In this section, we discuss the kinematics and morphology of the individual targets. In addition, the velocity fields are thoroughly described in Paper I. For all galaxies we show isovelocity contours for the H $\alpha$ emitting gas, overlaid on broad band images. The labels of the isovelocity contours are given in Paper I.

$\begin{figure} \par\includegraphics[width=12.6cm,clip]{ms1027_fig5.eps}\end{figure}$

Figure 5: Mass model for ESO400-G43, ESO400-G43B, Tololo0341-407E and Tololo0341-407W. For further explanations see caption of Fig. 3. For ESO400-G43 no disk and halo component are shown since it was impossible to construct a dynamical mass model.

$\begin{figure} {\resizebox{12.8cm}{!}{\includegraphics{ms1027_fig6a.ps}} } {\resizebox{12.8cm}{!}{\includegraphics{ms1027_fig6b.ps}} } \end{figure}$

Figure 6: The H $\alpha$ iso-velocity contours overlayed on an R-band CCD image of ESO350-IG38. The velocities to which the contours correspond can be found in Paper I. North is up, and east to the left. The size of the field is 1 by 1 arcmin, corresponding to $23 \times 23$ kpc. Note the irregular morphology at all isophotal levels. The faintest visible structures are $\mu _R \approx 26$ mag/arcsec².

This galaxy has a heart shaped morphology, with three bright starburst nuclei (Fig. 6). The outer morphology is distorted out to very faint isophotal levels (see Fig. 6 where the faintest visible structures are $\mu _R \approx 26$ mag/arcsec²). This is seen more clearly in broad bands that in H $\alpha$ demonstrating that the light originates mainly in stars. Hence, the large scale distribution of stars in this galaxy is highly asymmetric. At all isophotal levels an extension in the south-eastern direction is evident. This may be a tidal tail in development, or the remnants of a such. H $\alpha$ images reveal faint arms extending from the south-eastern and south-western nucleii. An HST/WFPC2 archive image reveals that the three very bright starburst nuclei at the center of ESO350-IG38 are composed of numerous individual super-star clusters with luminosities up to M_V = -15. The total number of luminous star clusters is very high (Östlin 2000). Emission from the south-east nucleus is dominated by a resolved very high surface brightness "nucleus'' with M_V = -17.5 with no apparent internal structure. The properties of these central regions bear a close resemblance to the classical colliding galaxies NGC 4038/4039 as seen in HST data (Whitmore & Schweizer 1995).

The H $\alpha$ line profiles of this galaxy are broad, up to FWHM=270 kms^-1, and have a non-Gaussian shape. This suggests that two or more non-virialised components may be present. In the centre, double peaked lines are present consistent with the presence of a counter-rotating disk (see Fig. 3 and Paper I) or high velocity blobs. These properties indicate that the centre is not dynamically relaxed, while the outer velocity field shows a very slow rotation. The estimated stellar mass density exceeds by far what can be supported by the observed amount of rotation ( $\approx$ 30 kms^-1). Thus, the galaxy is either not in equilibrium, or it is not primarily supported by rotation. These properties strongly suggest that the starburst was triggered by a merger process.

This galaxy is the most massive in the whole sample and also has the highest star formation rate. The inferred (core-collapse) supernova frequency is one every 7 years. Surprisingly, it seems to be rather devoid of cool gas (Bergvall et al. 2000), suggesting that the starburst is about to run out of fuel.

$\begin{figure} {\resizebox{10.4cm}{!}{\includegraphics{ms1027_fig7a.ps}} } {\resizebox{10.4cm}{!}{\includegraphics{ms1027_fig7b.ps}} } \end{figure}$

Figure 7: The H $\alpha$ iso-velocity contours of the primary dynamical component in ESO480-IG12 (see Paper I) overlayed on an R-band CCD image. North is up, and east to the left. The size of the field is 1.7 by 2.5 arcmin, corresponding to $30 \times 45$ kpc.

4.2 ESO480-IG12

An intriguing finding with this galaxy is that it has strong outer morphological distortions (Fig. 7) and that it is apparently aligned with a chain of galaxies in a background (uncatalogued) cluster. Spectra have been obtained for three of the galaxies with the smallest angular distance from ESO480-IG12, which however are at much higher red-shift ( $\sim$ 30000 kms^-1), but it cannot be excluded that one of the galaxies apparently belonging to the cluster is in fact a low mass companion of ESO480-IG12.

