5 Summary and conclusions

We have analyzed the morphological and kinematic structure of the nebula around the LBV candidate S 119.

We find that S 119 is surrounded by a nebula that represents a spherically expanding shell. On the western rim of this shell, a massive outflow occurs with a much higher velocity than the expanding shell. If one assumes a scenario, in which LBV nebulae are due to wind-wind interactions, the LBVN should be filled with hotter gas, since LBVs-unless in outburst-are hot stars with fast stellar winds. S 119 is a hot star, classified as Ofpe/WN9. The lack of detectable X-rays can therefore be explained by either (a) a relatively old outflow in which the originally hot gas had already sufficient time to cool, or (b) an LBV nebula which was not filled with hot gas in the first place, or (c) a velocity of the outflow which is too small to cause shocks with a sufficiently high post-shock temperature, or - of course - a combination of two of these reasons.

HST images are characterized by the spherical overall appearance of this nebula. The edge of the spherical component is much better defined in the northern and eastern directions where we also find the largest surface brightness. It does not exhibit a smooth distribution of surface brightness, but rather is very patchy and shows many filaments and knots, some of which extend - in particular in the western directions - beyond the nebula's main body and occasionally are detached from the main body. This is the reason for the less well defined edge of the nebula in this direction.

High-resolution long-slit echelle spectra show that the spherical component of the nebula is expanding with about 25 km s^-1 and that the center of expansion is at a radial velocity of 156 km s^-1. This is remarkable as it is considerably lower than that of the LMC to which-due to its projected position on the sky-the star seems to belong. Both numbers are in good agreement with earlier results (e.g., Nota et al. 1994).

In all spectra a high velocity component is present in addition to the spherical expansion. It corresponds to a relative radial motion of up to $\sim$100 km s^-1 faster than the spherical main body of the nebula. This high velocity material is concentrated on the western side of the nebula. It is worth noting that on this side the nebula also seems to be somewhat frayed between the filaments NW-2 and S-1 (see Sect. 3). As yet this makes S 119 the only LBV (candidate) nebula with such an outflow. The radial velocities in the outflow increase linearly with distance from the nebula. This Hubble-type velocity law reminds one of the strings found in and around the Homunculus nebula of $\eta$ Car (Weis et al. 1999). However, due to the rather different morphological structure, projection effects (which were ruled out in the $\eta$ Car strings) may play a role in the radial velocity structure in the S 119 outflow. One may, for instance, think of a conical structure with an opening angle that increases outwards. Due to our current lack of understanding of the physical nature of such Hubble type velocity laws, it is not possible to draw firm conclusions on the relation between the two phenomena. The largest radial velocity, relative to the star, amounts to 127 km s^-1 at a projected distance of $6\hbox{$.\!\!^{\prime\prime}$ }9$ from the star, corresponding to $\sim$1.7 pc. One may deduce a minimum dynamical age of $\sim$ $1.3\times 10^4~$yrs. The dynamical age of the shell assuming a radius of 4 $.\!\!^{\prime\prime}$5 and an expansion velocity of 25.5 km s^-1  amounts than to $\sim$ $4.2\times 10^4~$yrs. One has to be aware, however, that it depends strongly on the formation mechanism of the outflow how meaningful a dynamical age is. In addition it was shown that the radial velocity of the system seems to deviate from that of the LMC so that the distance to S 119 has to be questioned. This distance on the other hand severely affects the determination of the dynamical age. If S 119 were only 30 000 pc away the dynamical age of the nebula would already go down to $\sim$ $2.5\times 10^4~$yrs. That would than be comparable to the lifetime of the star as an LBV. The dynamical age and the stars position in the HRD, that is being a hot star makes it quite likely that S 119 is already on its way to leave the LBV phase as does He 3-519 (Davidson et al. 1993).

The brightness difference between the west side and the east side (Nota et al. 1994, see also Sect. 3) can most likely be accounted for by the outflow. However, one can only speculate whether the outflow is due to a density gradient in the ambient medium or whether it is caused by asymmetric flows in the S 119 system. Given the sphericity of the nebula's main component we are inclined to hold environment effects responsible for the outflow. This is also consistent with the brightest part of the nebula occurring in the same direction in which the HST image shows ionized diffuse gas which may easily indicate a higher density in that direction.

Acknowledgements

We have made use of the ROSAT Data Archive of the Max-Planck-Institut für extraterrestrische Physik (MPE) at Garching, Germany. Part of the work was carried out on a workstation provided by the Alfried Krupp von Bohlen und Halbach-Stiftung to the ITA. This support is gratefully acknowledged.