It is well known that for a single, autosomal locus with differential viabilities convergence to a stable equilibrium is assured. However if new alleles are introduced sequentially by mutation the final equilibrium may depend on the particular sequence. Thus the equilibrium to which the population moves will be randomly determined. When there is an internal, stable polymorphism there is no such randomness in the final equilibrium as there is only the one stable point. However the time taken to reach that equilibrium will vary with the order of occurrence of mutations, and with the specific set of viabilities. The object here is to discuss the possible routes by which an internal polymorphism can be reached, and to begin the elucidation of which sets of alleles can be stable in their own space.