A galaxy classification system that is inadequately documented will not be well-understood by the people who use it. In the case of Hubble's revision, Sandage (1961) and Sandage and Bedke (1994) document the system with large numbers of high quality photographs, indicating what the notation means, variations within a class, and how effectively the system can handle the large number of distinct morphologies. Hubble's revision represents a point of view that has served extragalactic astronomy for a very long time, and the extent to which it has been illustrated guarantees that it will always be useful. However, the advantages of de Vaucouleurs' revision are in a way outweighed by the disadvantage that no major atlas has ever been prepared illustrating his point of view. If an observer wants to see what de Vaucouleurs means by a type such as (R)SB(rs)ab, they often have to refer to the Sandage atlases where the organization will not favor placing such a type into its appropriate context. Although for spirals the stage parameter (Sa, Sb, Sc, etc) will in general be consistent between the two systems, de Vaucouleurs has a clearly different point of view for Es and S0s. The Hubble classification of nonbarred S0s depends on the importance of dust, while the classification of SB0s depends on the appearance of the bar. In de Vaucouleurs' revision, the classification of S0s into early, intermediate, and late categories depends on the development of structure, particularly rings, and can have little to do with dust.

Another important area of disagreement is on the classification of rings. The (r) and (s) varieties are a well-established aspect of galaxy morphology that both the Hubble and de Vaucouleurs revisions recognize as a part of normal galaxy morphology. However, this does not mean that Sandage and de Vaucouleurs see these varieties in the same way. There is, in fact, considerable disagreement on this aspect of galaxy classification. For example, some galaxies classified as (r)-variety by de Vaucouleurs (1963) or in the Third Reference Catalog of Bright Galaxies (RC3, de Vaucouleurs et al. 1991) are classified as (s)-variety by Sandage and Tammann (1981, 1987; hereafter RSA) and Sandage and Bedke (1994). In the latter sources also, outer pseudorings are generally not recognized. The disagreement results because the spiral structure in many galaxies is very ring-like. In the Carnegie Atlas, strong "near rings" such as those in NGC 3081, 1433, 6782, and 7217 are noted but not formally recognized in the classification because the rings are made of tightly wound spiral structure. In the case of NGC 6782, SB94 state that "the `ring' is two inner spiral arms that start in the nucleus [hence the subtype (s) rather than (r)] and nearly overlap after each has unwrapped by about half a turn." Apparently, this effect is caused by a combination of the leading dust lanes in the bar and the tight spiral pattern that defines the inner ring in this case. In the case of NGC 3081, where the ``near ring'' has a very high contrast in the disk, SB94 state that the galaxy has "one of the most complex morphologies of the RSA galaxies. There are rings within rings at the edges of disks within disks, as intricate a structure as in nesting, concentric Russian dolls." None of these rings, however, is recognized in the Carnegie classification of SBa(s). The de Vaucouleurs classification of the galaxy is instead (R)SAB(r)0/a.

The virtue of paying attention to pseudorings was beautifully highlighted by M. P. Schwarz in a 1979 PhD thesis at Mount Stromlo Observatory. Using a test-particle dynamics code, Schwarz was able to simulate the morphology and other aspects of inner and outer rings and pseudorings in terms of gas dynamics at bar resonances. This was a real breakthrough for morphological studies that provided meaningful interpretations of some of de Vaucouleurs' classifications. The atlas will definitely cover this issue, as part of our goal of presenting modern interpretations of features.

The classification of magellanic spirals is an important contribution of de Vaucouleurs. The SB(s)m category has well-defined characteristics (de Vaucouleurs and Freeman 1973) including a single main arm and a characteristic asymmetry, as typified by the Magellanic Clouds. The late-type categories Sd, Sdm, Sm, and Im, and the extent to which the standard families and varieties exist at these types, will be well-illustrated in the atlas.