4 Data reduction

4.1 Galaxy positions  

Precise galaxy positions were independently derived from four deep plates (Nos. 8417, 8753, 8776 and 8788) taken under 1''-2'' seeing conditions and digitised in the 12bit mode. We usually applied the Gaussian centring algorithm of the standard MIDAS software package onto the galaxy cores. In a few doubtful cases of irregular galaxies or blended cores the centre position was estimated by eye. The reference frame for the astrometric reduction was realized using the 40 faintest stars from the PPM catalogue (Röser & Bastian [1991]) in this field. Each plate was separately modelled by a 2nd order 2D polynomial fit. For each galaxy, the obtained four positions were averaged, yielding the mean galaxy position with a typical standard deviation $\sigma$ of $0\hbox{$.\!\!^{\prime\prime}$}25$.The mean total error of the derived galaxy positions (including possible systematic ones) is expected to be $\le0\hbox{$.\!\!^{\prime\prime}$}5$.

We compare the obtained galaxy positions with the data from the Reference Galaxy List of the Lick Northern Proper Motion program (NPM, Klemola et al. [1987]), which contains 50517 galaxy positions with a typical rms error of $0\hbox{$.\!\!^{\prime\prime}$}3$.There are 35 NPM galaxies in our field, all of them could be cross-identified with galaxies of our survey. The differences between the NPM positions and ours are shown in Fig.1; the mean residual is $0\hbox{$.\!\!^{\prime\prime}$}56$, none exceeds 1''.

1mm

  

Figure 1: Comparision of galaxy positions derived in this paper versus NPM values (Klemola et al. [1987]). The plot shows the residuals in right ascension $\alpha$ and declination $\delta$ for all 35 galaxies common in both catalogues. The mean positional difference is $0\hbox{$.\!\!^{\prime\prime}$}56$

4.2 Photometry

Our photometry is based on the those four plates mentioned in Sect.4.1 in combination with plate 8841. The availability of five plates simplifies the plate fault removal and increases the signal-to-noise ratio. Again, subframes of $200''\times200''$ have been extracted which are centred on the galaxy cores ($400''\times400''$ for NGC1275). Surface photometry is carried out according to the standard procedures (e.g. Okamura [1988]):

The measured relative photographic transmission T is converted into the relative intensity I of the incident light by means of the characteristic curve which was approximated by

$$\log_{10} I = c_1 + c_2\,\log_{10}(D)+c_3\,D^{c_4}+c_5\,D$$ (1)

(Lehmann & Häupl [1988]), where I designates the intensity of the light the emulsion was exposed to, and $D=-\log_{10}\,(T/T_{\rm max})$ the resulting photographic density. The coefficients c₂ to c₅ are derived for each plate separately from a least-square fit onto the calibration wedge. c₁ is determined by the aperture photometry procedure described below.

In the next step, the local sky background has to be subtracted. The local sky background is assumed to be constant in each sub-frame of (usually) $200''\times200''$.Since the galaxy images cover only a small fraction of each frame, the mean local sky background has been assumed to be equal to the well defined maximum of the intensity histogram of the image.

It is well-known that background subtraction is rather difficult when very faint signals of extended objects have to be discriminated against a much brighter sky background, especially in the irregular environment of a galaxy cluster (e.g. Capaccioli & de Vaucouleurs [1983]; Okamura [1988]). The photometric errors resulting from our simple background subtraction procedure are expected to become significant for the faintest detectable features. However, it is not our intention to achieve highly accurate photometry of very faint structures, although their visual inspection is very important in the context of morphological classification (see Sect.5). The apparent magnitudes are derived from the integrated brightness of the galaxies within the $\mu_B=25$magarcsec^-2 isophote. The latter corresponds to 5% of the sky background level which is well above the noise level.

Prior to the determination of isophotal parameters, plate faults as well as foreground stars and overlapping galaxies have to be eliminated carefully. The availability of five co-centred images of each galaxy simplifies the removal of spurious objects. At each pixel position, both the highest and the lowest measured intensity values are omitted. Only the three medium intensity values are averaged to construct a final, "clean'' image for each galaxy.

Due to the low Galactic latitude the investigated field is strongly contaminated by foreground stars. Since the approximation of stellar images by a Gaussian distribution and subsequent removal does not always lead to satisfying results, we have decided to remove disturbing object images manually. Stellar images projected on the outer parts of the galaxy images are either removed by the subtraction of the fitted Gaussian profiles, by simply replacing the affected pixel values by those from similar unaffected parts of the image, or by means of the MIDAS procedure MODIFY/AREA. The removal of disturbing images near the galaxy centre is more complicated and time-consuming; in the case of E or S0 galaxies good results are achieved by first determining the galaxy centre and then replacing the affected pixel values by those on exactly the opposite side of the image at the same distance from the centre. In a few cases, when none of the methods mentioned above worked properly, we have used a pixel editor to reconstruct the undisturbed image as well as possible. Throughout the cleaning process, the images have been monitored on the display in order to avoid any misfits. Although manual star removal is time-consuming and seems to be somewhat inaccurate, it rather well preserves galaxy features that may be removed by automated cleaning procedures.

For the absolute flux calibration of the images, the results of the photoelectric and photographic aperture photometry of A426 galaxies by Weedman (1975) and Strom & Strom (1978) serve as reference. We have only considered apertures $\ge17''$ to avoid seeing and sampling effects. NGC1275 has been excluded from the calibration sample mainly due to a close foreground star. For the remaining 50 galaxies, the internal nuclear magnitudes have been determined from numerical integration of the relative intensities within the corresponding aperture. After subtraction of the mean offset between the published and the measured nuclear aperture magnitudes, the remaining rms difference is 0.08mag (Fig.2). The systematic error of the absolute calibration calculates as 0.08mag$/\sqrt{50}=0.01$mag.

1mm

  

Figure 2: Photometric zero point calibration: aperture photometry of galaxies of the present catalogue ($B_{\rm cat}$), compared with results published by Weedman ([1975]) and Strom & Strom ([1978]) ($B_{\rm lit}$). Mean standard deviation $\sigma(B_{\rm lit}-B_{\rm cat})=0.08\,$mag

From the numerical integration of the calibrated intensities within the $\mu_B=25\,$magarcsec^-2 isophote, the apparent magnitude B₂₅ of each galaxy has been derived. The total error is estimated to be of the order of 0.1...0.2mag. No colour correction has been applied since the Tautenburg B band almost exactly matches the Johnson B band for 0<(B-V)<2 (van den Bergh [1964]; Börngen & Chatchikjan [1967]; Andruk et al. [1994]). Major and minor axis as well as position angles (PA) of the galaxies have been also derived from the 25magarcsec^-2 isophotal ellipse and are included in the catalogue.

The sky background is $\mu_B=21.7~\pm ~0.1$magarcsec^-2 for all five plates. Surface brightness fluctuations introduced by the granularity are of the order of 0.07magarcsec^-2 at the 10$\mu$m scale for surface brightnesses up to 19magarcsec^-2.

4.3 Radial velocities  

We have derived radial velocities for five of the catalogued galaxies from low-dispersion spectra taken at Calar Alto (Sect.2). The radial velocities of two IRAS galaxies could easily be measured thanks to the presence of prominent emission lines in their spectra. No strong emission lines have been found for the other galaxies, but absorption features of a late type stellar population were unambiguously identified: CaII K&H $\lambda \lambda 3933, 3968,\ G$ band $\lambda 4307$,MgI $\lambda 5175$, NaI $\lambda 5893$. The resulting heliocentric radial velocities are included in the catalogue. The typical uncertainties are $\sim200$kms^-1.