The astrophysical context of diffusion in stars

Some arguments are presented in support of the concept of diffusion being responsible for abundance anomalies in A and B stars. It is shown how chemical separation of elements solves some puzzling astrophysical problems which arise if one assumes that all the observed abundances are caused, without alteration, by nuclear reactions, and then, how element separation fits in with our understanding of the stability of the outer regions of stars. The creation of spots or, rather, rings by diffusion is discussed, and is shown to depend on a number of insufficiently known parameters, which probably vary from star to star. This may well explain the variety of observed abundance patterns.