LOSSES AND GAINS BY THE SOIL, 



273 



should be left for sufficient time to allow readjustment to take 

 place. 



All the water which falls on the soil in such an apparatus must 

 either percolate or evaporate. The percolation will therefore be 

 in excess of that in a free area, where a portion of the water runs 

 off. 



(2) The second method 1 consists in determining the quantity 

 and composition of the water from tile drains in the field. Tile 

 drains run when the water does not pass into the subsoil rapidly 

 enough to keep the soil from becoming saturated in the vicinity 

 of the drains. The quantity of drain water does not, therefore, 

 represent the quantity of water which percolates, but its com- 

 position should represent to a certain extent the composition of 

 the percolating waters. 



Quantity of Loss by Percolation. The composition of the 

 drainage water from some of the plots at Rothamsted 2 is given 

 in the following table : 



ANALYSIS OF DRAINAGE WATERS IN PARTS PER MILLION. 



[f we assume a downward percolation of 10 inches of drain- 

 age water, i part per million of water corresponds, in round num- 

 bers, to 2 l /4 pounds per acre per annum. According to this 

 estimate, from i l / 2 to 334 pounds phosphoric acid, 2 l / 2 to i2 l / 2 

 pounds potash, and 9 to 36 pounds nitrogen are lost in the drain- 

 age waters a year, according to the condition of the soil. 



At Rothamsted, the average amount of nitrogen contained in the 

 crops (wheat) in 30 years on the unmanured plots was 18.6 and 

 20.3 pounds ; the estimated loss by drainage from tiles, 10.3 and 



1 Jour. Agr. Sci., 1906, p. 377. 



2 Hall, An Account of the Rothamsted Experiments, p. 237. 



