Evidence for the Neutral Hypothesis of Protein Polymorphism