The effective point spread function of the ROSAT telescope is significantly broader during the All-Sky-Survey than during the pointed observations. This is due to the scanning mode where the PSF is averaged over a wide range of off-axis angles. For most of my sources the number of detected photons is too low to recover this resolution. For consistency I choose a source independent 90 percent error radius r90 of 30 arcsec, regularly used in ROSAT source identification (Hasinger priv. com.). Motch et al. (1997a) calculate an upper limit of r90=32 arcsec for their extensive identification program and find that this is a conservative limit. The result of my identification program supports this finding, see Fig. 7 (click here).
Figure 7:
Separation between the X-ray sources and their counterpart.
left: number of counterparts found in a given radius
interval. Distance in arcsec.
right: same distribution but normalized by the radius
I searched the astronomical catalogs provided by SIMBAD and the
NASA Extra Galactic Database (NED) for objects at the X-ray position.
In SIMBAD a search radius of 1 arcminute and in NED a search radius
of 5 arcminutes was used. Any object within 30 arcsecond of
the reported X-ray position was considered a counterpart to the
source as long as there was no compelling evidence found to contradict
this.
Of the 89 sources coinciding with clouds, 11 were readily identified
with bright (mV< 10.6) stars and 6 others with nearby low mass dwarf
stars and one emission line star (mV,14.6).
Two sources coincide with known quasi-stellar objects (QSOs) and
three others with known radio sources.
The Abell cluster 0496 was found only 1.5 arcminutes distant from
RX 
. Table 2 (click here) shows the names of all
identified SIMBAD sources.
In addition to the DSS images, I created finding charts from data in the Automated Plate Measuring machine catalog. The APM group at Cambridge independently scanned plates from the Palomar Observatory Sky Survey. The significant advantage of the APM catalog over the DSS scans is the additional color information as both red (E) and blue (J) plates have been scanned. The detection limits are around 21.5 mag in J and 20 mag in E. The APM data base is an object list, including an automated classification into stellar and non-stellar objects, rather than a collection of images. This can cause confusion when large extended structures such as bright emitting clouds are encountered. A comparison between the images from the DSS and the APM scans can then resolve ambiguities in the morphology of the source. I give finding charts for all objects of which spectra have been taken in Fig. 9.