4 Correlations between kinematic and photometric parameters

The photometric parameters used for these correlation were taken from Burstein et al. (1987), RC3 and the ESO-Uppsala Sky Survey, kinematic quantities are results of this work.

To calculate the distances of the sample galaxies two different models were used:

1.

The distances were calculated by assuming an undisturbed Hubble-flow and H₀ = 75 km s^-1 Mpc^-1, using the system velocities taken from Faber et al. (1989).

2.

The distances were taken from the catalogue of Kraan-Korteweg (1986), where a model of the Virgo cluster infall was added to the Hubble flow.

4.1 Correlations between anisotropy and photometric parameters

Binney (1978) was the first to utilize a $(v/\sigma )$ vs. $\epsilon$-diagram in order to get a statistic reference if a group of galaxies suits into the model of an oblate rotator or whether it rather consists of dynamical hot systems, where chaotic star motion dominates.

In Fig. 1 $(v/\sigma )^*$ vs. $\epsilon$, $(v/\sigma )$ vs. $\epsilon$ and v vs. $\epsilon$ are plotted for all sample galaxies where these parameters are available. Galaxies with $(v/\sigma)^{*}>1$ are expected to be rotationally supported systems. These are (sorted by increasing $\epsilon$): NGC 1404, NGC 2872, NGC 3260, NGC 4105, NGC 7196, NGC 3557, NGC 3268, NGC 6861, NGC 2271, NGC 7049 and NGC 1537. The last four objects are S0 systems, the others are elliptical galaxies with fast rotation.

The correlation between anisotropy and luminosity was first described in Davies et al. (1983). Galaxies with low luminosity are mostly rotationally supported systems while galaxies with high luminosity are generally systems with high anisotropy. The existence of this correlation is explained by the increasing influence of chaotic relaxation and the increasing influence of merging with increasing galaxy mass (and therefore luminosity) (Davies et al. 1983; Bender et al. 1992). In Fig. 2 the correlation between $(v/\sigma )^*$ and M_B is plotted.

The galaxy NGC 3557 is represented by the data point with the greatest luminosity and doesn't follow the general trend. This E3-galaxy features high luminosity and a rotational velocity of $v_\textrm{\scriptsize rot}$ = 217 $\pm$ 11 km s^-1.

Figure 1: $(v/\sigma )^*$ vs. $\epsilon$, $(v/\sigma )$ vs. $\epsilon$and v vs. $\epsilon$ - diagrams. The dotted line represents the theoretical model of an oblate rotator

Figure 2: The correlation between anisotropy and absolute luminosity in the B filter. The distances of the galaxies needed for the calculation of MB were in the upper diagram taken from the catalog of Kraan-Korteweg (1986) and in the lower diagram derived under the assumption of an undisturbed Hubble flow. The dotted line represents the dividing line between rotational supported systems and anisotropic galaxies $(v/\sigma )^* = 0.7$ given by Bender et al. (1992)

4.2 $\kappa$-space

$\kappa$-space was introduced by Bender et al. (1992) and was originally defined to examine the physical properties of dynamically hot galaxies. The axes of $\kappa$-space are proportional to the logarithm of galaxy mass ($\kappa _1$), mass-to-light ratio ($\kappa_3$), and a third quantity that is mainly surface brightness ($\kappa _2$) (Burstein et al. 1997). The $\kappa _1$/$\kappa_3$-projection represents a side-view of the fundamental plane, the $\kappa _1$/$\kappa _2$-projection shows it nearly face-on.

All sample galaxies where the necessary input parameters $\sigma_{\rm c}$, $R_{\rm e}$and $I_{\rm e}$ were available are represented in Fig. 3. The data point of one galaxy, NGC 6861, lies within the "zone of exclusion'' (ZOE) $\kappa_1+\kappa_2>8.0$ defined by Burstein et al. (1997). All galaxies are giant sequence (Gas-Stellar Continuum) members. In the giant sequence $\kappa _2$ increases while $\kappa _1$decreases, it contains bulges, normal ellipticals (giant and intermediate) and compacts.

