The task of summarizing the views of any sociologist seems to me to be rendered difficult by a certain vagueness in outline of the positions laid down, a certain tendency for a description of fact to run into an analogy, and an analogy to fade into an illustration. It would be discourteous to doubt that these tendencies are necessary to the fruitful treatment of the material of sociology, but, as they are very prominent in connection with the subject of gregariousness, it is necessary to say that one is fully conscious of the difficulties they give rise to, and feels that they may have led one into unintentional misrepresentation. With this proviso it may be stated that the writings of Ward produce the feeling that he regards gregariousness as furnishing but few precise and primitive characteristics of the human mind. The mechanisms through which group “instinct” acts would seem to be to him largely rational processes, and group instinct itself is regarded as a relatively late development more or less closely associated with a rational knowledge that it “pays.” For example, he says: “For want of a better name, I have characterized this social instinct, or instinct of race safety, as religion, but not without clearly perceiving that it constitutes the primordial undifferentiated plasm out of which have subsequently developed all the more important human institutions. This . . . if it be not an instinct, is at least the human homologue of animal instinct, and served the same purpose _after the instincts had chiefly disappeared_, and when the egotistic reason would otherwise have rapidly carried the race to destruction in its mad pursuit of pleasure for its own sake.”[G] [G] “Pure Sociology,” p. 134. Italics not in original. Passages of a similar tendency will be found on pp. 200 and 556. That gregariousness has to be considered amongst {26} the factors shaping the tendencies of the human mind has long been recognized by the more empirical psychologists. In the main, however, it has been regarded as a quality perceptible only in the characteristics of actual crowds—that is to say, assemblies of persons being and acting in association. This conception has served to evoke a certain amount of valuable work in the observation of the behaviour of crowds.[H] Owing, however, to the failure to investigate as the more essential question the effects of gregariousness in the mind of the normal individual man, the theoretical side of crowd psychology has remained incomplete and relatively sterile.