Tables 1-3 are only available in electronic form at http://www.aanda.org

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

http://cxc.harvard.edu/ciao/

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...=2.0 keV)

http://asc.harvard.edu/ciao/download/doc/expmap_intro.ps

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... PWDetect

See http://www.astropa.unipa.it/progetti_ricerca/PWDetect

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... XIMAGE

http://heasarc.nasa.gov/xanadu/ximage/ximage.html

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... scales

This effect occurred for only 3 sources at large off-axis angles, where the PSF is particularly elongated and thus significantly different from the symmetric PSF assumed by PWDetect. Note that, quite obviously, this effect is also not included in the simulations of source-free fields.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... convention

http://cxc.harvard.edu/cdo/scipubs.html

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... XTRACT

http://www.astro.psu.edu/xray/docs/TARA/ae_users_guide.html

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... TARA

http://www.astro.psu.edu/xray/docs/TARA/

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... FTOOLS

http://heasarc.gsfc.nasa.gov/docs/software/ftools/

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... hints

http://www.astro.psu.edu/xray/docs/TARA/ae_users_guide/node35

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... arcsec

http://www.astro.msfc.nasa.gov/Ch4/Ch4_15-03_Tsujimoto.pdf

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... (PIMMS

http://heasarc.gsfc.nasa.gov/docs/software/tools/pimms.html

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... (PSC)

See http://www.ipac.caltech.edu/2mass

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... range

We only consider COUP sources for which mass estimates are given by Getman et al. (2005a).

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... difference

A direct comparison with the results of Hillenbrand (2005) is however not possible because her estimate of disk lifetime (i) refers to star in the $0.3{-}1.0 ~M_\odot$ , i.e. less massive than the ones we observe in Cygnus OB2, (ii) is based on a different, and more efficient, indicator of disks presence than the one we can use here, i.e. the H-K color excess measured with respect to the photospheric value as determined from spectral types.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... diagram

The use of the J-band is justified because (i) in the presence of disk excesses the J-band is the most representative of the photospheric emission and (ii) the mass ranges in which the mass-luminosity relationship is degenerate are narrower than for a similar relationship in the H and K bands.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... stars

There are only 28 stars with masses in the $2.8{-}5.3~M_\odot$ range and for which this interval formally corresponds to the uncertainty on the mass estimate. We therefore expect a small fraction of the 57 intermediate mass stars to lie in the $2.8{-}3.0~M_\odot$ range and to have a substantial model-predicted convective envelope.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... types

A number of X-ray detected stars with no spectral type appear to be of high mass but with lower X-ray luminosities than those with spectral types. In the discussion of high mass stars, however, we will only consider stars with spectral types.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... 9

One undetected B stars has an unknown luminosity class.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... spectrum

The temperatures reached by hydrodynamic shocks in the winds of single stars are usually not high enough to produce significant Fe K $_\alpha$ emission. In wide early-type binary systems, however, stellar winds collide with velocities close to $v_\infty$ and are thus heated to sufficiently high temperatures. In such cases, the cooler surrounding wind material, excited by the high energy radiation, can produce the 6.4 keV fluorescent Fe line. The observation of the Fe-complex can thus be a used as diagnostic for colliding-wind binaries (Raassen et al. 2003).

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.