2. Observations and data reduction
2.1. Surface photometry
We obtained CCD images of Hickson 96 in the Johnson BVR bands. Fig. 1 shows the V band image of the entire group.
Offset positions are in arcsec relatives to the center of NGC 7674 , . The images were obtained at the prime focus of the 3.5m telescope of the Centro Astronómico Hispano-Alemán in Calar Alto (Spain). The detector was an RCA CCD of 1024 640 pixels of size, giving a scale of 0 25 pixel-1, and a field of view of 4 2 2 4. The exposure times were 1000s, 700s and 500s in B, V and R respectively. The seeing of each frame was measured from the FWHM obtained by fitting an analytic Moffat function to the profiles of stars in the field of the group. It was found to be 1 0 in B, 0 8 in V and 1 0 in R.
The reduction and calibration of the data were carried out using standard techniques and produced fluxes that are accurate at the 2% level. Bias exposures taken through the run were found to be constant and were used to construct an average bias which was subtracted from each image. Pixel-to-pixel variations were evaluated with dome flat-fields. All flats, after normalization by their median values, show similar form and no important differences. Images were divided by the averaged flat in each filter. The atmospheric extinction was determined from observations of three selected fields in the open clusters NGC 272, Be87 and NGC 366. The stars were chosen to cover a wide range of colours to take into account colour effects in the standard system. The rms errors of the standard stars in the final calibration are smaller than 0.05 mag in all colours. We subtracted a constant sky background from each frame. The errors due to variations in the sky were always smaller than 1%. Colour indexes given in this paper and noted with subindex 0, have been corrected for galactic absorption (using the extinction value given by Burstein & Heiles, 1984, with the reddening law from Savage & Mathis, 1979), internal extinction (de Vaucouleurs et al., RC3, 1991) and K-effect.
We have compared our photometric data with those by Longo & de Vaucouleurs (1983) and Hickson, Kindl & Auman (1989). V magnitudes are consistent to better than 2% and colour indexes to better than 8% for NGC 7675. Agreements are better than 0.1 mag for apertures smaller than for NGC 7674. At larger radii the photoelectric photometry is probably affected by the close companion H96c and a star.
Table 1 summarizes the long slit spectra taken for this study. The format is as follows: Column 1) spectrum identification; 2) date of observation; 3) spectral dispersion; 4) exposure time in seconds; 5) spectral range and 6) position angle of the slit in degrees. The first four spectra were taken with the reflective aspherized grating spectrograph Carelec (Lemaître et al. 1990) used at the Cassegrain focus of the 193 cm telescope at the Observatoire de Haute Provence. The first of them was obtained with a thick front illuminated Thomson CCD with 576 384 pixels of 23µm size. The reciprocal dispersion of 33Å mm-1 gives 0.76 Å pixel-1 and a spectral resolution of 1.9 Å. The spatial scale is pixel-1. The slit passed through the centers of H96a - c. The next three spectra (through the centers of H96b - d, a - c, and c - d) were taken with a thinned back illuminated RCA CCD with 512 320 pixels. The size of a pixel was 30µm and the dispersion was 260Å mm-1 yielding 7.8Å pixel-1 and a spectral resolution of 15.6 Å. The spatial scale is 1 3 pixel-1. The slit width of 2 5 provided good sampling with the seeing at Haute Provence during that run. The spectra were reduced using the usual methods including the instrumental correction by spectrophotometric standard stars observed each night. Suitable procedures were written inside the ESO MIDAS package.
Table 1. Long slit spectra.
Two additional long-slit spectra were taken along the minor and major axis of NGC 7675 at the 3.5 m telescope on Calar Alto, using the Twin spectrograph at the Cassegrain focus. As detector we use a TEK CCD camera with 1024 1024 pixels of 24µm. The spatial resolution was 0 9 pixel-1. We use a slit width of 250µm, corresponding to . The observing conditions were good, with a seeing better than 1 2. In this run we also obtained a total of 15 exposures of standard radial velocity giant stars of G and K type, in order to obtain accurate redshifts and velocity dispersions. These were calculated using the cross-correlation technique derived by Tonry & Davis (1979). The wavelength calibration was tested using sky lines present in the spectra. The rms of the central wavelengths in the lines for both spectra was less than 0.008Å, corresponding to 2 km s-1. We show one low resolution spectrum for each galaxy in Fig. 2.
In order to calculate the heliocentric velocity for each galaxy we have used the high resolution spectra for H96a and c, and compared the results with the values obtained for the low resolution spectra. In the case of H96b we have used a mean of the values obtained from the 3 different available slits (major axis, minor axis and p. a. = ). For H96d we only have a low resolution spectrum. For the emission lines we assume that the center of each measured spectral line coincides with the center of the best Gaussian fit. For several cases, where H , and [NII] or [OIII] were blended due to the low resolution, we have separated the components by a multiple gaussian fit made with a routine included in MIDAS. The results of these measures are given in Table 2. There is a good agreement between our values and those obtained by Hickson et al. (1992).
Table 2. Velocities of HCG 96 members.
© European Southern Observatory (ESO) 1997
Online publication: June 30, 1998