Because of their variability, it is an impossible task to compare directly two QSO surveys of the same region of the sky made at different epochs. However we now have, for a large fraction of the sky, the possibility to extract from the APS database, for any object, the O magnitude as measured on the Palomar Sky Survey plates with an accuracy of about 0.2 (Pennington et al. 1993). By doing this for all known QSOs found in the course of a number of different surveys of the same area of the sky, we may hope to reach as near as possible from an ideal survey complete to a well defined limiting magnitude.
We have extracted the APS O magnitudes, when available, for all objects in the QSO catalogue (Véron-Cetty & Véron 1998) brighter than B = 17, with MB < -24.0 and z < 2.15, located in the 2400 deg2 of the FBS at > 30. Whenever this O magnitude exists, we give it the preference. There are 18 such QSOs with O < 16.0 (from a sample of 105 UV-excess stars, we have found that = -0.17 with a rms error of 0.25), and seven with B < 16.2. Thus our ``complete'' sample contains between 18 and 25 QSOs brighter than B = 16.2, or 0.0075 to 0.010 deg-2, i.e. 1.2 to 1.6 times larger than the PG surface density. It should be possible, when the APS database will be completed, to get the O magnitudes of the seven objects for which they are not yet available.
A.M. Mickaelian is grateful to the CNRS for making possible his visit to OHP. A.C. Gonçalves acknowledges the Fundação para a Ciência e a Tecnologia, Portugal (PRAXIS XXI/BD/5117/95 PhD. grant).