The derived RC shows a rapid rise followed by a Keplerian like decline shortly after, then levelling out at a constant velocity of 30 km s-1 for radii larger than five arcseconds. Between approximately one and two kpc the RC falls slightly faster than the Keplerian prediction, but this is within the uncertainties. The receding and approaching sides are in reasonable agreement.
A close look a the line profiles in the centre revealed double lines which made us attempt a decomposition of the velocity field (see Fig. 2). This yielded a regular secondary component in the central part only. After decomposing the velocity field the RC looks very different (see Fig. 2). The central high velocity part in Fig. 1 is replaced by a secondary counter rotating component spinning at high velocity. To get good agreement between the receding and approaching sides of the secondary component, its systemic velocity was adjusted. Note that the RC of the secondary component is based on a total of nine pixels only. The centre of the primary component nearly coincides with that of the total non-decomposed velocity field. The primary component has a nearly constant RC with a velocity of 40 km s-1. This differs from the non-decomposed value of 30 km s-1 because it is based on fitted, i.e. smoothed, data, and slightly different (1 pixel) centre coordinates. Moreover, since fitting Gaussian profiles to the data requires higher S/N, the radial extent of this RC is somewhat smaller than for the non-decomposed velocity field. A word of caution is necessary here: the velocity difference between the primary and secondary components is close to the FSR of the used FP2, meaning that the lines almost overlap. In effect the relative velocity of the secondary component with respect to the primary, is uncertain. Therefore we cannot exclude that the counter rotating component is an artifact. However there is a clear signature of asymmetric line profiles in the centre and some sort of multicomponent gas is needed. Observations with a FP with higher FSR (e.g. FP1) should reveal if the second component represents a counter rotating disc or not. At this stage we urge the reader to view the RC of the second component in Fig. 2 as one possible interpretation. If our interpretation is correct, the mass of the secondary component is on the order of .
This galaxy has the greatest H luminosity in the sample and the ionised gas mass may be as great as , comparable to the dynamical mass estimate in Table 4.
The H emission is concentrated to two, or perhaps three, central H II regions. In between the H II regions the isovelocity contours are squeezed. Double and very broad components are observed in the emission line profiles, especially west of the centre (see Figs. 3 and 4). Two components have been extracted with quite different intensity levels. If we do not decompose the profiles and simply compute the velocity from the total H line, the first component clearly dominates in the brightest part, but in fainter regions west of the centre, unreasonable features appear like the confusing situation with three different symmetry axes. Anyway, this would not affect the RC of the primary component since the regions with double lines are close to the minor axis. For the primary component, the velocity field shows a strong gradient along the major axis; but with a plateau in the north-west region. This coincides with the western H II region and roughly with the region where the double features in the lines are most pronounced, and thus close to where the second component has its maximum intensity. The kinematical centre is well defined along the major axis and coincides roughly with the continuum peak intensity. Due to the plateau, the kinematical centre is less well determined along the minor axis, however different choices of centre along this axis gives consistent RCs. In essence the RC does not sensitively depend on the centre coordinates or the decomposition model.
The RC derived for the main component shows solid body rotation out to a radius of kpc, after which it levels out, although it is here based on the receding side only. The RC has a rather large dispersion indicating that the physical situation might be more complex than the assumed disk anatomy. For the second component, we could not obtain any sensible RC. However, the duality of the velocity peaks in the western part of the galaxy are very significant, and some sort of multicomponent gas is needed to explain the data. The ionised gas mass is of the order , thus significantly smaller than the dynamical mass estimate which is of the order .
The H emission is concentrated to an extended bright central starburst region. In addition there is diffuse H emission extending in a tail towards the east. There are also suggestions of a small H arm emanating towards the south from the western side of the starburst. The velocity field is very irregular and does not contain a single axis of symmetry. Moreover the velocity gradient is steep in the western parts and roughly east-west orientated; while in the eastern regions the gradient is much lower, not always positive and lacking a well defined position angle. The eastern extended tail has almost no velocity gradient.
East of the centre, just at the border of the starburst region, we observe what appears to be the superposition of two patterns with different orientation (see Fig. 5). We will refer to the hypothetical component east of the centre as the perpendicular component, since its PA is roughly perpendicular to the major axis of the galaxy. However, we do not observe any double component in the profiles (in either FP2 or FP1 data). Anyway, we tried to decompose the velocity field without any conclusive results. This could mean that if there are two components present, the linewidth is too large with respect to their velocity separation or/and where the components overlap, one only sees the one with the strongest Hemission.
