropic two-integral dynamical models (Cortés et al. 2008); (7) $\lambda _{R_{\rm 1/2} iso}$ over the inner R_1/2 derived from isotropic two-integral dynamical models (Cortés et al. 2008); (8) k, ratio between observed λ_R measured over the inner 25'' to λ_R within the inner 25'' derived from isotropic two-integral dynamical models; (9) λ_R estimated in the inner R_1/2 from two-integral dynamical models corrected by isotropy; (10) luminosity-weighted ellipticity within the inner R_1/2.

Download table as:  ASCIITypeset image

In Figure 12 (middle panel), we display λ_R versus the concentration parameter C₃₀, which is a measure of the bulge-to-disk ratio. We see that galaxies with low λ_R tend to have a high C₃₀ value, although the sample is too small to derive further conclusions. In the case of the relationship between λ_R and the slope of the velocity dispersion profile α (Figure 12, bottom panel), we see that galaxies classified as [T/A], [T/N], and [N] can span a wide range of α independent of the value of λ_R. In the case of the [T/C] galaxies, they have flat velocity dispersion profiles in spite of their amount of rotational support. The possibility of this being caused by 3:1 or higher mass-ratio mergers is explored in the following section.

N-body (e.g., Bendo & Barnes 2000; Jesseit et al. 2007) and hydrodynamic simulations (e.g., Kronberger et al. 2007; Kronberger et al. 2008) have started to produce 2D kinematics that can be compared with the observed velocity and velocity dispersion fields of galaxies. In order to search for clues about what physical processes have acted on these galaxies, and which ones are driving galaxy evolution, we compare our observed kinematics with the results of simulations of gravitational interactions (i.e., mergers and tidal encounters) and ICM–ISM stripping.

7.1. Gravitational Interactions

7.2. ICM–ISM Stripping

The kinematic peculiarities of the sample galaxies are difficult to explain by internal mechanisms, and are presumably due to interactions within the cluster or circumcluster environment. Can these peculiarities be caused by a sole mechanism, or do we need more than one mechanism acting independently to explain them? In order to answer these questions, we focus on the probable scenarios for the galaxy peculiarities suggested by the stellar and ionized gas kinematics for all the sample galaxies.

In the Appendix, we present a detailed discussion on the kinematics of each sample galaxy, and in Table 8 we summarize the kinematic and morphological indicators of interactions for each galaxy and give our best guess for the types of interactions that the galaxies have experienced. All galaxies show evidence for ongoing, recent, or past interactions, and several galaxies show evidence for multiple interactions.

Two galaxies in our sample, NGC 4351 and NGC 4457, show possible evidence of ongoing ram pressure stripping based on our kinematic data. One galaxy that has been proposed as an example of past ram pressure stripping, NGC 4580 (Koopmann & Kenney 2004, Crowl et al. 2008), is confirmed in our kinematic data to be undisturbed and have a kinematically cold stellar disk, inconsistent with any past strong tidal interaction and perfectly consistent with the past ram pressure stripping scenario. All but one of our galaxies, NGC 4651, are very gas poor (Chung et al. 2009) and may have experienced gas stripping. Several of these galaxies show clear or possible evidence for ongoing or past stripping based on H i mapping (Chung et al. 2009), but not from our data.

Among the most intriguing objects in our sample are those four galaxies with truncated/compact Hα morphologies. All have flat stellar velocity dispersion profiles. While NGC 4064, NGC 4424, and NGC 4606 have small bulges, all have much lower angular momentum parameters λ_R than NGC 4580, a small-bulge galaxy with similar mass, indicating that their stellar disks have been heated. The only one of the four with a prominent bulge, NGC 4694, has a central dip in its velocity dispersion, suggesting a young circumnuclear stellar disk. These four galaxies have likely experienced gravitational interactions that have disturbed the stellar disks and driven gas to the central kiloparsec. NGC 4424, with the strongest kinematic and morphological disturbances, likely experienced a recent intermediate-mass ratio merger (Kenney et al. 1996; Cortés et al. 2006) or a close high-velocity collision with another galaxy. NGC 4694, with a complex H i distribution including a tail connecting to a nearby dwarf, clearly has experienced galaxy-galaxy tidal interaction, probably involving accretion or a "partial merger", that is, the nearby dwarf was disrupted and part of it merged with NGC 4694. The detailed type of gravitational interaction is not as clear in NGC 4606 and NGC 4064, but their kinematic properties suggest intermediate-mass ratio mergers or close tidal encounters. These galaxies have probably also experienced ICM-ISM stripping since they are gas-poor, have very truncated H i and Hα disks, and in the case of NGC 4424 an H i tail (Chung et al. 2007; Chung et al. 2009).

