For each galaxy, total B, V, R and I magnitudes were measured from the CCD images. They were determined from apertures chosen interactively to assure that the diaphragm is large enough to contain the whole galaxy but still small enough to limit the magnitude error due to the sky error. Foreground stars within the aperture were removed interactively. Table 4 gives the magnitudes.
and obtained in this work were checked against the magnitudes given in the ESO-LV catalogue. It was found that the ESO-LV magnitudes are systematically brighter than those measured here. For the galaxies observed in photometric conditions, one obtains with a scatter of and with a scatter of . These differences are in agreement with independent studies. Peletier et al. (1994) found that their B-magnitudes are , and their R-magnitudes fainter than those in the ESO-LV catalogue. This result is difficult to interpret because the agreement of the magnitude differences indicate a systematic zero-point offset for the ESO-LV magnitudes. However, Paturel et al. (1994) compared magnitudes in the ESO-LV catalogue with those given in the RC3 catalogue (de Vaucouleurs et al. 1991) and did not find any significant zero point differences between the two catalogues.
CCD magnitudes for the I-filter are available for ESO-LV 463-0200, 351-0180, 351-0250, 412-0210 and 352-0720 (Mathewson & Ford 1996). Comparison with total I-magnitudes obtained in this work for the galaxies observed in photometric conditions gives a mean difference of with a scatter of . This shows that there exists a zero-point difference between the two magnitude systems with the magnitudes obtained in this work being systematically fainter. The reason for this difference could not be found.
For the galaxies observed in non-photometric conditions, comparison of and with the ESO-LV data and of with the data of Mathewson & Ford does not show any significant deviations from the magnitude differences obtained for the galaxies measured in photometric conditions. This indicates that for these galaxies the errors for integrated magnitudes and surface brightness due to the weather conditions can be regarded as smaller than .
Figure 1 (click here) shows surface brightness , apparent ellipticity and position angle PA plotted against the apparent semimajor axis r for the sample galaxies. and PA are plotted for the R-filter because the R-images are those with the highest signal-to-noise ratio and the ellipticity as well as the position angle profiles are very similar for different wavelength regions.
Figure 1: Surface brightness , apparent ellipticity and position angle PA against semimajor axis r. and PA are taken from the R images. For details see text
Figure 1: continued
For galaxies which are not edge-on, the inner parts of the profiles are dominated by the light distribution of the bulge while the outer parts are dominated by the spiral arms. For galaxies with prominent spiral arms (ESO-LV 339-0120, 235-0080 and 352-0400), apparent ellipticity and position angle vary strongly throughout the galaxy.
For edge-on or nearly edge-on galaxies (ESO-LV 463-0200, 351-0240, 352-0540 and 352-0720), the inner parts of the profiles are dominated by the bulge, while the outer parts are dominated by the overall shape of the disc. The apparent ellipticity is small in the center, but throughout the linear part of the surface brightness profile, increases systematically until it reaches a constant value, which is the true apparent ellipticity of the galaxy disc.
For each galaxy, and PA for in the R-filter are taken as a measure for and PA of the whole galaxy. This isophote has been chosen because it is well within the disc. Table 4 gives magnitudes, apparent ellipticity and position angle for each galaxy. Comparison with the values given in the ESO-LV catalogue shows a good agreement for almost all sample galaxies. The mean differences are with a scatter of 0.14 and with a scatter of . There is, however, one galaxy for which the -value given in the ESO-LV catalogue strongly deviates from that obtained from the profile. This is found for ESO-LV 352-0720, where is given in the ESO-LV catalogue. The galaxy image shows an object with very high ellipticity, so the value of 0.39 is not representative for the outer parts of this galaxy. The value of found in this work is much more realistic for the galaxy as a whole.
The surface brightness profiles obtained in photometric conditions were compared to those given in the literature. The ESO-LV catalogue contains low-resolution photographic surface brightness profiles in B and R for all sample galaxies. For B, a mean difference of with a scatter of was found. For R, the mean difference is with a scatter of . The profiles obtained in this work are systematically fainter than the ESO-LV profiles. Most of this deviation is explained by the magnitude zero-point difference. Part of the deviation is found in the outer parts of the profiles, i.e. the profiles obtained in this work are steeper than those in the ESO-LV catalogue. It is not clear what causes this deviation. The differences are probably not due to the uncertainty in the sky subtraction. As a check the images with the fitted ellipses were subtracted from the original image. For none of the sample galaxies, any significant parts of the galaxies are found to be remaining after subtraction.
I-profiles of four of the sample galaxies observed in photometric conditions (ESO-LV 463-0200, 351-0250, 412-0210 and 352-0720) are also found in Mathewson & Ford (1996). The mean surface brightness difference is with a scatter of . As for the B and R filters, this is larger than expected from the magnitude zero-point difference, but is in a better agreement than the - and -profiles with the ESO-LV catalogue.
In order to determine scalelength and central surface brightness of the disc, an exponential law was fitted to the surface brightness profiles of each sample galaxy. Since the linear parts of the profiles are the ones dominated by the disc, these parts were used for the fit. Bulge parameters were not fitted because the angular diameters of the bulges are so small that they are strongly influenced by seeing.
Table 5 shows the results obtained for and . The -profiles for ESO-LV 286-0630 and ESO-LV 352-0400 do not show any linear parts, so no fits were made for them. In order to obtain a scalelength which is not influenced by the inclination angle of the galaxy, is measured along the apparent semimajor axis. The errors for and are obtained by comparing the values for different sub-parts of the linear parts of the profiles. ESO-LV 463-0200 and 351-0180 could be observed twice. The sets of images were taken under different seeing conditions. The difference of and found between these two sets are also used to determine the error. The errors obtained for and are similar for each galaxy and filter. Typical values are and .
