Galaxy Morphology and Classification

1.2 Galaxy Types that Do Not Fit Into the Hubble Scheme

In 1936 it was believed that the ``tuning fork'' scheme described in Hubble's (1936) influential book The Realm of the Nebulae provided a complete framework for the morphological classification of normal galaxies. However, the discovery of the Sculptor and Fornax dwarf spheroidal galaxies by Shapley (1938, 1939) provided the first hint that the Hubble tuning fork diagram did not give an exhaustive description of the entire realm of galaxy morphology. The relation between the elliptical (E) and dwarf spheroidal (dSph) classes of galaxies has remained a subject of lively controversy (Jerjen & Binggeli 1997). Kormendy (1985) and Wirth & Gallagher (1984) regarded dwarf and giant ellipticals as physically distinct classes of objects. whereas Bingelli, Sandage & Tarenghi (1984) emphasized the continuity between the characteristics of giant and dwarf ellipticals. However, Kormendy (1987b) argues that the effects of seeing on ground-based observations of objects in the distant Virgo cluster may have blurred the distinction between elliptical and dwarf spheroidal galaxies. Gorgas et al. (1997) have found some evidence which suggests that the Mg/Fe ratio in the brightest (dwarf) spheroidal galaxies in the Virgo cluster is near solar. In this respect these spheroidal galaxies therefore resemble the disks of galaxies from S0 to Sc which have solar Mg/Fe ratios (Sil'chenko 1993), whereas ellipticals have high Mg/Fe values. This difference suggests that either (1) star formation started with a bang in E galaxies, but proceeded more slowly in the disks of S0-Sc galaxies, or (2) the mass spectrum with which stars were formed in ellipticals contained a larger fraction of massive objects that produced supernovae of Types Ibc and II on a short time-scale.

The class of cD galaxies identified by Morgan provides another example of a classification type that does not fit comfortably into the 1936 Hubble scheme. The fact that most cD galaxies only occur close to the centers of rich clusters of galaxies shows that galaxy morphology, and hence galaxy evolution, may be strongly affected by environment. Another class of objects that does not appear to fit naturally within the Hubble scheme is the amorphous (Am) galaxies (Sandage & Brucato 1979, Gallagher & Hunter 1987), some of which have also been called Ir II by Holmberg (1958). A complication is that galaxies such as NGC 3034 (= M82) and NGC 3077 probably started out as objects of quite different morphological types that subsequently evolved into Am galaxies by tidal capture of large amounts of gas. Furthermore the morphological peculiarity of some galaxies is, no doubt, due to the recent tidal capture of companions. Good examples are NGC 2685 (which Sandage (1961) refers to as ``perhaps the most peculiar galaxy in the Shapley-Ames Catalog'') and NGC 5128 = Centaurus A. Surprisingly, Hubble did not extend his ``tuning fork'' diagram to include irregular galaxies, even though such objects exhibit the same dichotomy that is observed among spirals. The Small Magellanic Cloud is, for example, a normal irregular whereas the Large Magellanic Cloud is a barred irregular. Also Hubble never gave an objective criterion for distinguishing very late-type spirals from irregulars. Another weakness of the Hubble system is that its classification of elliptical galaxies along the sequence E0-E7 is a function of both the intrinsic shape of these objects and the direction from which they are being viewed. Van den Bergh (1976b) has noted the existence of ``anemic'' spiral galaxies, which appear to be spirals that have been modified by the cluster environment. Furthermore it is found that galaxies in the cores of rich clusters, such as A957 (Abraham et al. 1994), do not fit comfortably into the Hubble scheme, which grew mainly out of experience with classification of galaxies in the field and in small groups. Finally, observations with the Hubble Space Telescope show that the fraction of galaxies that fit comfortably into the Hubble system drops precipitously with increasing look-back time (Abraham et al. 1996a, van den Bergh et al. 1996).