Several studies have quantified two aspects of the spiral sequence that determine the Hubble classification: the bulge to disk ratio and the pitch angle of the arms. Kennicutt 37 showed that objectively measured pitch angles correlate on average with RSA type, but with a large scatter; a rough correlation of pitch angle with Morgan concentration class was also found. Simien and de Vaucouleurs 57 used bulge/disk decompositions to illustrate how the bulge-to-total luminosity ratio varies with Hubble type. A smooth variation with type was found, but most importantly, this study found no support for van den Bergh's belief that SO's form a sequence parallel to spirals. The bulge contribution for SO's is generally intermediate between pure spheroidal systems and spirals, thus supporting Hubble's placement of SO's as transitions between E's and SO's.
The above correlations are expected because they underlie the classification system. But what really makes the Hubble system important are the correlations with other parameters that were not part of the system. It has been known for a long time that Hubble types, especially de Vaucouleurs revised Hubble types, also correlate very closely with objective measures of color, surface brightness, and HI content 22, 11. Late-type galaxies are bluer, have more hydrogen, and generally lower surface brightnesses than early-type galaxies. Intermediate types have intermediate properties between these extremes. These correlations are illustrated with new RC3 data and revised galactic extinction and tilt corrections in Figure 1. They are remarkable, at least for spirals, SO's, and irregulars, and they suggest that the revised Hubble sequence has physical significance 21. Possible scenarios for the origin of the Hubble sequence are discussed by Larson in this volume.
Figure 1: Dependence of several photometric parameters on the coded numerical stage in de Vaucouleurs system. The top two panels illustrate the correlations for colors, while the lower panels show the correlations for hydrogen index H. I. and mean effective surface brightness µ'eo, all as defined and corrected in the introduction to RC3.