Astron. Astrophys. 330, 412-418 (1998)

3. Results

3.1. Gas and stellar kinematics

Fig. 1 presents stellar and gaseous velocity fields for the central part of NGC 2685, and Fig. 2 presents stellar and gaseous velocity fields for the central part of IC 1689. In both cases one can see that the stars rotate around the minor axes as it must be in disk galaxies, and the gas rotates around the major axes as if it is coupled with the more outer gas of polar rings.

In NGC 2685 the gas and star isovelocities are strictly orthogonal, and systemic velocities determined by the gaseous and stellar velocity fields are the same, namely, 860 km/s. In IC 1689 the picture is more complex. The stellar isovelocities are consistent with a circular rotation of stars in the plane of galactic disk. A systemic velocity determined by using this velocity field is 4470 km/s. An absence of systematic wavelength-scale shift is checked by measuring a night-sky emission line [OI] : a whole set of the night-sky line-of-sight velocity measurements shows a gauss-like distribution with a mean of -3 km/s and a dispersion of 17 km/s. Particularly, the measurement of [OI] in the spectrum of the nucleus has given a value of -16 km/s. Since our estimate of the systemic velocity of the stellar component in IC 1689 differs strongly from that of Hagen-Thorn & Reshetnikov (1997), we have checked the result of cross-correlation by a direct gauss approximation of the magnesium absorption lines (5167.3+5172.7) and in the spectrum of the nucleus. Though a accuracy of such approximation is lower than that of cross-correlation, we have obtained km/s and 4471 km/s, consequently, in full agreement with the above mentioned systemic velocity derived from the cross-correlation. A systemic velocity determined by using the gas velocity field is higher by some 150 km/s - 4620 km/s. Gas isovelocities in the center of the galaxy are turned by some 45^o with respect to both isophote axes, major and minor, and only at the radii more than 5" we see an ionized gas rotating in a plane orthogonal to the plane of the galaxy: it is an inner polar ring seen also on ultraviolet (U) broad-band image (van Gorkom et al. 1987). A line-of-sight velocity difference between the ring and the nuclear gas is km/s, in agreement with the long-slit data of Hagen-Thorn & Reshetnikov (1997). Excitation mechanisms of emission-line spectra are quite different in the ring and in the nucleus: the ratio [NII] in the ring is consistent with a star-forming activity, and the nucleus lacking at all (only [NII] is present) is a typical LINER. This result was firstly obtained by Hagen-Thorn & Reshetnikov (1997) too.

A two-dimensional velocity field gives a unique opportunity to clarify a geometry of rotation and mass distribution. In the case of planar circular rotation a azimuthal dependence of central line-of-sight velocity gradients would be a pure cosine law with a maximum at the line of nodes (isophote major axis):

sin i cos ,

where is deprojected angular rotation velocity, i is an inclination of rotation plane, and is an orientation of the line of nodes (isophote major axis). In the case of triaxial potential there must be a non-zero line-of-sight velocity gradient along the isophote minor axis.

Fig. 3 presents azimuthal dependencies of central line-of-sight velocity gradients for the ionized gas and stars in NGC 2685 and IC 1689, and Table 2 contains results of fitting these dependencies by cosine laws. (Photometric characteristics are taken from Peletier & Christodoulou, 1993, for NGC 2685 and from Reshetnikov et al., 1995, for IC 1689). They confirm the impressions given by Figs. 1 and 2.

Fig. 3. The azimuthal dependencies of central line-of-sight velocity gradients for the stars and ionized gas in NGC 2685 and IC 1689. The solid lines present cosine laws fitted by the least-square method

Table 2. Parameters of fitting the azimuthal dependencies of central line-of-sight velocity gradients for the stars and gas in NGC 2685 and IC 1689

The dependencies for NGC 2685 are quite perfect cosine ones; the dynamical major axis of the stars coincides with the photometric major axis while orientations of inner and outer isophotes are the same. All these facts provide evidence for a axisymmetric mass distribution and circular rotation of the main body of the galaxy. Thus, the nuclear ionized gas also rotates circularly, but in the plane which is orthogonal to the plane of the galactic disk. As it was noted earlier (e.g. Shane 1980), the nuclear ionized gas in this galaxy must be related to the neutral and molecular hydrogen of the polar ring. The only thing which is not clear is why the angular rotation velocities of the gas and stars differ in such a way: our estimate of the gas velocity dispersion inside , 200 km/s, is higher than the published estimates of the stellar velocity dispersion, 60-114 km/s (Whitmore et al. 1990, Di Nella et al. 1995, McElroy 1995), so one might expect that nuclear gas rotates slower than stars; but we observe an opposite difference. After excluding non-circular motions and strongly different rotation plane inclinations (both stellar and gaseous subsystems seem to be seen edge-on), only different spatial resolution remains: as gas velocity gradients are measured closer to the center than those of stars, their higher values may signify that there is an unresolved drop of rotation velocity near (let us remind that the seeing was not too good during the CCD observations of NGC 2685 in the green spectral range - see Table 1). In other words, the nucleus of NGC 2685 may be dynamically decoupled.

