in.] PERCOLATION OF WATER 85 



to supply the needs of the crop between the intervals of 

 rain. 



While the flow of water from a field-drain may be 

 taken as a rough measure of the amount of percolation 

 going on at any given time for that soil, the movement 

 may be followed more closely by means of a lysimeter 

 or drain-gauge. As a rule, the records of these instru- 

 ments are vitiated by the disturbance undergone by the 

 soil in filling them ; but the drain-gauges at Rothamsted 

 were constructed by building cemented walls round 

 blocks of earth in situ, and then gradually introducing a 

 perforated iron plate below to carry the soil. The 

 following diagram (Fig. 8) shows the mean monthly 

 records of rainfall and percolation through a depth of 

 20 inches and 60 inches respectively, over a period of 

 thirty-five years. 



It will be seen that of the total rainfall a little less 

 than one-half percolates through 60 inches of the 

 Rothamsted soil ; it should be remembered, however, 

 that the surface of these gauges is kept free from 

 weed or any growth. The total drainage through 

 20 inches of soil is practically the same as that through 

 60 inches, but rather a greater proportion of the rainfall 

 comes through the 60-inch gauge in the winter and 

 through the 20-inch gauge in the summer. In the 

 winter months the percolation reaches as much as 

 80 per cent, of the rainfall, in August little more than 

 20 per cent, of the rainfall finds its way through the 

 layer of soil. When the ground has become dried 

 to any depth in the summer, percolation may be much 

 hindered by the air within the soil and the want of a 

 continuous film of wetted surfaces to lead the water 

 down by surface tension. The top layer of soil becomes 

 thoroughly wetted and will not allow the air below to 

 escape ; only after some time are local displacements of 



