whose genesis is attributed to retrogressive (rückwärts fortschreitende" or "rückschreitende") erosion, is vaguely separated into several ill-defined classes and sub-classes determined by structure, climate, and various other conditions. The second of Löwl's categories is also recognized by Phillipson. Still more recently, Richthofen, neglecting antecedent drainage, designated the superimposed class of Powell epigenetic, and formulated a classification of the remaining types of continental depressions (Die Hohlformen des Festlandes) as (a) orographic depressions (Landsenken); (b) tectonic valleys, and (c) sculptured valleys; and the last two categories are separated into classes and sub-classes, corresponding fairly with those of Löwl, determined by their relations to structure and by various genetic conditions.
These several classifications have much in common; their differences are largely due to the diversity of the regions in which the investigations of their respective authors have been prosecuted; but combined they probably comprehend all the topographic types which it is necessary to discriminate.
The American classification and nomenclature, particularly, is unobjectionable as applied to montanic hydrography; but it does not apply to the perhaps equally extensive drainage systems and the resulting topographic configuration developed on emergent terranes either (a) without localized displacement or (b) with localized displacement of less value in determining hydrography than the concomitant erosion, terracing and reef building; neither does it apply to the minor hydrography in those regions in which the main hydrography is either antecedent or consequent; nor does it apply even to the original condition of the superimposed or antecedent drainage of montainous regions.
Upon terranes emerging without displacement and upon equal surfaces not yet invaded by valleys, the streams depend for their origin on the convergence of the waters falling upon the uneroded surface and affected by its minor inequalities, and for their direction upon the inclination of that surface. They are developed proximally (or seaward) by simple extension of their courses by continued elevation, and distally by the recession of the old and the birth of new ravines; and since in the simple case it follows from the law of probabilities that the receding ravine will retain approximately the old direction and that the new ravines will depart therefrom at high angles, the drainage systems thus independently developed become intricately but systematically ramified and more or less dendritic in form. Löwl, Phillipson, Richt-