Several of the sample galaxies have kinematically distinct cores resulting from gas infall followed by star formation, and are likely signatures of accretion or mergers. NGC 4698, with its orthogonally rotating bulge, is clearly an old merger. NGC 4429, with its misaligned central stellar and dust disk, and the tail in its LOSVD, is likely an old merger due to the lack of gas observed in this galaxy, which probably was consumed in the subsequent star formation that formed the kinematically distinct core. NGC 4450, with the tail in its LOSVD and the kinematical misalignment between its central gas and stars, has likely experienced a recent minor merger. NGC 4694, with its central sigma drop and evidence for a young central stellar disk, plus the clearly disturbed H i, has likely experienced accretion or a partial merger with the dwarf galaxy it is currently interacting with. While it does not have a kinematically distinct core, the disturbed outer disk of NGC 4651 clearly indicates a recent minor merger, and the prominent tail in its stellar LOSVD at r = 10''–20'' suggests that the merger affected the inner galaxy. The old mergers could have happened before the galaxies entered the central part of the cluster. Among the galaxies with evidence for recent minor mergers, NGC 4651 is far behind the main cluster, and NGC 4694 and NGC 4450 have at 3D distances beyond, but not far from, the virial radius of the cluster.

Non-merging tidal encounters are likely responsible for the misaligned inner and outer disks in NGC 4569 and NGC 4293, as well as their central kinematic and morphological PA offsets. Fast tidal encounters are expected to be common in clusters, and may be responsible for some of the other kinematic peculiarities we have observed. Some features can be produced by either mergers or non-merging tidal encounters, and their origin can be hard to pinpoint without a fuller picture of the galaxies properties.

The fact that most (10/13) of our sample galaxies have signatures of some kind of gravitational interaction shows that they play important roles in driving cluster galaxy evolution. At least half of our sample galaxies show evidence for ICM-ISM stripping, a fraction that underestimates the general importance of gas stripping in clusters, since we chose mostly galaxies that exhibited features not consistent with simple ICM-ISM stripping. Thus, a significant fraction of cluster galaxies experience both gravitational interactions and ICM-ISM stripping.

In this paper, we have studied the stellar and ionized gas kinematics of 13 bright peculiar Virgo Cluster galaxies to investigate the mechanisms responsible of their peculiarities. Our results are summarized as follows.

