Since then we performed diverse investigations of various aspects of LMC 4. The age of the association NGC 1948 at the NE edge of LMC 4 (Vallenari et al. 1993; Will et al. 1996) and of the globular cluster NGC 1978 (Bomans et al. 1995) were determined, as well as of the associations LH 54, LH 60, and LH 63 at the SW edge (Petr et al. 1994). Signs for X-ray emission from gas inside of LMC 4 (Bomans et al. 1994) were found, as well as highly ionised gas (C IV) with HST (Bomans et al. 1996). The neutral gas inside LMC 4 was investigated and it was concluded that the structure could not have been formed by the collision with a high velocity cloud (Domgörgen et al. 1995). Finally, Braun et al. (1997) investigated the stars inside LMC 4, notably those of the constellation Shapley III, and found they were all of the same age, being about 10-15 Myr. In fact, the similarity of all the ages derived for star groups inside and at the edge of LMC 4 pointed toward the need to identify a large scale star formation triggering event.
ROSAT X-ray data of the LMC were analysed by Blondiau et al. (1997). They found in particular near the dark cloud south of 30 Doradus an anticorrelation between X-ray emission on the one hand and neutral hydrogen and IRAS (Schwering, 1989, A&AS 79, 105; see maps on p.142-145) dust emission on the other hand. This anticorrelation has a nature well know from studies of the galactic halo (see, e.g., Kerp 1994). Blondiau et al. noted also that this exceptional region is very close to and aligned with the steep gradient in H I near the leading edge, a region from which the X-ray intensity is lower than the general background of the X-ray sky (see map of Snowden & Petre 1994).
How does the LMC move through the halo of the Milky Way to produce such a bow shock? From attempts to model the Magellanic Clouds and the Magellanic Stream it is known (Heller & Rohlfs 1994) that the LMC moves basically towards the east. In an analysis of Hipparcos data of LMC stars, this motion was confirmed (Kroupa & Bastian 1997). Including the known radial velocity of the LMC the full space velocity is ≅265 km s-1.
The rotation of the LMC is well know from radial velocity maps of stars and of analyses of the radial velocities of H I gas. Kroupa & Bastian (1997) found signs for a clockwise rotation from the proper motion of the Hipparcos LMC stars. Combining the radial rotation curve with the inclination, or combining the radial and the tangential rotation curve, one finds that the true rotation speed of the LMC at a radius of 2 kpc is ≅150 km s-1.
Given the above data for the dynamics, the maximum velocity difference between LMC gas and gas of the Milky Way halo is ≅450 km s-1, to be found near the SE edge. Gas flowing straight to the east (due to the LMC rotation) will temporarily pile up against the halo gas and will move to the side (toward the N) due to the proper rotation of the LMC.
First, the clockwise progression of age, starting at the leading edge as given by the fields listed by de Boer et al. (1998), should continue along the edge of the LMC and populations older than 50 Myr should be found at the western and southern edge. One should look for large star groups inside larger structures.
Star formation most likely takes place now in the huge molecular cloud south of 30 Doradus. It is to be expected that mapping in the near IR of the dark cloud will uncover numerous proto stars and proto clusters.
The star formation at the leading edge of the LMC needs, however, not be a continuous process. E.g., the piling up of gas may be interrupted because a region devoid of gas (such as an old LMC 4) is moving up to the leading edge. The sequence of presently known superstructures along the east and north, such as the dark cloud, the 30 Doradus region, LMC 4, and probably the Kim et al. (1997) shell, makes clear that gaps are present indeed.
The bow shock induced star formation model predicts a clock-wise increase in the age of superstructures, in step with the full rotation velocity of ≅150 km s-1. At a radius of about 2 kpc this means one revolution in ≅100 Myr. One might therefore expect at each location along the outer edge of the LMC stellar populations with ages of modulo 100 Myr. Investigations of new fields support this prediction (see Braun 1998). Is the age of the older group in the field of SN 1987A of 400 Myr as visible in the L,Teff diagram of Romaniello et al. (1998) of significance?
Gas at the leading edge may be brushed back over the disk of the LMC, which would probably result in X-rays from the turbulent medium present on both sides of the disk near the front of the LMC. Is there a relation with the interpretation by Blondiau et al. (1997) who, from their analysis of the ROSAT data found that soft X-rays are coming from behind as well as from the front of their dark region?
The inclination of the LMC with respect to the line of sight, with a value of 22° (Kim et al. 1997) to about 30° (Westerlund 1997), implies that the vector of motion of the LMC lies in the plane of the LMC itself. Thus the LMC slices its way through the Milky Way halo as a disk (see Fig. 1). The LMC appears not to tumble. However, the true values of the inclination as well as of the direction of motion are not known with such an accuracy that a slight misalignment can be excluded.
The radial rotation curve of the LMC presented by Kim et al. (1998, see also this workshop), shows that the inner part, up to a radius of about 2 kpc, has solid body rotation. This means that the geometry of the star forming regions near the leading edge may stay the same during the further rotation, and are not disrupted by shearing. Is it significant, that the shape and orientation (in the rotating frame) of the association LH 77 inside LMC 4 is very similar to the shape and orientation of the molecular cloud south of 30 Doradus?
As was formulated by de Boer et al. (1998), the bow shock scenario does not preclude that other mechanisms of star formation would operate in the LMC. Clearly a bow shock event will produce subsequent localised star formation, such as was found by Petr et al. (1994). Subsequent star formation thus may explain the luminous structure of N 11.
Also the dynamic interaction between LMC and SMC may play a rôle. The last close encounter of LMC and SMC took place about 200 Myr ago (Heller & Rohlfs 1994), about two and a half times the time of one LMC revolution.
Following the LMC orbit backward around the Milky Way, the LMC was located 60° back in the sky, or close to the galactic south pole, about 200 Myr ago. During that period the LMC has made about 2.5 revolutions. (The Sun would have moved once around the Milky Way center.) Has the plane of the LMC changed orientation over that part of the orbit? If not, was then the stripping of gas much more vehement in the past? Would this have anything to do with the present day structure of the Magellanic Stream?
| First version: | 29th | March, | 1998 |
| Last update: | 08th | October, | 1998 |