is the portion, however, in which he least allows himself to be guided by his fundamental conception of psychology as a natural science; it is also the portion in which he makes most use of his comprehensive learning and powers of acute analysis, on terms common to himself with all students of modern empirical psychology. To be sure, he starts the discussion with the very doubtful assumption, in which he agrees with Dr. Ward, that "extensity," or "bigness," and that in three dimensions, belongs to our most primitive sensation-states. And although he appears far better acquainted than Dr. Ward with the experimental data on which writers like Stumpf and Hering endeavor to place this view, he makes comparatively little use of these data. Their claims for "extensity" as a primitive quality of all sensations are almost wholly based upon an appeal to adult self-consciousness, — a form of appeal which I suppose nearly all expert writers regard as entirely irrelevant and inadequate. Certainly this appeal is no substitute for that minute analysis of the factors, and stages in growth, of the perceptive power of mind, in which Wundt and Helmholtz and many others have been so successful; and in which I have tried to take some small part. For the view, however, which regards "extensity" as an acquired quality of certain sensation-states, Professor James — in his customary off-hand way — declares (II, p. 31), "there is not a vestige of evidence." Moreover, the opinion with which it hangs together — namely, that our sensations are originally devoid of all spatial content — he confesses himself "wholly at a loss to understand." This seems a pity, because the opinion probably represents about nine-tenths, or at least four-fifths, of the most patient, intelligent, and elaborate work which the science of psychology has done upon the subject of perception by the senses, during the last quarter of a century.
We are not concerned at present, however, to discuss with Professor James his theory of "The Perception of Things." We wish only to remark in passing that almost all the points which are here treated, in any full and satisfactory way, avail themselves of the same inductive conclusions and meta- -