The northern extension is long and narrow while the south one is broader and more diffuse. South of the centre, there are several kiloparsec scale plumes. In addition, there are several small faint blobs in the southern part. High resolution images would reveal if these are extended or perhaps compact star clusters, like those seen in ESO338-IG04, ESO350-IG38 and ESO185-IG13. The overall morphology presents large-scale asymetries down to the faintest visible levels, compare e.g. the north and south-west extensions. This indicates an asymetric and non-equilibrium distribution of stars, which may be due to an interaction/merger or a strong warp.

The velocity field is irregular with double line profiles, almost over the entire galaxy, and two components have successfully been fitted in Paper I. The double lines could arise in a separate dynamical component or an outflow or expanding super bubble. The rotation curve and luminosity distribution suggest that dark matter dominates the dynamics (see Fig. 3).

$\begin{figure} {\resizebox{12.8cm}{!}{\includegraphics{ms1027_fig8a.ps}} } {\resizebox{12.8cm}{!}{\includegraphics{ms1027_fig8b.ps}} } \par\end{figure}$

Figure 8: The H $\alpha$ iso-velocity contours of ESO338-IG04 overlayed on a R-band CCD image. North is up, and east to the left. The size of the field is $60\hbox {$^{\prime \prime }$ }\times 53\hbox {$^{\prime \prime }$ }$ , corresponding to $11 \times 9.5$ kpc.

4.3 ESO338-IG04 (Tololo 1924-416)

This well-known BCG has an almost chaotic velocity field, with strong gradients and an extended tail with little internal velocity structure (Fig. 8). It has a companion at a projected distance of 70 kpc to the south-west (see next subsection). Approximately 10 $^{\prime \prime }$ east of the centre, just at the border of the central star-forming region, there is a velocity component whose kinematical axis is perpendicular to the photometric major axis of the galaxy (see Paper I). The radial light distribution indicates that the observed mean rotational velocity (if a such is at all meaningful to define in view of the irregular velocity field, see Fig. 3) cannot support the system gravitationally.

There is a 5 kpc long tail towards the east, and large scale isophotal assymetries down to the 26 mag/arcsec² level. At fainter levels the morphology becomes more regular, but a box-like shape remains on the western side. The tail has much bluer colours than the rest of the galaxy outside the starburst region, signifying a younger stellar population. It contains stellar clusters, and cannot be explained by a purely gaseous tail. The tail has almost no velocity gradient with respect to the centre ( $\Delta v \le 10$ kms^-1). On the other hand, the western half of the galaxy has a strong gradient and an implied rotational velocity of 80 kms^-1at a distance of 3 kpc from the centre (Fig. 5 in Paper I). This is identical to the rotational velocity in the companion ESO338-IG04B which has an equal photometric mass. Hence, the western part of the galaxy shows about the expected level of velocity difference with respect to the centre for rotational support to be possible. But what is happening at the eastern side in the tail? The colour of the tail suggest that it has a distinctly different stellar population from the rest of the galaxy. The most likely is that the tail is a remnant of a merger and that projection effects prevent us from seeing the true velocity amplitude. The parts of the galaxy on the eastern side which are not in the tail, do not emit in H $\alpha$ , hence we have no information on its kinematics. Where the tail meets the starburst region we see an increased H $\alpha$ velocity dispersion, perhaps due to a shock. This coincides with the location of the perpendicular dynamical component discussed above. The companion is probably too far away for tidal forces to have caused the starburst and peculiar velocity field.

Radio interferometric observations (Östlin et al. in preparation) reveal that the galaxy is embedded in a very large H I cloud, more than 7 armin across (corresponding to 80 kpc at the distance of ESO338-IG04). The H I cloud has irregular morphology with two main components and no single axis of rotation. ESO338-IG04 appears to be located in the eastern H I cloud. The morphology and velocity field of the H I complex is consistent with a close interaction/merger of two gas-rich galaxies or H I clouds. The companion (see next subsection) is detected in H I but lies further away.

Although we cannot exclude that the starburst in ESO338-IG04 is triggerred by interaction with the companion, a merger appears more likely in view of the complex velocity field and the non-rotating arm. HST observations of this well known starburst has revealed that in addition to many young compact star clusters, it contains a system of intermediate age ( $\sim$ 2 Gyr) globular clusters (Östlin et al. 1998), a fossil of a previous dramatic starburst event.

$\begin{figure} {\resizebox{12.8cm}{!}{\includegraphics{ms1027_fig9a.ps}} } {\resizebox{12.8cm}{!}{\includegraphics{ms1027_fig9b.ps}} } \end{figure}$

Figure 9: The H $\alpha$ iso-velocity contours of ESO338-IG04B overlayed on an V-band CCD image. North is up, and east to the left. The size of the field is $83\hbox {$^{\prime \prime }$ }\times 83\hbox {$^{\prime \prime }$ }$ , corresponding to $15 \times 15$ kpc. Note the low surface brightness galaxy north-west of ESO338-IG04B.

