The radio sources were identified with their optical counterpart on the basis of its positional coincidence with the radio centre. If a compact radio core was detected, then it undoubtedly coincided with the identification. If a source was an extended one and no core could be detected, then a much larger area near the "centre" of the source had to be considered. Usually the source was identified with the brightest galaxy or stellar object in the "centre". Majority of the sample sources were identified with galaxies and quasars, as well as galaxy or quasar candidates seen on the Palomar Observatory Sky Survey (POSS) prints. The tentative samples of GB/GB2 radio galaxies and quasar candidates were analyzed by Machalski & Condon (1985b, 1986). Later on, deeper identifications were collected from the literature and our own research (e.g. Machalski & Magdziarz 1993b). In Notes to Table A3 (click here) some ambiguous identifications are discussed. Unfortunately, still about 34 per cent of the sources in Part 1, about 45 per cent in Part 2, and about 52 per cent in Part 3 remain unidentified.
Red (R) and blue (B) magnitudes of the identified objects are either photometric ones in the R Kron-Cousins VRI system (Cousins 1976) and in the UBV system of Johnson (1966), or estimated from the POSS(E) and POSS(O) prints, respectively. In particular, UBVRI photometry for the brightest elliptical galaxies and quasars in the sample was provided by Machalski & Wiśniewski (1988), and absolute MR and MB magnitudes as well as four intrinsic colours of the elliptical galaxies, corrected for galactic extinction, aperture, and reddening (K-correction), was given by Machalski (1988). Statistics of the optical type is given in Sect. 6.
Spectroscopic redshift is available for majority of the sample quasars, and for many brighter galaxies. References to the available photometry and redshifts are provided in Table A3 (click here). For elliptical galaxies without spectroscopic redshift but with photoelectrical photometry, a redshift estimate, based on corrected R-magnitude and B-R colour, is provided by Machalski (1988). A statistical analysis, described in Sect. 6 suggests that at least 11-12 sources in the sample may have redshift greater than 3.16.
The sample sources are identified with X-ray sources observed with the
Imaging Proportional Counter (IPC) on the EINSTEIN observatory
(Wilkes et al.
1994), and the Position Sensitive Proportional Counter (PSPC) on the ROSAT
observatory. The latter data were taken from the "WGA" catalogue, and an
updated version of the "First ROSAT source Catalogue
of Pointed Observations".
An X-ray source is considered as a counterpart for a radio source if its
normalized distance to the radio position is
where: 
and 
are the differences between the X-ray
and radio position, and 
and 
are the
combined X-ray-radio position errors
in right ascension and declination, respectively. Majority of the X-ray sources
are identified with compact radio sources (mostly quasars and BL Lac objects)
for which accuracy of the VLA position is 
arcsec.
Because the X-ray data available are not homogeneous, and the GB/GB2 sky region is only partly covered by the above X-ray observations, their details (epoch of observation, integration time, resultant counts, etc.) are not cited in this paper. Only a reference to the original data is given.