• 1.  
  We present 2D maps of the stellar velocity V, and the stellar velocity dispersion σ and the ionized gas velocity (Hβ and/or [O iii]) of the central 30'' × 45'' for galaxies in the sample. We show maps of the higher-order moments of the LOSVD h₃ and h₄ for the three galaxies with the highest signal-to-noise ratio. The stellar rotation curves and velocity dispersion radial profiles are determined for 13 galaxies, and the ionized gas rotation curves are determined for 6 galaxies. We show deviations from circular gas and stellar motions by subtracting models of circular motions from the data. We measure the kinematic position angles for the gas and stars, and compute the differences between the optical and kinematical major axes. We provide improved position angle and inclination values for some galaxies.
• 2.  
  The stellar velocity fields in the centers of our sample galaxies are largely consistent with a rotational pattern, except for the strongly barred NGC 4064, and NGC 4698 which has an orthogonally rotating core.
• 3.  
  NGC 4064 exhibits an S-type pattern in its stellar velocity field, and big differences between the optical and kinematical PA within the inner 10'' due to strong non-circular motions caused by a central bar. The bar in NGC 4064 is unusually strong. It may be the strongest stellar bar observed to date in terms of deviations from circular motions for the stars. Large misalignments between the optical and kinematical major axes are also found in the stellar velocity field of NGC 4293, however, the presence of a bar in this galaxy is not clear.
• 4.  
  Nine galaxies exhibit non-circular gas motions and/or systematic differences between the stellar and ionized gas kinematics. The causes of these anomalies are varied. There is a velocity gradient along the minor axis caused by a nuclear outflow in NGC 4569. There are large differences between the kinematic PA of the gas and stars in NGC 4698 and NGC 4450, even though both gas and stars have predominantly circular motions. We propose that this is caused by the gravitational interaction of the gas with the potential of the disk and orthogonally rotating bulge in NGC 4698, and recent gas accretion or a minor merger in NGC 4450. Two relatively face-on galaxies in our sample, NGC 4351 and NGC 4457, show possible evidence of ongoing ram pressure stripping based on distinctive curvatures in gas isovelocity contours consistent with ram pressure stripping.
• 5.  
  Stellar velocity dispersion radial profiles exhibit a range of behaviors. Some galaxies have sharply decaying profiles, including NGC 4569, NGC 4651, and NGC 4450. Some have roughly flat profiles, such as NGC 4424, NGC 4606, NGC 4064, and NGC 4580. Others have more complex profiles with maximum values outside the center, including NGC 4698, NGC 4429, and NGC 4694. These galaxies have kinematically distinct cores. There is not a one-to-one correlation with a bulge-to-disk ratio, as there are galaxies with both large and small bulge-to-disk ratios that have relatively flat profiles.
• 6.  
  The V/σ ratios are bigger than unity where the exponential disk starts to dominate the light in most but not all of the galaxies. Galaxies such as NGC 4569, NGC 4580, and NGC 4651 are largely supported by rotation (V/σ ⩾ 1 at 0.05 R₂₅), but others such as NGC 4424, NGC 4429, and NGC 4698 have V/σ < 1 over the entire array, so they are supported by random motions as far out as we can measure.
• 7.  
  Signatures of kinematically distinct stellar components have been found in several galaxies. Pinched isovelocity contours and analysis of the higher moments of the LOSVD (h₃, and h₄) reveal cold circumnuclear stellar disks in NGC 4429, and maybe in NGC 4450. In the case of NGC 4429, this disk has a mass one-fourth the mass of the bulge, and is misaligned with the outer disk, suggesting a merger origin. NGC 4694 has a central σ drop and evidence for a young central stellar disk that is probably related to an ongoing gravitational interaction. The stellar velocity field in NGC 4698 clearly reveals the existence of an orthogonally decoupled core (Bertola et al. 1999), undoubtedly due to an old merger. While it does not have a kinematically distinct core, the disturbed outer disk of NGC 4651 clearly indicates a recent minor merger, and the prominent tail in its stellar LOSVD at r = 10–20'' and an anticorrelation between V–h₃ suggests that the merger has affected the inner galaxy.
• 8.  
  We have found no evidence of stellar counter-rotating components at least as big as 20% of the primary component. This indicates that disk galaxies with massive counter-rotating components are rare, and are only created under very special conditions.
• 9.  
  We have computed for sample galaxies the angular momentum parameter λ_R, which describes the relative amounts of rotational and random stellar motions. Only two of our galaxies are consistent with being strongly dominated by rotation. Random stellar motions are dynamically important in most of our galaxies, indicating the importance of minor-to-intermediate-mass ratio mergers and gravitational interactions in establishing their dynamical states.
• 10.  
  Several galaxies with small bulges and truncated/compact Hα morphologies have low ratios of V/σ, flat radial σ profiles, and values of λ_R which indicate that their stars have been dynamically heated by gravitational interactions. This may include 3:1 or higher mass-ratio mergers, which are not as disruptive as 1:1 mergers, allowing the galaxies to retain their fast rotator characteristics and disk-like morphology but puffing up the disks. While ram pressure stripping may be partly responsible for the truncated gas disks in these truncated/compact galaxies, their properties cannot be fully explained by simple ram pressure stripping, and require gravitational interactions.
• 11.  
  Comparison with simulations reveals that most (10/13) of our sample galaxies have signatures of some kind of gravitational interaction, showing that they play important roles in driving cluster galaxy evolution. At least half of the galaxies show evidence for ICM-ISM stripping, and several galaxies exhibit evidence for both. This implies that a significant fraction of cluster galaxies experience both gravitational interactions and ICM-ISM stripping.

