Sandage (1986) showed what the star formation rate (SFR) of galaxies of different Hubble type might look like. His curves are based on the study of Gallagher et al. (1984), who determined the SFR at three different epochs of galactic evolution. Sandage's figure establishes a connection of SFR and Hubble type but, as was already mentioned by Sandage, it “contains no physics”. In order to explore the background of this connection, however, it is necessary either to improve the observations or to model the evolution of galaxies self-consistently. However, the results of modelling the SFR are not reliable, if simplified models are used, which only describe some structural properties of galaxies. It is necessary to apply self-consistent models which take into consideration all relevant interaction processes between the gaseous and stellar components of a galaxy. Such models can be checked by comparison with observations like density and abundance distributions, star-gas content, velocities, velocity dispersions, mass-luminosity relations and age distributions of stars. A detailed model will show, whether the initial conditions, the feedback mechanisms during the evolution or the environment determine the evolution of a galaxy.