650 THE AMERICAN JOURNAL OF SOCIOLOGY

such as the cutting of the Suez canal, social facts; yet both undoubtedly deserve recognition in a philosophical statement of all the determining conditions in these two branches of knowl- edge.

The sort of conditions which I mean by the phrase "extra- social" will appear from the enumeration below. It does not claim to be complete, however. Their full discussion does not come within our province, seeing that they are extra-psycho- logical.

I. Group-selection (described above. Sec. 120). — In group- selection we have a condition of enormous importance in the development of social aggregations, especially in the instinctive and spontaneous periods ; that is, of so-called " companies." It holds, however, for all societies when the conditions are such that groups as groups come into competition. Not only real war, but commercial and social wars of all kinds, illustrate group- selection. The working of the principle is strictly analogous, indeed identical, with that of natural selection in biology, an anal- ogy excellently worked out by Bagehot in his remarkable work, Physics and Politics. It is one of the foundation stones also of S. Alexander's work. Moral Order and Progress. Bagehot acutely recognizes the distinction, without explicitly drawing it, between group-selection as a condition of evolution in the earlier stages of human aggregation and the operation of the real social force of "discussion" (described above under the heading "generali- zation") in the higher forms. It is, moreover, an additional proof that group-selection is a condition, and not a social force, that there is this difference between the lower and the higher ; for the lower are determined, as we have seen, very largely by bio- logical principles, such as instinct and physical heredity, and do not involve the social progress which the operation of the psy- chological forces brings in later on. Yet it is just there that group-selection is all-important.'

^ The corresponding truth has often been pointed out (see QoVY.^ Primary Factors of Evolution, chap. 7; Cattei.l, Science, N. S., Vol. Ill, p. 668; Baldwin, Psychological Hcview, Vol. IV, 1897, p. 219) that natural selection in biological evo- lution is not a force or cause, but a condition. Spencer's phrase, "survival of the

fittest," itself analyzes natural selection. The fitness is assumed ; it is due to earlier