Page:Popular Science Monthly Volume 57.djvu/172

162 been exactly what it was. . . . Many highly valuable compendia of Physical Geography, for the use of scientific students of that subject, are extant; but in my judgment most of the elementary works I have seen begin at the wrong end, and too often terminate in an ominum gatherum of scraps of all sorts of undigested and unconnected information; thereby entirely destroying the educational value of that study which Kant justly termed the 'propædeutic of natural knowledge.'" (Preface to 'Physiography,' 1878).

Here we find clear recognition of the need of introducing a consideration of causes, just as was urged by Guyot; and furthermore a recognition of the need of linking together in their natural relations all the items which together constitute the content of the subject. It may, however, be contended that the attempt to combine in a single course of study the elementary principles of chemistry and physics, of geology and astronomy, along with those of physical geography, is not practicable from an educational point of view; such a combination will not secure either the clear knowledge or the strong discipline that can be derived from systematic courses in two or three of these subjects, presented separately. Text-books like Hinman's 'Eclectic Physical Geography' and Mill's 'Realm of Nature,' in both of which a broad range of other than geographical subjects is covered, do not seem to-day to be in so much favor as those books which attend more closely to the true content of our subject. Indeed, with respect to physical geography, considered from the scientific and educational point of view, a report on College Entrance Requirements, recently published by our National Educational Association, presents the best definition and outline of the subject that has yet appeared. It advises the omission of irrelevant matter, however interesting such matter may be in itself. The principles of physics and the succession of geological formations with their fossils, the classification and distribution of plants and animals must be taught elsewhere; but much profit may be had from terrestrial phenomena by which the principles of physics are illustrated, and from the consequences of past geological changes in determining present geographical conditions, and especially from the physiographic controls by which the distribution of organic forms is determined.

The general scheme under which all land forms may receive explanatory description must consider chiefly the movement and erosion of the earth's crust. Deformation offers a part of the earth's crust to be worked upon. Various destructive processes of erosion work upon the offered mass, and the streams, with their transported waste, follow the depressions in the carved surface. So important is the element of erosion, and so leading is the part played by rivers in erosive