The azimuthal dependencies for IC 1689 look less definitive. The dynamical major axis of stars is turned by 11^o with respect to the line of nodes (outer isophote orientation), and a turn of photometric major axis between the innermost and the outermost regions is also 11^o, but in opposite sense (Reshetnikov et al. 1995). These small discrepancies may be inside errors of measurements, but if they are real, it may be an evidence for a mild triaxiality of the galaxy.

3.2. Stellar contents of the nuclei

Fig. 4 presents radial profiles of absorption-line index Mgb in the central parts of NGC 2685 and IC 1689 (for a definition of the absorption-line indices - see Worthey et al. 1994). In both cases one can see a prominent magnesium-strength break between the nucleus and the surrounding bulge. If we treat this break as due only to metallicity variations, we can estimate a value of metallicity break by using single-age model calibrations of Worthey (1994). It reaches 0.7 dex for NGC 2685 and 0.5 dex for IC 1689.

Fig. 4. Azimuthally-averaged magnesium-line strength profiles for NGC 2685 (two sets of measurements) and for IC 1689

For NGC 2685 we have plotted two independent measurement sets: in 1993 with IPCS as a detector and in 1994 with CCD. (The nuclear spectrum in 1993 had a low signal-to-noise ratio and is not plotted). In this particular case a accuracy of the Mgb index measurements has appeared to be good enough even with IPCS: the difference between two data sets does not exceed 0.1 Å anywhere except the points at . The discrepancy at is naturally explained by the seeing difference (see Table 1): a seeing FWHM of 2.5" in 1994 affects the measurements at this radius. Though a radius range under consideration is rather small, we may suspect a presence of Mgb gradient in the bulge of NGC 2685. If we take the measurements in the radius range 2.5"-7" - 4 points obtained in 1993 and 3 points obtained in 1994, - they are nicely fitted by a linear law:

.

Therefore even if we take a difference between the nuclear Mgb and the central bulge Mgb extrapolated by using this linear law, we would still obtain Å (). The metallicity gradient measured in the bulge, , is slightly higher than usual metallicity gradients in bulges of early-type disk galaxies which ranges from 0 to -1 (Balcells & Peletier 1994). It is interesting that Peletier & Christodoulou (1993) have noted that the nucleus of NGC 2685 is distinguished by its red colour, and in the bulge the gradient is absent.

In IC 1689 we detect an enhancement of the magnesium-line strength in both the nucleus and at a radius of , where the polar ring crosses the major axis.

Fig. 5 may help us to check if the magnesium-strength breaks in NGC 2685 and IC 1689 are due only to the metallicity differences. It presents a () diagram which provides age-metallicity disentangling. Since emission is quite absent inside in both galaxies, we are sure that absorption line is not contaminated by an emission in the nuclei and in the innermost bulge regions, and the age diagnostics is valid. But the outer points - two in NGC 2685 and three in IC 1689 - must be excluded from the consideration due to a noticeable emission contamination of their . After examining Fig. 5 we conclude that stellar population in the nucleus of NGC 2685 is rather old, of 10 billion years or older, and a age gradient along the radius is undetectable. Meanwhile the decoupled nucleus in IC 1689 may be as young as of 5 billion years, and the innermost bulge may be much older. Unfortunately, for this galaxy we have only one point in the inner bulge which absorption line is not contaminated by an emission, so for we can estimate only upper limits of the stellar population ages, and the result on age difference between the nucleus and the inner bulge is marginal. But we must keep in mind that if the nucleus is much younger than the bulge, the metallicity difference corresponding to the same is higher. In the case under consideration it reaches 0.7 dex instead 0.5 dex reported above.

Fig. 5. The diagram () for age-metallicity disentangling in the nuclei and bulges of NGC 2685 and IC 1689. The measurements are made at the positions shown in Fig. 4; the nuclear positions are identified. The models are taken from Worthey (1994), the ages are given in the legend in billion years. Two outer points for NGC 2685 and three outer points for IC 1689 are shifted down by a weak emission

© European Southern Observatory (ESO) 1998

Online publication: January 16, 1998
helpdesk.link@springer.de