The funding for this research has been provided by Fundación Andes, Chile; Fondap project grant 15010003, Chile. This work is part of the research on the Virgo Cluster and isolated spiral galaxies funded by NSF grant AST-0071251.

In this section, we comment on the velocity and velocity dispersion fields for each galaxy, as well as some relevant background information from other sources. We also explain the peculiarities, Hα type components, and features within the central r ∼ 25'' region that we have mapped.

NGC 4064. NGC 4064 has a strong stellar bar that turns into very open spiral arms at larger radii. Star formation is very strong in the central ∼1 kpc but virtually absent at larger radii. Disturbed dust lanes extend to ∼3 kpc (Cortés et al. 2006). These features suggest some kind of gravitational interaction, plus a process that removes the outer disk gas.

The central stellar isovelocity contours indicate rotation plus non-circular motions. The S-type twisting in the isovelocity contours (Figure 6) is consistent with the pattern expected for a strong bar (Figure 2). Due to the strong bar-like streaming motions, the mean stellar kinematic major axis (PA = 129°; PA∼ 100° for the central r = 10'') is very different from the photometric major axis of the inner galaxy, which reflects the bar (PA = 170°), and the photometric major axis of the outer galaxy (PA = 150°), which reflects the disk.

The stellar velocity dispersion field exhibits an off-center peak of 66 km s⁻¹, and an asymmetric shape with the line-widths greater on the NW side than the SE side. We think that this feature is real, although it is not clear whether it is a telltale signature of a minor merger or an uninteresting consequence of patchy dust extinction. Overall, the radial velocity dispersion profile is flat, which suggests some type of gravitational interaction.

The ionized gas emission is confined to the inner 10 arcsec and forms a bar-like structure (Figure 6) with star formation regions (traced by the Hβ emission) along the major axis of the bar. The ionized gas (Hβ and [O iii]λ 5007 emission) exhibits kinematic behavior roughly similar to the stars with barlike streaming motions (Figure 6). Strong CO emission in the central 10'' shows even stronger barlike streaming motions (Cortés et al. 2006).

A more complete analysis can be found in Cortés et al. (2006). Its truncated Hα and H i disks (Chung et al. 2009) suggest ICM–ISM stripping. Its location in the outskirts of the cluster (d_M87 = 2.9 mpc; Cortés et al. 2008) and recent quenching of the star-formation (just 425 myr; Crowl & Kenney 2008) suggest that this galaxy experienced the combined effect of a gravitational interaction and stripping at the outskirts of the cluster.

NGC 4293. The outer disk of NGC 4293 is tilted with respect to the inner disk, a clear signature of a gravitational interaction. Optical images shows two bright spiral arms ending at r = 80'' in a ring-like feature with PA = 75°. Beyond this, there is a sharp drop in surface brightness accompanied by a shift in PA that reaches PA = 66 ° at r = 150''.

The central stellar velocity field of NGC 4293 seems to be largely consistent with rotation. The perturbations in the isovelocity contours are likely due to the low signal-to-noise ratio of the spectra. However, the mean stellar kinematic position angle (PA = 42°) is very different from the inner disk optical PA (Ψ ∼ 35 °), and also different from the outer optical position angle (Ψ ∼ 24°; Table 4). Analysis of the Spitzer 3.6 μm image shows regular isophote shapes with PA = 77° in the central r = 20'' bulge-dominated region, so the inner stellar distribution is aligned with the rest of the galaxy inside r = 80''. Moreover, there is no clear photometric evidence of a bar, although we cannot rule out a bar extending largely along the line of sight. The stellar kinematic major and minor axes are largely perpendicular, and so there is also no clear kinematic evidence of a bar, although a map with higher S/N ratio will provide a better test of this. We have no other explanation for the large difference in kinematic and photometric PAs.

The stellar velocity dispersion shows a general increase toward the center. Not all of the structure apparent in the σ map is real. Strong dust lanes near the center may affect the line shapes in some locations.

[O iii] emission (Figure 6) was only detected in two regions separated by 10''. The scarcity of ionized gas emission made it impossible to obtain a reliable extended velocity field.

The misalignment between the kinematical and optical major axes, the disturbed dust lanes, and tilted outer stellar disk all strongly suggest a gravitational interaction.