Figure 3: $\kappa$-space in the parameterization published by Bender et al. (1992). The distances of the galaxies needed for the calculation of $R_{\rm e}$ were in the upper diagram taken from the catalog of Kraan-Korteweg (1986) and in the lower diagram derived under the assumption of an undisturbed Hubble flow. The dotted line in the $\kappa _1$/$\kappa _2$-plot marks the beginning of the "zone of exclusion'' (ZOE), which extends to the upper right. The full line in the other projections marks the fundamental plane

4.3 Statistics

 

 

Table 6: Kinematic types of the sample galaxies following Bender et al. (1994)

galaxy	kinematic type	galaxy	kinematic type
I1729	disky:	N3302	boxy, without rotation:
N1404	decoupled central component	N3309	normal E-galaxy
N1427	decoupled central component	N3377	normal E-galaxy
N1537	normal E-galaxy	N3557B	decoupled central component
N1549	E0; boxy, with weak rotation	N3557	normal E-galaxy
N1889	unknown	N3585	decoupled central component; disky
N2271	decoupled central component; disky	N3617	decoupled central component:
N2325	not classifiable	N3260	disky
N2434	boxy, without rotation	N3636	E0; boxy, without rotation
N2380	boxy, without rotation	N3904	decoupled central component
I2311	decoupled central component; disky	N3923	not classifiable
U4508	not classifiable	N4033	decoupled central component
N2663	boxy, without rotation	N4105	decoupled central component
N2699	decoupled central component	N4106	unknown
N2887	boxy, without rotation	U7354	not classifiable
N2865	boxy, without rotation:	N4261	boxy, without rotation
N2872	unknown	N4697	normal E-galaxy
N2888	decoupled central component	N5061	E0; boxy, with weak rotation
N2986	decoupled central component	N5237	not classifiable
N3078	decoupled central component	N5903	boxy, without rotation
N3087	decoupled central component	I4797	disky
N3136	not classifiable	N6861	not classifiable
N3125	not classifiable	N7029	not classifiable
N3224	normal E-galaxy:	N7049	normal E-galaxy
N3250	decoupled central component:	N7196	decoupled central component:
N3258	boxy, with weak rotation	I5297	boxy, without rotation
N3268	disky

Rotational velocity profiles, velocity dispersion profiles, h₃ and h₄-profiles were extracted for 49 galaxies. Bender et al. (1994) identified several kinematic types by their rotation, velocity dispersion, h₃ and h₄ profiles, finding correlations between kinematic profiles and photometric properties like diskyness or boxyness. Following this description, our sample galaxies were divided into the kinematic types, the kinematic types of the individual galaxies are listed in Table 6.

galaxies with decoupled
central component kinematics	17 galaxies	= 32%
disky	7 galaxies	= 13%
boxy, without rotation	10 galaxies	= 19%
boxy, with weak rotation ...	3 galaxies	= 6%
E0	4 galaxies	= 8%
normal E-galaxies	7 galaxies	= 13%
unknown	3 galaxies	= 6%
not classifiable	9 galaxies	= 17%.

Galaxies where not enough data for this evaluation was obtainable are classified as "unknown'', galaxies with profiles which did not fit into any of the representative object classes are found in "not classifiable''. Some galaxies fit into more than one class, they are multiply counted. Disky early-type galaxies are two-component systems with a hot bulge and a cold, fast rotating disk. This superposition will cause asymmetric LOSVDs which are characterized by strong negative h₃ values. Features in the major axis rotation curve that are caused by varying light contributions of the disk are reflected by similar features in the h₃ profile. Boxy ellipticals show none or only weak rotation and positive h₄ values on their major axes. Weak rotation causes some weak asymmetry in their LOSVDs and therefore slightly negative h₃ values. E0 galaxies do show a slight increase of velocity dispersion within their photometric core radii and smaller h₄ values inside their core radii than further out. Normal E-galaxies are galaxies with h₃ and h₄ $\approx$ 0.

