5 Conclusions

We have investigated NGC 6231 with broad band photometry finding that this cluster is located at a distance of d = 1990 pc. The absorption in the direction of NGC 6231 is quite normal although it is seen through nearby molecular clouds. The differential reddening and the large colour excess scatter may be due to a combination of factors such as anomalies of the intracluster material, and also to the presence of several binaries and variables. An isochrone fitting to the upper main sequence stars gives a nuclear age of 3-5 Myr for this cluster.

NGC 6231 shows a bend at $V\approx 13.5$ mag followed by a scarcity of ZAMS stars for 14 < V < 16. Besides, faint stars here exhibit the peculiarity of being all above the ZAMS without following its shape. Shobbrook (1983) stated that "... significant number of type A pre-main sequence stars might be found here in a search amongst still fainter members...". We attempted to know how many of the observed faint stars are field stars seen against the cluster surface integrating the function $\psi (M_V, ST, r, \Delta r)$ over the distance.

We are aware that the use of a local luminosity function might weaken our argument but, thinking as Underhill (1960) did, it should be proven first that $\Psi (M_V,ST)$ is anomalous in this particular direction. We consider, however, that to know exactly the field star luminosity function here, a deep photometric survey around NGC 6231 should be performed. Once the number of field stars was removed from our star sample a contour map with the star excess was built. This map clearly defines a zone of PMS stars in our HR diagram where, in addition, the PMS stars found by SBL98 are perfectly placed in.

The PMS zone found in NGC 6231 suggests an age spread of about 10 Myr, incompatible with the age derived from bright (massive) members. That is, the less massive stars formed in a timescale of 10 Myr whereas the most massive stars formed in the last 3-5 Myr, which is against a coeval process.

Concerning the IMF slope, this is the fourth very young open cluster included in our series which slope is flatter than the one given by Salpeter (1955). We have previously found x = 1.1 in Tr 14, 1.04 in Pismis 20, 1.09 in NGC 5606 (Vázquez et al. 1996, 1995, 1994, respectively). A preliminary analysis of another galactic cluster in our series, NGC 3293 (Baume et al. 1999), yields that x = 1.15, similar to the young LMC cluster NGC 1962-65-66-70 (Will et al. 1995) with x = 1.0 to 1.2 or Cyg OB2, another galactic young region with x = 1.0 (Massey & Thompson 1991). Notwithstanding the slopes of galactic and magellanic young clusters average x = 1.3 (see the compilation of Will 1996) it appears that the "extremely" young ones (ages < 10⁷ yr) show flatter slopes than the field stars.

Flat slopes, however, could be produced by overestimating the masses of very hot stars or by a mass segregation effect as suggested by RCB97 for NGC 6231. Concerning the first case, up to now there is no conclusive way to know the masses of very hot stars (except for a few cases), as they are mainly obtained from the theoretical HR diagram. A way to reduce the impact of our ignorance of stellar masses is using the procedure of Burki (1977) who detected clusters having abnormal high number of massive stars through a comparative analysis of their respective LFs. In relation to the mass segregation effect, more extensive surveys are necessary, searching for coronal members in several young open clusters.

This article is partially based in the Digitized Sky Survey that was produced at the Space Telescope Science Institute under US government grant NAG W-2166. Original plate material is copyright the Royal Observatory Edinburgh and the Anglo-Australian Observatory.

Acknowledgements

The authors are indebted to Dr. Garrison for the allocation of telescope time. G. Baume made this work while a grant from the CONICET. We want to thank the useful comments made by Dr. Raboud to improve this manuscript.