4.4 ESO338-IG04B (companion)

$\begin{figure} {\resizebox{12.8cm}{!}{\includegraphics{ms1027_fig10a.ps}} } {\resizebox{12.8cm}{!}{\includegraphics{ms1027_fig10b.ps}} } \end{figure}$

Figure 10: The H $\alpha$ iso-velocity contours of the primary dynamical component of ESO185-IG13 overlayed on a V-band CCD image. North is up, and east to the left. The size of the field is $50\hbox {$^{\prime \prime }$ }\times 50\hbox {$^{\prime \prime }$ }$ corresponding to $18\times 18$ kpc.

4.5 ESO185-IG13

ESO185-IG13 presents a clear example of classical morphological perturbations produced in galaxy mergers and/or collisions (Barnes & Hernquist 1992; Schweizer 1998). The broad band images of this galaxy (see Fig. 10) clearly reveal the presence of a tidal tail 30 $\hbox{$^{\prime\prime}$ }$ long in the north- east direction (about 10 kpc at the distance of the galaxy main body).

On CCD images, there are two small faint galaxies with fairly low surface brightness approximately 1.5 arcmin south-west of ESO185-IG13. None of these were detected in the Fabry-Perot data, hence their distances are unknown. An HST image of ESO185-IG13 (obtained from the archive) reveals the presence of numerous compact star clusters with luminosities up to $M_V \approx -15$ (Östlin 2000), a typical signature of colliding system (see Whitmore et al. 1993, 1995). Most of the bright cluster sources are concentrated to a central bar like structure. Both the HST and ground based images reveal the presence of arms in the center, though with no counterpart in the Fabry-Perot H $\alpha$ image or velocity field.

The analysis of the velocity field of ESO185-IG13 revealed the presence of two dynamically distinct components (Paper I). The secondary component is counter-rotating with respect to the main component (which spins faster and dominates the H $\alpha$ emission). The two components have marginally different position angles. If the two components represent gas discs that intersect, this configuration cannot be long lived (the rotation period of the secondary component is $\approx$ $4 \times 10^8$ years). Despite a regular rotation curve, the observed level of rotation cannot support the observed photometric mass (see Fig. 4). However, given the limited surface photometry available, and the lack of near-IR data, the estimated M/L for the disk component is rather uncertain. If the IMF has a flat low mass part (see Sect. 2.4) and if we have underestimated the amount of rotation (e.g. due to projection effects) and take into account the sum of the two components (as in Fig. 2), it might be possible to solve the mass discrepancy. In summary, this is a clear example of a BCG where the starburst has been triggered by a merger.

$\begin{figure} {\resizebox{12.8cm}{!}{\includegraphics{ms1027_fig11a.ps}} } {\resizebox{12.8cm}{!}{\includegraphics{ms1027_fig11b.ps}} } \end{figure}$

Figure 11: The H $\alpha$ iso-velocity of contours ESO400-G43 overlayed on an R-band CCD image. The grey scale have been chosen to show detail rather than depth. North is up, and east to the left. The size of the field is $50\hbox {$^{\prime \prime }$ }\times 50\hbox {$^{\prime \prime }$ }$ , corresponding to $19\times 19$ kpc. The source 15 $^{\prime \prime }$ north of the centre is an H $\alpha$ emitting region at the red-shift of ESO400-G43.

4.6 ESO400-G43

This galaxy has a regular outer morphology, embedded in a large H I cloud that is extended towards a companion galaxy with a projected distance of 70 kpc (Bergvall & Jörsäter 1988, see next subsection). At fainter isophotal levels peculiar morphological features start to appear. Approximately 15 $^{\prime \prime }$ north of the centre there is a bright H $\alpha$ emitting blob, apparently not rotating with the galaxy at the pattern speed expected at its location.

