ON THE EVOLUTION OF A BIOSYNTHETIC PATHWAY