TOPOGRAPHIC IMPORTA^'CE . QF THE DEPOSITS 153 



;\-liich the stream flows is increased. There is also extension of the de- 

 posits during rainy years, when most of the "water evaporates lower in 

 the course than usual. 



The conglomerate is likewise of imjDortance in forming and preserving- 

 terraces. There are three causes to which the ten-aces may be due, 

 luimely, intermittent uplift of the land, climatic changes, and periodic 

 eliange in. the rate of erosion effected by the conglomerate itself. If it be 

 .supposed that upward movement of the land has taken place in late Qua- 

 ternary times, the upper terrace may be explained on the assumption that 

 the conglomerate pavement was formed as it is at present being formed, 

 that an uplift then occurred, causing the rate of erosion to increase until 

 tlie stream had reached a stage of readjustment by cutting down to the 

 level of the present floor, and that the deposition of the modern pavement 

 then commenced. Or it may be supposed that a period of heavy or medium 

 rainfall, during which the stream bed was widened and a great mass of 

 boulders and pebbles of all sizes were gathered along its course, was suc- 

 ceeded hx a period of aridity, during which the stream, having little 

 l^ower to erode and transport, became subject to evaporation and satura- 

 tion with mineral matter and the gravel was ceinented; and that there 

 followed a similar alternation of periods, resulting in the cutting of the 

 channel through the terrace pavement and the formation of the present 

 ]iavement at a lower level. The conglomerate itself and its influence on 

 erosion ma}' be considered a third factor in the production of the ter- 

 races. The stream restricts itself by tlie deposition of the conglomerate 

 in some parts of its course, the water becoming concentrated into a 

 smooth, narrow channel through the rock itself; its rate of flow is in- 

 ci-eased; the amount of Avater is increased, owing to the reduction of 

 evaporation; there is little chance for the deposition of gravel or of 

 cement, and the previously dominant work of building up gives place in 

 that part of the stream to erosion. When the pavement is cut through, 

 I he stream is superimposed in a sharp channel upon the softer underlying 

 sandstone. In this formation it cuts down and gradually widens out 

 until a flat valley bottom is restored and the process of deposition oi 

 cemented gravel is recommenced upon the new grade surface. 



It is improbable that any one of these three factors is unassociated 

 with the others in the formation of the terraces. The likelihood is greater 

 that the terraces have resulted from a complication of causes, one or the 

 other playing the chief part and the others acting as modifying factors. 



