Page:The National geographic magazine, volume 1.djvu/48

32 duced above that such a classification is objectionable unless the greater geographic elements due to diastatic movements (in which the constructive action is veritable but different in kind from those in the other categories) be excluded, and this is impracticable without limiting the classification to subordinate phenomena. Moreover it is illogical and useless to unite the constructive phenomena of the remaining categories, since (1) the processes exemplify widely diverse laws, which must find expression in any detailed classification whether genetic or not, and since (2) the differences between the forms united are much greater than the differences between the forms separated in such a classification—e. g. the differences between a dune, a drumlin and a mesa (all constructive forms) are far greater than the differences between a fresh lava sheet and a deeply cut mesa, between a drumlin and the smallest drift remnant, or between a dune and a Triassic mound of circumdenudation; and this is true whether the distinction be made on analogic, homologic, or genetic grounds. Indeed it seems evident that while discrimination of constructive and destructive forms is necessary and useful in each genetic category, the use of this distinction as a primary basis of classification is inexpedient.

The classification of topographic forms proposed a few years ago by Davis, who regards "special peculiarities of original structure" as a primary, and "degree of development by erosion" a secondary basis, and Richthofen's arrangement of categories of surface forms as (1) tectonic mountains, (2) mountains of abrasion, (3) eruptive mountains, (4) mountains of deposition, (5) plains, and (6) mountains of erosion, in addition to depressions of the land (Die Hohlformen des Festlandes), are more acceptable, since they are based in part on conditions of genesis. But it is clearly recognized by modern students of dynamic geology that waterways are the most persistent features of the terrestrial surface; and the most widely applicable systems of classification of the surface configuration of the earth thus far proposed have been based substantially on the agencies of gradation. Thus Powell, Löwl and Richthofen classify valleys by the conditions of their genesis; Gilbert classifies drainage; and Phillipson, unduly magnifies the stability and genetic importance of the water parting, classifies the hydrography through