NGC 4351. NGC 4351 is lopsided. The continuum light in the central ∼30'', where there is strong star formation with an irregular pattern, is significantly offset from the centroid of outer disk isophotes. Otherwise, there are no obvious disturbances in the light distribution of the outer galaxy, which lacks star formation. It has been unclear whether the peculiarities of this galaxy are due to ram pressure or a gravitational interaction.

The surface brightness of the stellar light in the center of NGC 4351 is fainter than most other sample galaxies, and so the signal-to-noise ratio of the spectra is lower (∼20 per pixel in the center).

The stellar velocity field is generally consistent with a rotation pattern (Figure 6), except that all of the isovelocity contours curve toward the SW. While not all of the features that deviate from a pure rotation pattern are real, the unusual curvature toward the SW may be real. A similar pattern is observed with high significance in both ionized gas velocity fields (Figure 6) as well as the H i velocity field (Chung et al. 2009). It remains unclear whether this apparently similar disturbance to the stellar and gas velocity fields is due to a tidal interaction or ram pressure. While ram pressure does not affect stars, much of the stellar light we are measuring in this region could be from young stars that formed in gas disturbed by ram pressure. The galaxy's large systemic velocity with respect to the cluster, and the large projected ICM density at its cluster location, suggest the action of ICM-ISM stripping.

The stellar velocity field displays a mean kinematical PA of 53° ± 8°, similar to the optical PA measured by us at r = 100'' and very different from the optical PA of 80° given by Koopmann et al. 2001. Ionized gas velocities (Hβ and [O iii]) exhibit a mean kinematical PA of about 62°, which is very similar to H i kinematical PA given by Chung et al. (2009).

The stellar velocity dispersion over the central field is constant within the errors at ∼25 km s⁻¹, the lowest value in the sample. It is difficult to measure dispersion variations in this galaxy due to both the low signal-to-noise ratio and the fact that the measured σ (∼25 km s⁻¹) is of the order of the velocity resolution of the spectra.

NGC 4424. NGC 4424 is a strongly disturbed and highly peculiar galaxy. It has a heart-shaped stellar light distribution at intermediate radii and shell-like features in the outer disk, clearly indicating a merger or other major gravitational encounter (Kenney et al. 1996; Cortés et al. 2006). Star formation is very strong in the central ∼1 kpc, but virtually absent at larger radii. Disturbed dust lanes extend to ∼3 kpc (Cortés et al. 2006). It has an H i tail extending beyond the main body to the south (Chung et al. 2009), suggesting either ICM–ISM stripping or a collision with M49. The precise origin of this combination of peculiar features remains unclear.

The stellar velocity field is largely consistent with a rotation pattern. The small velocity range (400 km s⁻¹ ⩽V_los ⩽ 470 km s⁻¹) within the inner 25'' shows the remarkably low velocity gradient exhibited by the stellar kinematics, which was previously observed in the ionized gas by Rubin et al. (1999).

The stellar velocity dispersion field is relatively flat but shows three modest peaks slightly offset from the major axis. One is located just north of the nucleus, and two others are nearly symmetrically located 10'' about the nucleus, roughly at the same location of the outermost luminous H ii complexes.

Ionized gas intensity line maps show that Hβ and [O iii] emission are confined to two star forming regions roughly equidistant from the center (∼10''). The ionized gas velocity fields are completely different from the stellar velocity field, with isovelocity contours roughly perpendicular to the stellar velocity field. The amount of non-circular motions in the ionized gas velocity fields made it impossible to obtain reliable gas rotation curves.

NGC 4429. NGC 4429 is an S0 galaxy notable for a very regular dust disk that extends to 10''. It has no other known peculiarity. Emission lines were not detected.

The stellar velocity field has a well-defined spider diagram, consistent with a rotation pattern (Figure 6) except for minor variations along the minor axis to the north. The isovelocity contours are pinched in the inner 5'' due to a large central velocity gradient. Two peaks symmetrically located around the nucleus at 10'' suggest a rapidly rotating second stellar component associated with the dust disk. Similar features have also been observed in other early-type disk galaxies such as NGC 4526, NGC 4570, and NGC 4621 (Emsellem et al. 2004).