32% of all galaxies contain kinematically decoupled central components. The size of these central components was found to be $0.40\pm 0.28$ kpc and in all cases less than 1 kpc. The identification of decoupled central components is based on the kinematic profiles, which show features typical for disky galaxies, but h₃ has the opposite sign to v and the reversal in the rotation corresponds to the radius where also h₃ changes sign. On comparing these galaxies with results in literature, a good accordance was found, e.g. NGC 3078 and NGC 3250 where central depressions in the Mg₂ profiles were identified by Carollo et al. (1993). For two of the sample galaxies central surface-brightness profiles have been obtained by HST, NGC 3377 and NGC 4697 (Faber et al. 1997). Both have power-law profiles which lack cores. NGC 4261 and NGC 4697 do contain nuclear dust structures also indentified by HST (van Dokkum & Franx 1995).

The percentage of sample galaxies where a disk component was evident is altogether 49%, in 36% no disk component could be detected and in 15% of the sample galaxies this is uncertain. It may therefore be concluded that more than half of the sample galaxies contain stellar disk components, adding the fact that with greater spatial resolution even smaller disks might have been detected.

The luminosity distributions of individual kinematic object classes are shown in Fig. 4. Galaxies with kinematically decoupled components are evenly distributed in the luminosity range, in contrary to Bender (1996) who find that these components should be more frequent in luminous ellipticals. The dichotomy in the class of elliptical galaxies in boxy and disky objects is confirmed also from the kinematic point of view. However it is noteworthy that some disky ellipticals show the signature of a kinematically decoupled central component.

Figure 4: Absolute luminosity distributions of the sample galaxies with respect to the kinematic types introduced by Bender et al. (1994). The shaded distributions represent respective luminosity distributions of the subsamples with kinematically decoupled component, boxiness or diskiness. Note that "diskiness'' and "boxiness'' were infered from the galaxy's kinematic profiles, not photometric properties

It is an accepted fact that merging plays an important role in the evolution of early-type galaxies up to altering the original morphological type (Barnes 1996). The most notable signature of such an event are kinematically decoupled gaseous and stellar components (Bertola et al. 1990; Bender 1996). Statistics of our sample galaxies were used in order to verify this picture: ZCAT (Huchra et al. 1995) was searched for galaxies in the vicinity of each sample member using radial bins of 50, 100, 150 and 200 kpc around each galaxy and a redshift interval of $\pm 1000$ km s^-1. The resulting histogram (Fig. 5) reflects the well-known fact that generally elliptical galaxies are located in high-density environments (Dressler et al. 1994). However, it is noteworthy that the subsample of objects with decoupled components are found in groups of even higher density than average.

The peculiar central component kinematics in some ellipticals can possibly also be explained in ways different from the above scenario. E.g. the decoupled component could be due to streaming in a triaxial body, obliquely projected (Binney 1985; Franx et al. 1991; Statler 1994); but rotation amplitudes in the central components are in general too high and central metallicities are enhanced with respect to the main body, so this scenario cannot account for the formation of the majority of ellipticals with peculiar central components (Bender 1996).

Figure 5: Histograms of the number of sample galaxies with neighbours within a redshift interval of $\pm 1000$ km s-1 using ZCAT. The histograms range from a radial distance between sample galaxies and neighbours from 0<r<50 kpc (uppermost diagram) to 150<r<200 kpc (lowest diagram). The histograms for sample galaxies with kinematically decoupled components are shown as shaded. The last $N_\textrm{\scriptsize neighbour}$ bin includes all galaxies with more than 5 nearby galaxies. A trend that galaxies with kinematically decoupled components are more likely to be located in crowded environments (where $N_\textrm{\scriptsize neighbours}$ is higher) is evident