Astron. Astrophys. 321, 123-128 (1997)

2. The quasars sample

2.1. Definition of the sample

The SA94 sample is made up of 149 optically selected quasars ¹ listed in Table 1, observed in a rectangular area of the sky within the limits , (Epoch 1950.0), covering 17.66 sq. deg.

They have been mainly discovered in the UVx and objective-prism surveys described in Paper I and in the references to Table 1.

2.2. Calibration of the photographic material and error estimation

8 plates taken with the ESO La Silla and UK Schmidt telescopes and a copy of a POSS E-plate (E1453) have been analysed. In Table 2 a detailed list is given. The combination IIIa-F + RG630 filter was used in all cases except for the plate E1453, that is a combination of a IIIa-E + a red Plexiglas 2444, providing a transmission similar to that of a Wratten filter (peak sensitivity near ). The natural photometric system defined by the IIIa-F + RG630 combination (in the following defined ) is close to the Johnson-Kron-Cousins R and the color transformation is given by Bessel (1986):

Table 2. List of the plates

The combination IIIa-E + red Plexiglas 2444 provides a system for our purposes indistinguishable from the previous one, with no significant color term in the range , as verified on our photometric material.

The plate material has been scanned with the COSMOS microdensitometer (MacGillivray & Stobie 1984) in IAM mode. The resulting tables, one per each plate, containing the instrumental magnitudes and other useful parameters for the objects detected, have been merged together in one table. Only objects with at least 3 detections in the 9 plates have been accepted in this final table containing 82 984 entries. The astrometric error box defining a common detection has been defined as a circle of 3.5 arcsec of radius. In this way spurious detections (plate defects) are minimized to an acceptable level, while real measurements are in practice never discarded.

The calibration of the magnitudes of each plate has been carried out according to the following procedure (see also Paper I):

1. the magnitudes of 108 standard stars, transformed into the natural colour system of the plates, , according to Eq. 1, have been used to derive for each plate a polynomial regression between instrumental and calibrated magnitudes.
2. the median magnitude of all the objects has been computed and each plate has been re-calibrated against the median magnitudes of the 27 111 point-like objects.
3. a procedure of uniformization of the usually spatially variable response of the photographic plates has been applied. Each plate has been subdivided in sub-areas; for each of them the differences between the reference and the individual plate magnitudes have been computed and their distribution analysed. The zero-point shifts estimated in this way for each sub-area as a function of the magnitude have been smoothed and applied to the re-calibrated magnitudes. In the following we will refer to the magnitudes obtained in this way as .
4. for each plate the uncertainties of the magnitudes have been estimated by analysing the distribution of the differences as a function of the median magnitudes for all the point-like objects.
5. the completeness limit of each plate (see Table 2) has been estimated as the maximum of the histogram of the calibrated magnitudes.

© European Southern Observatory (ESO) 1997

Online publication: June 30, 1998
helpdesk.link@springer.de