206 NATURE AND PROPERTIES OF SOILS 



grade from 10 to 12 inches to the 100 feet. It is usually 

 formed by back-furrowing and scraping. The interval be- 

 tween the embankment depends on the slope. Since the terrace 

 is low and broad, it may be cropped without difficulty and 

 offers no obstacle to cultivating and harvesting machinery. 

 It wastes no land, and eliminates breeding places for insects. 



Small gullies, while at first insignificant, soon enlarge into 

 deep unsightly ravines. While they may be plowed-in or 

 otherwise filled up, such a procedure is generally a waste of 

 time, since the gullies form again with the next heavy wash. 

 A number of different methods are in use for the control of 

 gullying, depending on conditions. Staking is a very common 

 procedure, the size of the stakes increasing with the magni- 

 tude of the gully. The stakes are usually interwoven with 

 brush, although stone, straw, and other material may be 

 utilized. If brush or other loose material is used, it should 

 be staked to the ground or held down by stone or dirt. Other- 

 wise, the water will run beneath the fill and no benefit will 

 result. Dams of earth, concrete, or stone are often installed 

 with success. They must be supplemented by a tile-drain 

 outlet, however, with an elbow just above the dam. The dam 

 checks the water until it rises to the level of the elbow outlet 

 and is then carried away through the tile. Most of the sedi- 

 ment is deposited above the dam and the gully is slowly filled. 



109. Percolation losses and their control. — When at 

 any time the amount of rainfall entering a soil becomes greater 

 than its water-holding capacity, losses by percolation will 

 result. The losses will depend largely on the amount and 

 distribution of the rainfall and the capability of the soil to 

 hold moisture. The objectionable features of excessive per- 

 colation are two: (1) the actual loss of water, and (2) the 

 leaehing-out of salts that may function as nutrients to plants. 



The results from the Rothamsted lysimeter^ from 1871- 



^Hall, A. D., The Boole of the Rothamsted Escperiments, p. 22, New 

 York, 1917, 



