When men of science (or, more often, men who talk about science) speak of studying history or human society scientifically they always forget that there are two quite distinct questions involved. It may be that certain facts of the body go with certain facts of the soul, but it by no means follows that a grasp of such facts of the body goes with a grasp of the things of the soul. A man may show very learnedly that certain mixtures of race make a happy community, but he may be quite wrong (he generally is) about what communities are happy. A man may explain scientifically how a certain physical type involves a really bad man, but he may be quite wrong (he generally is) about which sort of man is really bad. Thus his whole argument is useless, for he understands only one half of the equation.
The drearier kind of don may come to me and say, "Celts are unsuccessful; look at Irishmen, for instance." To which I should reply, "You may know