My paper should, I am aware, be devoted to a statistical and social review. I have discovered, however, that the subject is eminently one which does not lend itself to statistics, and I propose instead to treat it from ...