3. Galaxy magnitudes

The magnitude histograms for the galaxies and stars misclassified as non-PSF-like objects are displayed in Figs. 7 and 8 respectively. The percentage of contamination by stars is very high: about 20%.

For a large fraction of the galaxies taken from the literature, Rmagnitudes were made available to us by C. Adami (private communication). In order to give magnitudes for all galaxies in the same photometric band, we estimated the R band magnitudes for the galaxies that we observed. This was done by identifying galaxies both in our photographic plate catalogue and in our CCD catalogue, and by finding the best linear fit (see Slezak et al. 1999 for details). This best fit was: $R=b_{\rm J} - 0.28$, with the slope fixed to 1.0, giving an rms fluctuation of 0.04. The galaxies in the literature for which Adami did not have magnitudes were identified with objects in our photographic plate catalogue, and we applied the same relation to derive their R magnitudes from their $b_{\rm J}$magnitudes. For some of the CfA galaxies, only $B_{\rm T}$ magnitudes were available; we roughly transformed them to R magnitudes, by calculating the mean (V-R) colour for the 239 galaxies detected in Vin our CCD catalogue (Slezak et al. 1999); this mean value is equal to 0.48, and therefore roughly corresponds to a G0 star for which (V-R)=0.52 and (B-V)=0.58 (e.g. Allen 1981), leading to $R\sim B$-1.1; for these CfA galaxies, we therefore took $R=B_{\rm T} -1.1$.

33 galaxies in our redshift catalogue have no identification with our photometric catalogue: 4 from Quintana & Ramírez (1990), 14 from Malumuth et al. (1992), 1 from Proust et al. (1987) and 14 from Huchra et al. (1992). They are the objects in Table 4 that have labels 2 to 5 in Col. (16), and their corresponding identification number in the literature, but no identification number from our photographic plate catalogue. Note that out of these 33 objects, half are located outside the area covered by our photometric catalogue, and three are located less than 1 arcmin from the cluster center, so we cannot separate them from the cD. The reason for the lack of identification of the other objects is not clear, since in all the cases where we did identify galaxies from our sample to galaxies in the literature the coordinates matched within a few arcseconds. The most likely explanation is that the coordinates given in the literature for these objects are not accurate; the fact that they may have not been detected in our photographic plate scan is unlikely since they are all very bright $(R\leq 16.8)$.