The stellar velocity dispersion field exhibits two peaks offset from the center and a dip in the central part probably due to a cold central stellar disk, as in NGC 7332 (Falcón-Barroso et al. 2004) and NGC 4526 (Emsellem et al. 2004), plus the effects of dust extinction on the line profiles.

The cold stellar disk embedded with a big bulge could be the result of gas falling to the center, which can happen due to the presence of gas in a merger (Jesseit et al. 2007). Since NGC 4429 presently has no gas and no signs of any recent disturbance, we think the merger must have happened a long time ago.

NGC 4450. NGC 4450 is a large spiral with an anemic star forming disk that appears to be truncated in the outer galaxy at r = 60'' (Koopmann et al. 2001). The central region that we mapped includes a bulge and the inner part of a large-scale stellar bar with a PA ∼ 10° that extends to r ∼ 50''.

The stellar velocity field shows fairly regular kinematics, mostly consistent with a rotation pattern. A considerable pinching in the isovelocity contours in the inner 3'' indicates a large central velocity gradient. The two plateaus in the stellar velocity field symmetrically located at r ∼ 5'' are similar to the two peaks in NGC 4429 and suggest the presence of a secondary rapidly rotating component. The mean central stellar kinematic position angle differs by −12° ± 5° with respect to the outer optical PA.

The stellar velocity dispersion is centrally peaked and relatively asymmetric.

Hβ emission is very weak and is confined to only a few fibers within the array. [O iii] λ5007 emission is brighter and more extended (Figure 6). The [O iii] velocity field seems to be dominated by rotation, but with a position angle offset by 30° with respect to the central stellar velocity field, indicating significant non-circular motions or possibly a central gas disk tilted with respect to the stellar disk. If it is a tilted gas disk, then it would probably not survive long due to gravitational torques and gas dissipation, so we think this gas was recently acquired.

NGC 4457. This nearly face-on galaxy has significant H i and star formation in the central 30'', and virtually nothing beyond. The inner ISM-rich part of the galaxy is dominated by one peculiar strong spiral arm.

The stellar velocity field displays a pattern consistent with rotation. The stellar kinematic PA is 70° and probably corresponds to the line of nodes of the galaxy.

As in NGC 4429 and NGC 4450, there are pinched isovelocity contours within the inner 5'' and two peaks in the stellar velocity located at around 5''.

The stellar velocity dispersion is relatively constant over the central field, although it has a modest peak in the very center. The azimuthally averaged sigma profile shows a possible secondary peak at r = 10'', although the irregular pattern in the sigma map and the low signal-to-noise ratio make this uncertain.

The ionized gas intensity maps (Figure 6) display emission near the nucleus and on the spiral arms. The ionized gas velocity fields show a dominant rotational component, but also large non-circular motions indicating some type of disturbance. The isovelocity contours appear to be curved systematically toward the west, although our map is sparse. However, this curvature is also present in the more extended and detailed Hα velocity field presented in Chemin et al. (2006). Both maps show a strongly blueshifted region close to the core of the galaxy. The kinematic PA of the ionized gas is close to that of the stars, but the Hβ and [O iii] velocities are systematically lower than the stellar velocities by 30–40 km s⁻¹. Along the stellar kinematic minor axis, ionized gas velocities show big differences with respect to the stars, becoming systematically redshifted on both sides of the nucleus, indicating non-circular motions in the gas. This gas isovelocity pattern resembles that observed in NGC 4351. We were unable to obtain reliable rotation curves for the gas due to the amount of non-circular motions in the velocity fields.

The gas kinematics seem consistent with ongoing ICM-ISM stripping. In fact, the presence of a peculiar Hα arm could also be explained as the result of ICM wind (Schulz & Struck 2001).

NGC 4469. This large spiral has an H i and star-forming disk truncated at r ∼ 90'', and an unusual extraplanar arm of gas and star formation due to ram pressure stripping. Its outer stellar disk (r  ∼  170''–270''; PA ∼ 25°) is slightly twisted with respect to its inner disk (r ∼ 70–140''; PA ∼ 20°). It has a bright nucleus, is a known LINER (e.g., Keel 1996), and has a nuclear outflow.

