In principle, galaxies can be classified in a two-dimensional diagram of total luminosity vs. surface brightness, where they concentrate in two different and ~ orthogonal branches. The spiral branch contains also the irregular galaxies, the dwarfs of various flavors, and the dwarf spheroidals. It is well separated from the elliptical branch, which contains also the cDs at one end and the low luminosity ellipticals at the other. Weird and rare galaxies, such as the low surface brightness objects and the "Malin I" types, as well as the BCDs, deviate by various amounts from these, apparently well-defined, sequences.
Problems appear when considering the proposed population of dwarf spirals, a class of ~ 6 kpc and low luminosity objects, or galaxies that appear elliptical but image processing reveals spiral arms. Are these the dwarf lenticulars (dS0) introduced by Sandage & Binggeli (1984)? Among the dEs there are at least two kinds of flavors: there are "disky dEs" and "bulgy dEs" according to the shape of their outer isophotes, and there are nucleated dEs (dE, N) as well as pure dEs. A revisit of the photographic material analyzed by Binggeli et al. (1985) yielded the interesting statistic that while 10% of the dEs at absolute magnitude -12 are nucleated, this fraction grows to almost 100% for dEs brighter than -16 (Binggeli et al. 2000).
Oh & Lin (2000) simulated the fate of globular clusters (GCs) around dwarf ellipticals and showed that they will eventually sink to the centers of these dEs, transforming them into the nucleated version. This process apparently takes a few Gyrs and happens mostly in the inner regions of a cluster. The sinking of GCs may explain also the off-center position found by Binggeli et al (2000) for the nuclei of dEs but contradicts (a) the nuclear magnitudes measured by Durret (1997), which at -10 mag are too bright to be GCs, as well as (b) the preliminary results from the HST snapshot program to evaluate the frequency of GCs near dEs (Miller et al. 1996). The latter found that dE, N galaxies tend to have more GCs than their non-nucleated cousins, rather the opposite of what one could expect.
The contribution of DGs to the total population is normally evaluated through the luminosity function. Here the literature sees significant confusion regarding the VC, primarily at the faint end of the distribution. While Caldwell & Armandroff (2000) claim a slope of -1.2 for the faint objects, Phillipps et al. (1998) estimate a faint-end slope of -2. At this meeting, Sabina Sabatini claimed an even steeper slope of -2.2 from a blind survey of some ten square degrees extending from M87 westward, significantly enhancing the number of low luminosity and low surface brightness galaxies in the cluster. Tully (private communication) supports a lower faint-end slope for the Virgo luminosity function and indicated that a flat low-luminosity end is seen also for the spiral-rich Ursa Major cluster at a distance from us that is similar to that of the VC. The differences in luminosity functions result from the difficulty of accounting properly for the contribution of background galaxies, in absence of redshift information, primarily for the low surface brightness galaxies.
Note also that, at present, there is no claim of any correlation between the proposed low surface brightness galaxies discovered by the Cardiff group in the direction of the Virgo cluster and any of the intracluster planetary nebulae (PN) candidates (e.g., Arnaboldi et al. 2002), although one would expect that with ~ 300 PNs and ~ 5 LSB galaxies per square degree, some PNs should be visible in these objects allowing an unbiased distance determination.
The dynamical properties of the different galaxy populations were recently studied by Conselice et al. (2001) for the VC core, a region six degrees in diameter around M87. This shows that the dEs form a population different from that of the other galaxies, or that they are a mixture of two different populations; one at the typical cluster redshift and another at a higher redshift band, about 1000 km/sec away from the cluster redshift. From this, Conselice et al. claim that the dEs were accreted by the VC not as dEs, but as a different type of progenitor galaxies (perhaps spirals, cf. Conselice et al. 2000). This was also proposed by Mao & Mo (1998) on theoretical grounds.
This general picture of DGs in the VC raises (at least) the following questions: (a) what is the relative importance of nature vs. nurture in DGs, (b) are the VC DGs the building blocks of galaxies or are they parts of large galaxies torn off during violent, cluster-related processes, (c) what happens to DGs as they accrete onto the VC, (d) what is the fate of the debris from disrupted DGs, and (e) what process triggers star formation (SF) in DGs. Some of these questions may have answers in the present workshop.