4.4.1 Morphology and structure of BCGs
BCGs are smallish galaxies with high central surface brightness. The detailed morphology and surface photometry of BCGs have been studied by many investigators. The central morphology is often irregular due to the presence of active star formation, but this contains very little information on the extended light distribution which likely traces the dominant stellar mass.
Loose and Thuan (1986a) defined four subclasses depending on the morphology of the central star forming region and the surrounding host galaxy. Kunth et al. (1988) reached the similar conclusion that BCGs constituted a ``mixed bag'' of morphologies, including objects that appeared to be isolated extragalactic H II regions (e.g. Pox186), irregular morphologies, as well as more common cases with symmetric outer envelopes suggesting the presence of an old population. Salzer et al. (1989b) classified the emission line galaxies in the UM survey according to morphology and emission line properties. Telles et al. (1997) divided H II galaxies into two classes: Type I which have irregular morphology and are more luminous, while Type II have symmetric and regular outer structure. The existence of regular haloes, if corresponding to stellar emission, is in itself suggesting fairly high ages since relaxation times are of the order of a few times 108 years. Of course the stars may be younger if they formed later on in an already relaxed gaseous disc.
Quantitative surface photometry reaching faint isophotal levels began with the development of CCD detectors. Bergvall (1985) found that ESO 338-IG04 (=Tol1924-416) followed an exponential like surface brightness distribution in the I-band, suggesting the presence of an old stellar disc. Loose and Thuan (1986b) on the contrary found Haro2 to follow a more elliptical like light distribution, with some suggestion of redder colours in the halo. Similar results were obtained by Kunth et al. (1988).
The shape of surface brightness profiles in BCGs has been subject to some debate and both exponential and r1/4 laws have both been claimed to best match the data. Papaderos et al. (1996a) have proposed that profiles in general can be fit with a three-component model, with an exponential light profile in the outer parts. Exponential outer profiles were also found by Telles and Terlevich (1997). On the other hand Doublier et al. (1997) find r1/4 profiles in a substantial fraction of objects, while the rest have exponential light profiles. We note however that these studies do not compare easily because only a few objects are in common. For most BCGs the shape of the profile changes with radius, meaning that the fitting shape will be uncertain and critically depend on the sensitivity limits. Bergvall and Östlin (1999) go deeper than other published studies (to levels fainter than µB = 28 mag/arcsec2), and claim that r1/4 types laws are favoured when using deep red (R or I band) data, but that discs provide decent fits to the outer parts, and are favoured for B-band data. The shape of the luminosity profiles depends on how one constructs them, and especially the amount of central excess (identified with the ``starburst'' component) depends critically on the chosen method (Marlowe et al. 1997). Several investigators have found high underlying surface brightness (µB = 20 to 23 mag/arcsec2) and short scale lengths in BCGs as compared to other dwarf galaxies (Papaderos et al. 1996a, Telles and Terlevich 1997, Marlowe et al. 1997). Bergvall and Östlin (1999) find much lower central surface brightness values when using deeper data for a sample of luminous BCGs. Thus there might be differences arising from the different nature of the objects, but also from different observational methodology. Now, one can ask what meaning the shape of the luminosity profile really has? Most dIs and LSBGs, and faint dEs, are well described by exponential like laws, while r1/4 laws are found in ellipticals of high and low (e.g. M32) luminosity. Recently, Jerjen and Binggeli (1997) showed that brighter dEs gradually approach the luminosity profiles of Ellipticals, although they never become as curved as r1/4. A systematic homogeneous survey at faint isophotal levels, including all known types of low luminosity galaxies would be illuminating and moreover useful in understanding relations between dwarfs. For an example, see Fig. 8 where we show a luminosity profile of IZw18.
In general, the central parts of BCGs contain one or a few star forming knots (often found to be composed of many individual bright star clusters), which may be identified with the ``starburst'' region. The central knots in most cases give rise to excess surface brightness. To quantify the strength of the starburst the excess light may be integrated and compared to the underlying galaxy light. This gives rather modest starbursts in most BCGs, amounting to a brightening with typically less than one magnitude in the blue (Marlowe et al. 1999, Papaderos et al. 1996b). Given, that it must have a low mass to light ratio, this suggests that the starburst contains only a minor fraction of the integrated stellar mass, and that the subsequent fading in luminosity will be very moderate. Of course, some of these galaxies may already have passed their SFR peak and be in the process of fading. However, the amount of central excess depends on how luminosity profiles were constructed and to what depth the profiles are fitted, and moreover depends on an a priori assumption of the true shape of the underlying galaxy. If whatever creates the increased star formation does so, not only in the centre, but throughout the galaxy in question, the ``burst strength'' will be underestimated. Colour profiles can here yield useful additional information.