The central stellar velocity field is largely consistent with a rotation pattern, although velocities on the NW side of the minor axis are systematically blueshifted by a few km s⁻¹ with respect to pure rotation, perhaps due to the effects of the prominent dust lanes present in this galaxy (Figure 2) since they are strong on the NW side. In the central 10'', there are other apparent deviations from pure rotation without any clear pattern, and these might also be caused by the effects of dust. This galaxy exhibits a small difference between the mean stellar kinematic PA (32°) and the optical photometric PA (23°)(ΔPA ∼ 9°), which could be the result of a tidal interaction.

The stellar velocity dispersion has a strong central peak and drops sharply with radius. The dropoff is steeper along the minor axis than the major axis, which is a reflection of the highly inclined viewing angle.

The ionized gas intensity maps (Figure 6) show that both Hβ and [O iii] emissions are associated with the central starburst. Hβ, but not [O iii], is also located to the NW where the star formation ring is located, and [O iii], but not Hβ, is detected south of the nucleus.

The gas velocity fields are very disturbed, indicating significant non-circular motions. Along the minor axis, the gas velocities reach a maximum value of −290 km s⁻¹ with a velocity amplitude of about 70 km s⁻¹. These are likely due to the nuclear outflow known to exist from the extended radio and Hα features along the minor axis (Chyży et al. 2006).

We could not obtain ionized gas rotation curves due to the amount of large non-circular motions in the gas.

NGC 4580. NGC 4580 has H i and Hα emission only out to a radius of 26'', where there is a prominent star formation ring. The outer disk has stellar spiral arms.

The stellar velocity field displays a nice spider diagram consistent with a rotation pattern (Figure 6), and shows no significant non-circular motions. The central stellar kinematic major axis is very close to the photometric major axis.

The stellar velocity dispersion is relatively constant over the central field, and the irregular features in the dispersion map are probably not real.

The ionized gas intensity maps exhibit Hβ emission over much of the array (Figure 6) with the strongest emission associated with the star formation ring. The [O iii] emission is poorly correlated with the Hβ emission with peaks in regions where Hβ is weak.

The Hβ velocity field shows an overall pattern of rotation with the same kinematic position angle as the stars. The gas isovelocity contours show wiggles indicating localized non-circular motions that may be associated with the star forming ring.

The undisturbed, stellar, and ionized gas kinematics indicate that there are no signatures of gravitational interaction or ongoing strong ICM-ISM stripping. The presence of a truncated Hα disk with stellar spiral arms in a gas-poor outer disk indicates that ICM-ISM stripping probably took place within the last Gyr (475 Myrs ago; Crowl & Kenney 2008).

NGC 4606. NGC 4606 has a truncated/compact Hα morphology, meaning that there is strong star formation in the central 1 kpc but none beyond. It has a disturbed stellar body with non-elliptical isophotes, indicating a strong gravitational encounter. It is an apparent pair with the spiral NGC 4607, separated by 17 kpc and 600 km s⁻¹.

The stellar velocity field is dominated by a rotation pattern (Figure 6) but some minor distortions are present. The stellar kinematical major axis PA = 39° in the inner 20'' agrees well with the PA = 44° measured from the optical photometry of the outer galaxy.

The stellar velocity dispersion is essentially constant with a value of 49 ± 3 km s⁻¹. Two apparent peaks in σ are located about 12'' north and south from the center, although the significance of these features is low. Simulations show that flat velocity dispersion profiles (Bendo & Barnes 2000) and off-axis σ peaks (Jesseit et al. 2007) can be produced in mergers with mass ratios ∼3:1.

The ionized gas emission is confined to the inner 10'' (Figure 6). The strongest emission arises from an elongated bar-like string of luminous H ii complexes, similar to those in the other truncated/compact galaxies, NGC 4064 and NGC 4424. There is a westward extension in [O iii] emission with no Hβ counterpart, along the minor axis. The Hβ velocity field exhibits a gradient along the major axis that could be consistent with rotation. Ionized gas emission is too scarce to derive further conclusions.

NGC 4651. Over most of the optical disk, NGC 4651 looks like a relatively normal spiral galaxy, but the outer parts are strongly disturbed with shell-like features and a peculiar straight tail (Chung et al. 2009), probably due to a minor merger.

