Rivers as Regional Bonds: The Columbia-Snake Example

Abstract

IN 1900, William Morris Davis warmly approved a statement made by Guyot half a century earlier that it was not enough describe, without rising to the causes, or descending to the consequences.' The causes of rivers have been studied in detail by geomorphologists and others, but all too often their consequences have been left to the inadequate treatment of novelists and historians. This paper represents an attempt to rescue rivers from the geomorphologists on the one hand and the romancers on the other.2 Rivers have a definite relation to man, though it varies from place to place and from time to time. Nevertheless, their consequences, like their causes, are susceptible of systematic analysis. Rivers are more than conduits for water of a certain length, width, depth, and speed, for consumption, irrigation, power, or navigation. Along most of their courses they are hindrances to local cross-country movements, and sometimes they form effective international barriers. More generally, however, rivers serve as regional bonds, because roads and railroads parallel their easy gradients and population is concentrated along their level or fertile valleys. Two spectacular types of rivers exemplify this role of regional connection: rivers flowing through mountain ranges; and rivers flowing across deserts. The Potomac is an example of the former, the Nile the classic example of the latter. The Potomac in places provides a water-level route for the twoand three-track main line of the Baltimore and Ohio Railroad on