THE UPPER MISSOURI RIVER SYSTEM. 599 



on the globe does there exist a better example from wbicli to study 

 these principles of surface erosion than in the Upper Missouri River 

 system. This Avill be best seen when we consider a little more closely 

 the proper bed of the river. 



For two or three months of each year, between March and June, 

 the river is high, and this state of high water is tolerably uniform 

 from year to year, so as to be in a manner normal. Supplied chiefly 

 from melting snows at greater and greater altitudes as the season ad- 

 vances, it persists with only slight fluctuations until the supply is ex- 

 hausted, when the water slowly falls to its low-water mark, where it 

 remains the rest of the year with only a small amount of variation, 

 because the rainfall is so light. There thus exist two distinct and 

 somewhat uniform conditions of the water, each occupying its regular 

 part of the year. Owing to this regularity of high water, the maxi- 

 mum bed of the river produced by it is somewhat uniform and clearly 

 marked, while it also bears a tolerably uniform relation to the deeper 

 channel represented by the low-water state. Examined in time of low 

 water, this river-bed seems to be three or four times as wide as the 

 river itself. The stream, then, usually flows in serpentine curves 

 which cross and recross the bed. The bed itself is also crooked much 

 as is the channel, only its curves are as much longer as it is wider. 

 The whole valley is usually also winding with much more ample 

 curves, and the river-bed crosses and recrosses it in a manner similar 

 to that in which the channel crosses and recrosses the bed. The river 

 itself generally hugs one of the banks of the bed, but it is always at a 

 curve, or bend, such as will tend to wear the bed on the convex side 

 and thus render it more crooked. The distance traversed by the chan- 

 nel in crossing from one side of the bed to the other is small, compared 

 with the distance traversed while in close contact with the bank of the 

 river-bed, which it is perpetually extending into the general valley. 

 The reason why it does not constantly grow wider is, that on the 

 abandoned side the surface is being constantly raised by deposits of 

 material which the water, more sluggish on this side, can no longer 

 hold. As the river shifts its position in the valley, a strip of land of 

 varying width is formed each year to be gradually assimilated to the 

 permanent valley (see Diagram No. II). . 



If, now, we take the more general view and regard the entire valley 

 as one homogeneous product, we can better study the process by which 

 it has been formed. Beginning with the channel of the river we shall 

 find that, except where crossing the bed, its cross-section presents a 

 figure approaching more or less closely to a right-angled triangle with 

 the right angle at the bottom, or deepest place. One side will then be 

 formed by a steep wall or bank, which may become perpendicular 

 above the surface of the water, but is not usually so below. The other 

 side of the triangle represents the general bottom of the river, which 

 gradually grows more shallow toward the remote side of the river-bed. 