It has been found by many authors that the surface brightness profiles of spiral galaxies often show deviations from an exponential law, which sometimes makes the definition of the linear part difficult. As can be seen in Fig. 1 (click here), such deviations are present in the galaxies studied here. The error of for is mainly due to these deviations.
The selection of the profile parts used for the fit was checked with the B-I and V-I colour profiles. For an exponential disc, colour profiles are linear. So the chosen parts should be those for which the colour profiles show least deviation from linearity, which was confirmed for the present sample.
For the 11 sample galaxies for which redshifts are available (see Table 2), disc scalelengths were determined in kpc. The mean values are: , , and . These results are in agreement with values found for other samples of spiral galaxies (e.g., Kent 1985; van der Kruit 1987; Andredakis & Sanders 1994; Courteau 1996).
For 13 of the 14 sample galaxies, the disc scalelength decreases systematically with increasing wavelength. ESO-LV 235-0080 shows the opposite behaviour, the disc scalelength increases systematically with increasing wavelength. For ESO-LV 351-0250, disc scalelengths increase slightly from I to V, but is deviating. ESO-LV 351-0250 is the only galaxy in the present sample for which an active nucleus was found. Maia et al. (1996) detected [OIII] emission at a rest wavelength . It is not clear yet if the surface brightness profiles of active galaxies show systematic differences compared to those of normal galaxies. Table 6 gives the mean scalelength ratios and their scatter for the different colours.
The systematic increase of disc scalelengths with decreasing wavelength has already been found by various authors. Elmegreen & Elmegreen (1984) obtained for face-on galaxies. The value found in this work is larger, which is due to the fact that the present sample does not include face-on galaxies only but also galaxies with larger inclination angles, which show larger ratios (see below).
Peletier et al. (1994) found that for bright galaxies with MK < -22 the ratio of the disc scalelengths in B and K, , increases systematically with increasing apparent ellipticity. Figure 2 (click here) shows disk scalelength ratios , and for the present sample. The trend found by Peletier et al. for is found for the optical and near infrared wavelength regions as well.
Figure 2: Disc scalelength ratios plotted against apparent ellipticity measured in the R-filter at . a) shows the scalelength ratios between B and I, b) shows the ratios between V and I and c) between R and I
The larger the difference in wavelength between the two bands, the larger is the increase of the scalelength ratio with increasing apparent ellipticity. These results confirm that significant colour gradients are present in the discs of the sample galaxies and that these gradients are larger for larger inclination angles.
Byun et al. (1994) simulated galaxy images for different amounts of dust, determined disc scalelengths and calculated scalelength ratios (see their Fig. 10). Unlike the galaxies investigated in this work, the model galaxies presented in Fig. 10 do not have bulges. For central face-on optical depths in the V-band, , the simulated ratios increase systematically with increasing inclination angle by . For larger amounts of dust, , the ratio is constant or decreases slightly.
This result is not in agreement with the ratios measured in this work (see Fig. 2 (click here)a). For edge-on galaxies, is about larger than for face-on galaxies. Byun et al. obtained their scalelengths from major-axis radial profiles, while in this work scalelengths are measured from elliptically-averaged profiles. Byun et al. stress that for galaxies with a significant large differences exist between the two types of profiles, especially for highly inclined objects. In the present work however, the fitting ranges for the scalelength measurements are chosen to be in the outer parts of the galaxy where the influence of the bulge is small, so the observed scalelengths should not be significantly affected. Furthermore, even if one excludes the highly inclined galaxies (e.g., objects with ), Fig. 2 (click here)a is still not in agreement with any of the model curves. So even though the comparison is difficult, evidence remains that the present data show a larger increase of scalelength ratio with increasing inclination than is predicted by Byun et al. (1994).
Before any final conclusions can be drawn the following steps are necessary: (i) the number of sample galaxies must be enlarged, (ii) for model galaxies with a bulge, scalelength ratios must be calculated from elliptically-averaged profiles, and (iii) the possibility of intrinsic colour gradients within the stellar disc due to population gradients must be taken into account. It is planned to simulate galaxy images with a wider variety of parameters than those used by Byun et al. in order to see if the observed results can be reproduced.
I would like to thank Dr. R.S. Stobie for allocation of observing time at SAAO/Sutherland. It is a pleasure to thank Drs. J.W. Menzies and J. Caldwell for assistance during the observations, Mrs. I. Bassett for prereducing the CCD images, and Profs. W.F. Wargau and W. Seitter for many useful discussions concerning galaxies and their properties. Special thanks go to the referee, Dr. F. Simien, for important and helpful comments, and to Dr. R. Dümmler for carefully reading the manuscript. Financial support of this work by the Deutsche Forschungsgemeinschaft DFG under the number Se 345/22-1 and by the Ministerium für Wissenschaft und Forschung des Landes Nordrhein-Westfalen/Germany (Lise-Meitner-Stipendium) is gratefully acknowledged. Finally, I thank ESO for the use of the MIDAS software for the calculation of the surface brightness profiles. This research has made use of the Simbad database at CDS, Strasbourg, France, and of the Lyon-Meudon Extragalactic Database (LEDA) supplied by the LEDA team at the Observatoire de Lyon.