During the ROSAT All-Sky Survey (RASS) about 60000 X-ray sources were
detected in the soft X-ray range between 0.1 and 2.4 keV with the PSPC
detector (Voges 1994). This large number of sources
does not allow a complete
identification of all optical counterparts. However, a statistical
classification of
the RASS source content can be obtained from much smaller complete
subsamples if variations of
the column density of neutral hydrogen,
, the background, and the
integration time are taken into account. Such complete samples of X-ray
sources are obviously an important and valuable basis for the study of X-ray
properties of AGN, galaxy clusters, and coronal and other galactic X-ray
emitters. At low galactic latitudes the RASS is investigated in this way by
the ROSAT Galactic
Plane Survey (e.g. Motch et al. 1991).
Our objective is the investigation of the RASS at high galactic latitudes
(
).
Our knowledge of the composition of the X-ray sky at high galactic latitude is based to a large extent on the results of the EINSTEIN Medium Sensitivity Survey (EMSS) (Gioia et al. 1990; Stocke et al. 1991). A basic result was that 51% of the EMSS sources are AGN (QSOs, Seyferts) and 26% are of stellar origin. The remaining 23% are BL Lacs (4%), normal and cooling flow galaxies (3%), clusters of galaxies (12%), and unidentified sources (4%). Recently, deep pointed ROSAT surveys such as the Lockman Hole studied by Hasinger et al. (1993), (1994) and the QSF3 field by Shanks et al. (1991) contributed new information, in particular concerning the composition of the soft X-ray background. In the deep sample studied by Hasinger et al. more than 60% of the sources were found to be AGN whereas the stellar sources contribute only 8%. Shanks et al. obtained a similar result.
In the RASS we can expect a composition different from these samples:
The sources in the EMSS were found serendipitously in fields
of otherwise "interesting'' objects.
Therefore this survey is
not representing the average X-ray sky with respect to the background and,
in particular, 
which could introduce a bias in
the composition of the sample. If e.g. 
in the locations of
the EMSS targets is lower than on the average in the RASS at high galactic
latitude, the sample would be biased towards extragalactic sources.
Furthermore, compared to ROSAT the EMSS
energy band is harder. Therefore, in the EMSS preferentially harder X-ray
sources were selected. Differences in the relative fractions of the various
object classes due to the energy bands are e.g. suggested by the results of
the EXOSAT High Galactic Latitude Survey (HGLS) (Giommi et al.
1991) which covered an energy band very similar to that of
ROSAT (0.05-2keV). This survey contained a significantly higher fraction of
stellar counterparts than the EMSS. In the deep ROSAT surveys quoted
above the detected sources are mostly at X-ray fluxes one to two order of
magnitude below the flux limit in the RASS. In addition, the column
density of neutral hydrogen is generally low in the directions of the deep
surveys (e.g. the Lockman Hole) and the area of sky covered is quite small.
Therefore we started an optical identification program for northern RASS X-ray sources at high galactic latitude. The goal was to identify optical counterparts of all X-ray sources in the selected sample. This investigation is a collaboration of the Landessternwarte (LSW), Heidelberg, the Max-Planck-Institut für Extraterrestrische Physik (MPE), Garching, and the Instituto Nacional de Astrofisica, Optica y Electronica (INAOE), Puebla. A similar investigation is being carried out in the southern hemisphere as an ESO key program (Danziger et al.)
Details of the X-ray survey data and their reduction have been discussed in Voges et al. (in preparation, Paper I). The full catalog of the identifications is published in Thiering et al. (in preparation, Paper III). A detailed statistical analysis of the results is given by Krautter et al. (in preparation, Paper IV). In the present paper we describe the methods of the optical identifications and we discuss first results. In Sect. 2 (click here) the RASS sample is discussed. The optical observations and the identification procedure are described in Sects. 3 (click here) and 4 (click here). In Sect. 5 (click here) preliminary results are presented.