The rotation curve shows a rapid increase followed by a dramatic decline, that drops faster than the keplerian prediction (see Fig. 5 and Paper I). This is obviously unphysical for an equilibrium disc. A possible interpretation is that the cause of the super-keplerian rotation speed is infall towards the centre. A B-R image of ESO400-G43 shows a shell-like structure with radius $\sim$ 1.5 to 3 kpc. This may be a sign of superbubbles or stellar resonances, like the "shells'' found in many merger remnants. In the north-west, starting 10 $\hbox{$^{\prime\prime}$ }$ from center, there are signs of an arm/plume both in the velocity field and broad band images (Fig. 11). Our kinematical data for the ionised gas is in good agreement with the results by Bergvall & Jörsäter (1988). The published H I kinematical data show a rotation curve with a rotational velocity of $\sim$ 50 kms^-1 (Bergvall & Jörsäter 1988), i.e. similar to the central value in the H $\alpha$ rotation curve before the super keplerian decline. Bergvall & Jörsäter (1988) suggested that this galaxy was a genuinely young galaxy and that the super-keplerian drop was due to a non-relaxed disk. The latter conclusion remains likely, although it is evident that the galaxy contains an old underlying population (Bergvall & Östlin 2001). It is likely that a rapid infall of gas towards the center is fuel for the starburst but - what gave rise to the infall? The most efficient infall mechanisms are bar instabilities and mergers, but there are no signs of a bar in ESO400-G43. The northern bright H II region might be an infalling blob, or something that has just plunged through the H I disc. Slightly more than an arcminute to the north-east there is a positive detection of a massive H I cloud, which may be interacting with ESO400-G43 (Bergvall & Jörsäter 1988).

The low rotational velocity outside the centre cannot, by far, support the stellar mass gravitationally (Figs. 2 and 5). The observed H $\alpha$ line widths (see Table 4) could be sufficient for gravitational support if the linewidth traces virial motions. However, this can not explain the observed shape of the rotation curve. Clearly, this galaxy is badly perturbed dynamically. The only likely explanation is interaction with the companion ESO400-G43B or that a merger with a smaller galaxy has occurred.

$\begin{figure} {\resizebox{12.8cm}{!}{\includegraphics{ms1027_fig12a.ps}} } {\resizebox{12.8cm}{!}{\includegraphics{ms1027_fig12b.ps}} } \end{figure}$

Figure 12: The H $\alpha$ iso-velocity contours of ESO400-G43B overlayed on an R-band CCD image. North is up, and east to the left. The size of the field is 1 by 1 arcmin, corresponding to $22 \times 22$ kpc.

4.7 ESO400-G43B (companion)

This physical companion to ESO400-G43 (Bergvall & Jörsäter 1988) is a regular dwarf galaxy with a regular rotation curve (Figs. 5 and 12). The shape of the rotation curve does not follow the light distribution which indicates the presence of a dark matter halo (Fig. 5). The properties of ESO400-G43B are fairly similar to those of ESO338-IG04B, but it has more intense star formation activity. The time scales for gas consumption and buildup of the observed stellar mass are both on the order of 3-4 Gyr. Thus, this galaxy is at the limit of being classified as a starburst galaxy. The star formation occurs in a central extended HII region. It is quite probable that interaction with ESO400-G43 increased the star formation activity.

$\begin{figure} {\resizebox{12.8cm}{!}{\includegraphics{ms1027_fig13a.ps}} } {\resizebox{12.8cm}{!}{\includegraphics{ms1027_fig13b.ps}} } \end{figure}$

Figure 13: The H $\alpha$ iso-velocity contours of Tololo0341-407 overlayed on an R-band CCD image. North is up, and east to the left. The size of the field is $37\hbox {$^{\prime \prime }$ }\times 37\hbox {$^{\prime \prime }$ }$ , corresponding to $11 \times 11$ kpc.

4.8 Tololo0341-407

This galaxy is the faintest BCG in the sample. It has two apparent nucleii embedded in an irregular envelope (Fig. 13). The velocity field shows the two nuclei to belong to distinct dynamical systems. In Paper I we successfully decomposed this galaxy into two dynamical components coinciding with the optical components on the broad band image. This may be two dwarf galaxies currently coming together. The eastern component (Tololo0341-407E) has a very high SFR per unit mass, whereas Tololo0341-407W is similar to the more luminous BCGs in this respect. Notice that both galaxies are far less massive than the other galaxies in the sample.

This is the only galaxy in the sample for which we do not have multi-colour CCD photometry, hence we could not determine M/L. As an estimate for M/L we took the median for the other galaxies, and for $M/L_{\rm disk}^{\rm min}$ and $M/L_{\rm disk}^{\rm max}$ we used the extreme values of the distribution for the other galaxies. Hence, the mass estimates for this system are more uncertain than for the other galaxies. Nevertheless, the maximum disk solutions are close to the predicted photometric rotation curves, and in the Eastern component we see marginal evidence for a dark matter halo (Fig. 5).

4 Discussion of individual objects