5 Results

In this section we discuss the results of the photometric decomposition, compare our results with previous work and present details on individual cases. The surface brightness profiles of the galaxies, their bulge and disk components are shown in the appendix A together with the ellipticity, position angle and Fourier coefficient profiles prior and after disk subtraction. The determined disk and bulge parameters are listed in Table 2 together with the estimated disk-to-bulge ratios D/B and other relevant parameters.

  

Table 2: Derived parameters of the bulge and disk components. For each object, we list the disk-to-bulge ratio D/B, the total magnitude $V_{\rm Tg}$, effective surface brightness $SB_{\rm eg}$ and effective radius $r_{\rm eg}$ of the galaxy; the total magnitude $V_{\rm Tb}$,the effective surface brightness $SB_{\rm eb}$ and effective radius $r_{\rm eb}$ of the bulge component; the total magnitude $V_{\rm Td}$, effective surface brightness $SB_{\rm ed}$ and effective radius $r_{\rm de}$ of the disk component; the inclination i, and the face-on corrected central surface brightness SB₀ and mean scale length of the disks $r_{\rm d}$. The disk parameters were corrected to face-on. The radii are all given in kpc. Typical errors are $\Delta SB_0$ $\simeq$ 0.2 and $\Delta r_{\rm d}$ $\simeq$ 20% - 30%

5.1 Comparison with previous work

We have compared the results obtained here for three galaxies (NGC 3585, NGC 4697 and IC 2552) with those given in previous work by Scorza & Bender (1995) and Scorza (1993), and in the case of NGC 3585, also with the recent decomposition made by Michard (1997). The comparison enables us to test the automatized version of the method applied here, and particularly the assumption on the shape of the disk profiles, which in Scorza & Bender (1995) were taken to be the sum of two or more exponentials, and here left free.

The surface brightness profiles of the comparison galaxies, their bulges and disks are shown in Figs. 2a, 2b and 2c. Shown are the bulge and disk surface brightness as determined here (full line) and by Scorza & Bender (1995) (dashed line). In the case of NGC 3585 (Fig. 2a), the dotted line shows the decomposition of Michard (1997). The differences in the surface brightness profiles of the bulges at large radii are probably caused by differences in the determination of the sky background in the three mentioned works. This seems to be indeed the case since the profiles in all works lay very close to each other at small radii, but diverge at large radii (in a more extremely way Michard's). The same holds for the other comparisons shown in Figs. 2b and 2c.

The surface brightness profiles of the disks show at small radii (< 3 arcsec) significant differences which are most probably due to the uncertainties caused by seeing. Outside these radii the profiles have a similar run in all decompositions, but at larger radii the difference becomes again noticeable. The disk profiles derived here are slightly brighter at large radii than those derived by Scorza & Bender (1995). This can in part be due to the non-exponentiality of the disk profiles.

  

Figure 2: Comparison with previous work. Surface brightness profiles SB (in mag/arcsec2) of the bulge and disk components of: a) NGC 3585 as determined here (full line), by Scorza & Bender (1995) (dashed lines) and by Michard (1997) (dotted line); b) NGC 4697 as determined here (full line) and by Scorza & Bender (1995) (dashed lines), and c) IC 2552 as determined here (full line) and in Scorza (1993) (dashed lines)

Another difference resides in the higher central surface brightness values of the latter work, explained by the fact that the authors combined two exponentials to construct their models. Thus, the smaller scale lengths and higher central surface brightnesses seem to be due to a systematic difference between the two decomposition techniques and not to an error in the determination of the disk parameters. The same is true for the slightly higher apparent magnitudes of the disks derived in this work. However, the influence of the different shapes of the disk profiles as found here and in Scorza & Bender (1995) on the D/B ratio is in comparison quite small. The values found here are $\sim$2% higher. It is worthy to mention that the inclination values found with both methods agree perfectly. The difference is smaller than $3^\circ$.

The luminosity profile of the disk in NGC 4550 determined here, agrees well in the range r=10-20 arcsec with the profile derived by Rix et al. (1992) from the kinematic decomposition of the VLPs. As mentioned in the introduction, the asymmetric VLPs observed in disky ellipticals make an embedded disk component the most plausible explanation for the peakness of the isophotes.

5.2 Discussion of individual cases

We turn now to discuss in detail the decomposition of six galaxies belonging to the three types of objects described in Sect. 3.

5.2.1 Bulge-dominated objects

Several clean examples for this type of objects have been discussed in Scorza & Bender (1995). We present here further details for ESO 208-21 and NGC 3818.

ESO 208-21: The disk of this galaxy is fully embedded in the bulge component, as indicated by both the a₄ Fourier coefficient and ellipticity profiles (see Fig. A1 in the Appendix). These show single maxima at intermediate radii and decline at large radii, where the bulge dominates. The surface brightness profile of the disk was found to be exponential over a large radius range and to turn slightly steeper inside 10 arcsec. The disk is very compact, with a mean effective surface brightness of 19.6 mag/arcsec² and an effective radius of 0.36 kpc.

NGC 3818: This object is representative of the class having a fully embedded disk in a boxy bulge. The a₄ profile of the original galaxy has negative values at large radii (see Fig. A15 in the Appendix). The disk has been removed in a way to give the bulge constant boxiness. Uncertainties associated with this procedure, e.g. on the D/B ratio, are discussed in Scorza & Bender (1995). The disk profile is exponential all throughout.

5.2.2 Disk-dominated objects

NGC 4831: The a₄ Fourier coefficient of this galaxy shows a single maximum at 16 arcsec (see Fig. A22 in the Appendix) and beyond this radius decays while the ellipticity continues rising. Like NGC 4564 (described in detail in Sect. 4.1) two steps were applied in the decomposition of this object. The solution with the brighter disk yielded a D/B = 1.04. The disk profile of this galaxy is exponential at large radii and turns steeper at small radii. The disk was found to have an inclination of $65^\circ$degrees.

NGC 7041: This is also an object with a dominant disk, as indicated by the high a₄ and ellipticity values. The disk has a profile well represented by a power+exponential law (see Fig. A27 in the Appendix). The decomposition made here agrees with the classification of the RSA (S0) and RC3 catalogue (T=-2.7).

5.2.3 Barred objects

IC 2552: This object shows the correlated behaviour of the a₄, position angle and ellipticity described in Sect. 4.1 (case III). After the bulge-model subtraction, an elongated bar-like structure is observed. At the radius where this sub-structure becomes dominant, the ellipticity, a₄ and position angle increase abruptly, and fall again at radii where the sub-structure fades. Outside this radius the galaxy turns very round (ellipticity $\sim 0.04$). The twist observed in this galaxy is $30^\circ$. The surface brightness profile of the sub-component is exponential at large radii and flattens in the inner parts (see Fig. A4 in the appendix). This suggests the presence of a face-on disk with a bar. Figure 3a  shows the residual bar-like structure which extends 8 arcsec from the center.

  

Figure 3: Residual bar-like structures in a) IC 2552, showing outer rectangular and inner peaked isophotes (the scale is $20 \times 20$ arcsec), and b) in NGC 4024, showing rectangular isophotes. The scale here is $22 \times 22$ arcsec

NGC 4024: This objects shows the same characteristics as IC 2552 (see Fig. A16 in the Appendix). After the bulge-model subtraction, the bar-like structure shown in Fig. 3b becomes visible.