In Fig. 7 we show a map of the velocity dispersion as derived from the FWHM linewidth of the FP1 data. The velocity dispersion has a fairly constant level of km s-1. Where the major axis of the perpendicular component cross the major axis for the whole galaxy the velocity dispersion is higher and peaks at 160 km s-1. Still, the shape of the H line is consistent with one single broad component.
It is not unambiguous how to derive a RC for this galaxy. The different RCs are however consistent in the way that they all are very irregular, signifying a non equilibrium system. The general feature is an approaching side with continuous steeply rising velocity, and a receding side with very small velocity gradient. This means that the assumption of a regularly rotating disc in equilibrium must be far from reality in this galaxy. In effect the kinematical centre is not well defined.
As a cure we tried to mask away certain points of the velocity field (those in the eastern arm and the approaching side of the perpendicular component) to check if we could obtain a more regular RC. The result is shown in Fig. 6a, but the RC is still far from regular. Moreover, the decline of the average RC outside 2 kpc is faster than the Keplerian case, which is not necessarily significant since the velocity field is so perturbed.
For the hypothetical perpendicular component we could extract a regular RC (Fig. 6b) only when using a narrow sector (). The implied rotational velocity is very small, but the regularity and the agreement between the approaching and receding sides still makes this component well defined.
The RC obtained from the FP1 data (Fig. 7) largely agrees with that from FP2 data (Fig. 5), but the rotational velocity of the receding side is km s-1 larger for the FP1 data. The differences can be attributed to three effects: Firstly, the different spectral resolution of FP1 and FP2. Secondly, the slightly different choice of centre coordinates (an independent determination was made for the FP1 data). Thirdly, the FP1 data is somewhat deeper and reach larger radii, and since the velocity field is deprojected these outer points enter also at smaller radii. In view of this, the agreement is very good.
A last note on this galaxy is that here it is questionable that the expression "rotation curve" is adequate since there appears to be no regular rotation present. Still, the presented RCs provide important information on the complexity of the system. However, the dynamical mass estimate in Table 4 should be viewed sceptically. The ionised gas mass is of the order .
This galaxy is a physical companion of ESO 338-IG04 and is located approximately six arcminutes south-west of it, corresponding to a projected distance of kpc. The H image contains at least four bright H II regions (Fig. 8). The galaxy has an irregular, somewhat "bent" shape in H but the velocity field is still quite regular. Along the north-west side there are twists in the isovelocity contours which may indicate the presence of a spiral arm. Just north-east of the brightest H II region there is a steep velocity gradient which we interpret as the kinematical centre, and which coincides with the centre of broad band images. The RC shows good agreement between the two sides and a gradual flattening with increasing radius. The estimated dynamical mass is and the ionised gas mass is of the order of .
On broad band CCD images, about north-west of the target galaxy, there is a small low surface brightness galaxy, which is not detected in the monochromatic H images.
Double components with rather different intensity have been extracted from the original lines. Both components have similar position angles but the gradients are reversed, i.e. the secondary component is counter-rotating with respect to the primary (Figs. 9 and 10). The brightest component is very similar to the total non-decomposed velocity field (compare the upper left and right panels in Fig. 9). The second component is more diffuse and regular than the first one. While the centre of the secondary component coincides with the continuum peak, the kinematical centre of the primary component is slightly offset from it ( to the north).
The RC of the primary component shows a rapid increase followed by an almost flat slowly rising part. The agreement between the approaching and receding sides is good. The secondary component also yields a regular RC, with a low rotational velocity though. The regularity of the RC of the secondary component suggests that it is caused by a dynamically well defined object, e.g. a counter rotating disc. The estimated dynamical mass of ESO 185-IG13 is and the ionised gas mass is of the order of . The dynamical mass of the secondary component is
The RC has a strange behaviour. Following the initial solid body rise to the plateau's, there is a rapid decline after which the RC levels out and stays flat out to a radius of 15 arcsec, after which the approaching side (the only one that has H signal) declines further. The remarkable thing is that after the maximum rotational velocity the decline is faster than for the Keplerian case. This is unphysical for an equilibrium disc, thus indicating that the velocity field is distorted. The "super-Keplerian'' decline can be eased somewhat, but not completely, by allowing the position angle to be variable. In the south west regions at a radius greater than double features in the H lines appear at low S/N. However, this feature is present in many consecutive pixels and is in total significant, which means that a secondary component might be present here. In the centre there are no signs of double lines and our attempts to decompose the velocity field failed. Due to the super-Keplerian behaviour of the RC the estimated mass apparently decreases with increasing radius (see Table 4). Thus the mass stated in Table 4 is very uncertain. The ionised gas mass is of the order of a few times , comparable to the dynamic estimate.
Copyright The European Southern Observatory (ESO)