3 Classification of box/peanut bulges

Shaw (1993a) and Lütticke (1996) derive objective parameters to characterize quantitatively b/p structures by the measurement of the excess luminosity and the total or maximal fraction of the b/p distortion (depending on the radial distance from the center) in relation to the observed bulge luminosity. However, the application of such a classification method is unpractical for a large sample of galaxies and strongly depends on the modelled luminosity distribution for the disk and the elliptical part of the bulge.

The a₄ isophote shape parameter is used by Combes et al. (1990), Shaw (1993a), Lütticke (1996), and Merrifield & Kuijken (1999) to quantify the degree of boxiness of bulges by measuring the deviations from perfect ellipses (e.g. Bender & Möllenhoff 1987). However, it is problematic that the determination of the extreme value of a₄ depends on the fitted region of the bulge. Fitting only the inner parts of bulges results in some galaxies in the undetection of the boxiness of the bulge which is most prominent in the outer parts. For instance the clear boxy bulge of NGC 1055 (Shaw 1993b; Lütticke et al. 2000b, hereafter Paper III) is undetected by Merrifield & Kuijken (1999) using the a₄parameter. An additional disadvantage of the determination of this parameter is the influence of the masked stars in the foreground, dust, bars, and the extreme nature of b/p isophotal distortions for the ellipse fitting. Therefore the region of the galaxy which is used for the fitting is important. Shaw (1993a) uses the whole galaxy, Lütticke (1996) the disk subtracted galaxy, and Merrifield & Kuijken (1999) mask out a wedge-shaped region of each image within 12 degrees of the disk major axis and only fit "on the side of the galaxy where the disk projects behind the bulge''.

These disadvantages lead to the fact that the a₄ parameter is not suitable for uniform classification of a large sample of bulges. Additionally, the resolution of the DSS images is too low to determine the a₄ parameter for the whole sample. Therefore the bulges of the investigated sample are classified by their degree of b/p shape derived by visual inspection from contour plots of the galaxy images.

Reshetnikov & Combes (1998) use similar arguments to choose a "straightforward procedure of eyeball estimation'' with isophotal maps from the DSS instead of objective criteria for the detection and classification of warped disks.

The bulges are divided in three types of b/p bulges ( ${\bf 1 - 3}$), elliptical bulges, and unclassifiable bulges.

1: peanut-shaped bulge;

2: box-shaped bulge;

3: bulge is close to box-shaped, not elliptical;

4: elliptical bulge;

5: unclassifiable bulge.

Type 1 bulges are described by a depression along the minor axis on both sides of the main axis. In this way the bulge looks like a peanut (Fig. 1, top). The depth of the depression can be used as a characteristic parameter for these peanut-shaped bulges (Lütticke 1999). The box-shaped bulges (type 2) are defined by isophotes parallel to the major axis. Therefore the bulge appears like a box (Fig. 1, second image). Some bulges of this class have a very prominent box form (Fig. 3, top). They could be type 1, but due to low resolution of the images it is not possible to see the eventual depression along the minor axis. These bulges are called 2+. The same class is used for bulges, which are on one side boxy and on the other peanut-shaped. Type 3 (Fig. 1, third image) is less well defined. These bulges possess a general flattening of the isophotes parallel to the major axis which is less pronounced as the flattening of type 2. However, this flattening differentiates these bulges from the purely elliptical bulges of type 4 (Fig. 1, bottom). The limits between the classes are not sharp: obvious between 1 and 2, clear between 2 and 3, but sometimes indistinct between 3 and 4. The classification of bulges can therefore in some cases be ambiguous, but a check of the whole sample (two independent classifications by the author and one of the co-authors) shows that more than 90% of the bulges are well defined by one of the different bulge types.

Figure 1: Examples for the different bulge types from the DSS. Isophotes are logarithmically equidistant

Several reasons are differentiated why bulges are not classifiable:

5.1: Inclination is too far away from edge-on;

5.2: Stars in the foreground projected onto the bulge;

5.3: Galaxy is strongly perturbed;

5.4: Dust conceals the shape of the bulge;

5.5: Bulge is too small;

5.6: Signal-to-noise ratio of the image is too low.

For galaxies which meet more than one of the above reasons only the most important one is listed (Tables 6 and 7). The classification for all galaxies in the sample with some minor changes after inspection of CCD images (Sect. 4.2) are given in Table 6.