selection; but it will itself form the basis of favorable variations without which natural selection can do nothing. It seems to me easy to understand how survival of the fittest may result in progress, starting from such functionally-produced gains: but impossible to understand how it could result in progress if it had to start from mere accidental structural increments due to spontaneous variation alone.
Thus it becomes clear why certain countries have by mere geographical position necessarily produced certain high types of human intelligence, while in certain other countries the race has never progressed beyond a very low level. There are places like Central Africa, where the physical conditions do not tend to produce any great diversity of occupation; and here the general average of intelligence does not tend to rise high. On the other hand, there are places, like Greece, Italy, the West European peninsulas and islands, where the physical conditions tend to differentiate the population into many groups, agricultural, mercantile, sea-faring, military, naval, and professional; and here the general average of intelligence tends to rise very high indeed. Of course, one must allow much influence to the time-element; for every such increase in differentiation involves yet further increases in the sequel, and brings the social organism, or parts of it, into contact with new environments. The Ægæan is not now of the same importance in this respect as during the days when coasting voyages from island to island were the utmost possible stretch of navigation: the science acquired there has widened the sphere of navigation itself, first to the entire Mediterranean, then to the open Atlantic, finally to all the oceans of the whole earth. But in principle it has always seemed to me (as against the really accidental view advocated by Mr. Bagehot) that the "philosophy of history," the general stream of human development, could be traced throughout to perfectly definite physical causes of this sort. Mr. Bagehot, basing himself on the pure Darwinian theory of spontaneous variations, believed that the differences between races of men were due to mere minute physical sports in their nervous constitution: it appears to me rather that they are due to the action of a definite environment, thus effecting a differentiation of circumstances, and in many cases calling into constant functional activity the highest existing faculties of the various social units in the most diverse ways. We may not thus (though vide post) be able to account for the particular character and genius of a Pericles, an Aristotle, a Hannibal, a Cæsar, a Newton, or a Goethe; but we can thus at least account for the general average of intelligence which made Greece, or Carthage, or Rome, or England, or Germany, capable of producing such an individual, as a slight variation on the common type, due to the convergence of separately rich and varied lines of descent. The real illuminating point is this—that such men do arise from time to time among the most intelligent na-
