5 Conclusions

We have analyzed the large-scale 21 cm- HI emission in a $6\hbox{$^\circ$ }\times 6\hbox{$^\circ$ }$ field around the nebula NGC2359, and found a large expanding bubble roughly centered at the nebula and its exciting star, the Wolf-Rayet HD56925. The systemic velocity of the bubble is $\sim$64 km s^-1 and it is expanding at 12 km s^-1. At an assumed distance of 5 kpc, the bubble has a size of $70 \times 37$ pc and a HI mass between 700 and 2400 $M_\odot$. By simple estimates of energetics, ages and stellar wind parameters, we think that this feature is a wind-blown bubble produced by the O-progenitor of the WR star. The presence of a nearly spatially-coincident IRAS shell reinforces our hypothesis.

We have also mapped the CO J=  $1 \rightarrow 0$ and $2 \rightarrow 1$ emission adjacent to NGC2359 and found three CO components in this region. Two of these components, with radial velocities of 34 (component A1) and 67 (component A2) km s^-1, have narrow profiles (up to 2 km s^-1) and do not show any morphological or kinematical effects which indicate the disturbance by the nebula or the WR star. However, the third component (the component B, emitting at 54 km s^-1), clearly bounds NGC2359 at its southern and eastern border. The profiles are broad, with linewidths of 4-5.5 km s^-1. A velocity gradient of a few km s^-1 is noted towards the south and south-eastern interface with the nebula. A weak "bridge'' in velocity of small angular extension seems to connect this component with the component B2 at the southern part.

We have observed the ¹³CO J=  $1 \rightarrow 0$ line at a few points in the region where the component B is detected. We have found a striking variation of the ¹²CO/¹³CO ratio both in position and in velocity. This ratio seems to decrease near the peak of the CO emission, for velocities higher than 53 km s^-1. This particular emiting region is not resolved by the beam, and is probably 0.2 pc wide.

From the observational data, we have estimated the physical conditions for each CO component using the LVG approximation. The highest excitation temperature and density are found in the component B. It is remarkable that high temperatures (up to 80K) are derived for the CO emission projected onto the HII region.

In view of the kinematics, the morphology and the physical properties of the CO, we think that the component A2 was shocked and accelerated by the expanding bubble to reach the radial velocities of the component B. The shock front is still acting at the southern part of the region. The origin of such shock might be related to the WR phase of HD56925 or, more probably, to previous episodes of RSG of LBV. From the ¹²CO/¹³CO ratio we infer the presence of a thin layer of dense gas which represent the more recently shocked material. Our data, however, cannot resolve spatially the region where a thin layer of shocked material is expected. CO observations with better resolution and sensitivity is needed to disclose the main features of the region.