GROUND, ROCK, WATER 141 



out, whereas immediately downstream from it there may be a shallow. 

 If, however, the bowlder or ledge serves as a dam and the water falls over 

 it, then on its downstream side will be a pool cut by the fall. (See 

 Plate 24.) 



Farther down the stream some of the material will be deposited 

 which was cut above, and the shape of the river bank will be modified 

 by the formation of bars. In general the larger material is deposited 

 first, as the current slackens, so that gravel bars may lie in a consider- 

 able current, but sand bars only in comparatively slack water. In its 

 lower reaches, a stream is constantly depositing material at one place, 

 but eroding at another place the material which it has previously 

 deposited. The stream tends to swing from side to side of its bed, 

 rebounding as it were from shore to shore as it progresses. The points 

 of its greatest friction with its banks are the points where its current 

 is thus deflected. The current tends therefore to lie on the outer side 

 of each curve of the stream. The steep newly cut banks and the deep 

 water will be on this outer and convex side and the flat newly formed 

 bars reaching gradually into the stream will form points on the concave 

 side, or extend slightly downstream from such points. (See Plate 

 25.) In very small streams, these effects are likely to be offset by 

 variations in material or even by growth of trees and shrubs on the 

 banks, but where a designer is endeavoring to give character to a small 

 artificial stream, some exaggeration of these essential characteristics, 

 which by association will lead the observer to think of larger streams, 

 will give his work individuality and verisimilitude. Even when in 

 reality the current is sluggish and could not cut for itself any consider- 

 able bed, it may be effective to model the bed as though it were the 

 work of a powerful current which might be imagined to run in the 

 Spring, of which the present trickle from pool to pool might seem only 

 the diminished Summer flow. In designing artificial brooks or rivers it 

 should be remembered that the essence of a stream is continuity. Any 

 water-body simulating a natural stream should have an obvious place 

 whence it apparently comes and a place into which it apparently goes ; 

 and with streams as well as with lakes, this effect of continuity may be 

 produced by the extension of the water beyond the portion seen, and the 

 consequent suggestion of greater extent than actually exists. 



