Well then, as in Astronomy we get an "exact science," because the facts and phenomena are on such a tremendous scale that we only see a minute portion of them--just a few details so to speak--and our ignorance therefore allows us to dogmatise; so at the other end of the scale in Chemistry and Physics we get quasi-exact sciences, because the facts and phenomena are on such a _minute_ scale that we overlook _all the details_ and see only certain general effects here and there. When a solution of cupric sulphate is treated with ammonia, a mass of flocculent green precipitate is formed. No one has the faintest notion of all the various movements and combinations of the molecules of these two fluids which accompany the appearance of the precipitate. They are no doubt very complex. But among all the changes that are taking place, one change has the advantage of being visible to the eye, and the chemist singles that out as the main phenomenon. So chemistry at large consists in a few, very few, facts taken at random as it were (or because they happen to be of such a nature as to be observable) out of the enormous mass of facts really concerned: and because of their fewness the chemist is able to arrange them, as he thinks, in some order, that is, to generalise about them. But it is certain as can be that he only has to extend the number of his facts, or his powers of observation, to get all his generalisations upset. The same may be said of magnetism, light, heat, and the other physical sciences; but it is not necessary to prove in detail what is sufficiently obvious. But now, roughly speaking, there is a third region of human observation--a region which does not, like Astronomy (and Geology), lie so far beyond and above us that we only see a very small portion of it; nor, like Chemistry and Physics, so far below us and under such minute conditions of space and time that we can only catch its general effects; but which lies more on a level with man himself--the so-called organic world--the study of man, as an individual and in society, his history, his development, the study of the animals, the plants even, and the laws of life--the sciences of Biology, Sociology, History, Psychology, and the rest. Now this region is obviously that which man knows most of. I don't say that he generalises most about it, but he knows the facts best. For one observation that he makes of the habits and behaviour of the stars, or of chemical solutions--for one observation in the remote regions of Astronomy or Chemistry--he makes thousands and millions of the habits and behaviour of his fellowmen, and hundreds and thousands of those of the animals and plants. Is it not curious then that in this region he is least sure, least dogmatic, most doubtful whether there be a law or no? Or, rather, is it not quite in accord with our contention, namely that Science, like an uninformed boy, is most definite and dogmatic just where actual knowledge is least.