The central stellar velocity field is well ordered with a pattern strongly dominated by rotation. The stellar kinematical major axis PA = 82° in the inner 20'' agrees well with the PA = 80° measured from both the optical photometry and the H i velocity field of Chung et al. (2009) at intermediate radii (r ∼ 30–90''). While to align the DensePak array we used PA = 71° from the optical photometric major axis measured at R₂₅ (Koopmann et al. 2001), this PA is characteristic of the outer galaxy (r > 100'') where the galaxy appears to be disturbed.

The stellar velocity dispersion is centrally peaked and decreases strongly with radius. A clear elongation is detected with high significance along the major axis of the galaxy. This feature is likely associated with a deviation from the Gaussian distribution in the LOSVD, which contains a strong h₃ component (see Section 5.3).

This is just inside the radius of the unusually bright, tightly wound spiral arms at r  ∼  20'', suggesting that both might be related to the minor merger.

The Hβ intensity map exhibits emission from star formation in the tightly wound spiral arms at r  ∼  20'' (Figure 6), but very weak emission from the center. The [O iii] intensity map shows strong emission over the nucleus and the spiral arms. Both ionized gas velocity fields appear largely similar to the stellar velocity field beyond the central 10'', although both Hβ and [O iii] maps show wiggles in the isovelocity contours that could be associated with spiral arms. Within the inner 10'', the [O iii] map exhibits a twisting in the isovelocity contours indicating non-circular motions in this gas.

NGC 4694. NGC 4694 has strong emission from the nuclear region due to star formation and/or an active galactic nucleus, and otherwise it has weak star formation in the central kiloparsec, and none beyond. A disturbed complex of H i extends across the minor axis to the nearby faint galaxy VCC 2062, indicating some kind of gravitational interaction, perhaps a minor merger or gas accretion event (Chung et al. 2009).

The stellar velocity field is largely consistent with a rotation pattern, although it shows some modest apparent deviations from circular motions that have low significance. Along the minor axis (PA = 56°), the stellar velocities do not display any gradient. The PA of the stellar kinematical major axis in the center agrees well with the PA measured from optical photometry in the outer disk.

The stellar velocity dispersion is practically flat with a small dip in the center. Some of the apparent variations exhibited in the outer parts are likely artifacts due to a low signal-to-noise ratio.

The ionized gas intensity maps show emission within the central 10''. The nucleus is relatively strong with fainter features extending to the NW and SW (Figure 6). The ionized gas velocity fields appear disturbed. Along the major axis, ionized gas velocities are systematically lower by 25 km s⁻¹ than the stars, and in places show large velocity gradients.

NGC 4698. This remarkable galaxy NGC 4698 has a bulge that rotates orthogonally with respect to the disk (Bertola et al. 1999). The disk appears to be undisturbed, suggesting that this galaxy is the product of an ancient merger.

The stellar velocity field shows at least two distinct kinematical components (Figure 6). The inner 10'' exhibits S-shaped isovelocity contours near the minor axis due to the orthogonally rotating core. The effect is less evident that in previously published data (Bertola et al. 1999), due to the smoothing effect introduced by the width of the fibers. In the outer 10'', the velocity field is consistent with a rotation pattern in most respects, except that the isovelocity contours are concave. These concave contours mean that the line-of-sight velocity increases as we move outward from the major axis, which is probably due to the effect of the orthogonally rotating bulge. This galaxy exhibits almost no stellar rotation within the inner 10''. Beyond the inner 10'', the LOS stellar velocities rise dramatically, due to the increasing influence of the stellar disk.

The stellar velocity dispersion has a modest peak at the center, but otherwise remains flat along the major and minor axes.

Hβ emission was not detected. On the contrary, [O iii] emission is very strong in the center and detected over the entire array (Figure 6). The [O iii] velocity field is well ordered but very different from the stellar velocity field.

The pattern seems to be largely consistent with rotation, although with a PA that changes with increasing radius from 120° at the center to 150° at 25''. (In order to produce the [O iii] rotation curve in Figure 8, we used rings with different PAs.) In the inner r ∼ 5'', the [O iii] kinematic PA is roughly perpendicular to the local stellar kinematic PA. At r  ∼  25'', the PAs are offset by ∼20°. We suspect that the gas distribution is non-planar, as the equilibrium plane of the gas is expected to gradually change from larger radii, where the outer disk dominates the gravitational potential, to smaller radii, where the orthogonally rotating bulge dominates the potential.