8 Conclusions

The main results of our XMM-Newton observation are that in Omega Cen the binaries do not seem be confined to the core or even within the half mass radius, as previously thought, and that the majority of the faint X-ray sources may be CVs, RS CVn or BY Dra binaries. Obviously, the small number of identifications currently available prevents us from reaching definite conclusions. The spectral and timing information provided by XMM-Newton, together with the accurate positions provided by Chandra for a wide sample of faint X-ray sources should encourage follow-up investigations at other wavelengths (optical, UV, radio). In particular, these observations should tell us whether those sources found well outside the core of $\omega$ Cen really belong to the cluster. More identifications are also required before reliable comparisons between $\omega$ Cen and other clusters with different structural parameters can be drawn. Such comparisons are critical to a better understanding of the dynamical evolution of globular clusters in general.

Acknowledgements

The authors would like to thank A. Cool and F. Verbunt for helpful discussions and the referee, H. Johnston for valuable comments. We would also like to thank M. Watson and P. Marty for their help during the analysis of the extended emission region, V. E. Zavlin for providing us with the pure H neutron star atmosphere model, and G. Hasinger for discussions about the interpretation of the Lockman Hole results. Finally, we are grateful to G. Lamer and J. Tedds for help with the astrometric correction.