On the liaiii and Drainage - Waters at Rothamxtcd. 269 
Part II. Tiik Amount and Composition of the Drainage- 
Waters FROM UNMANURED FALLOW LaND. 
In any inquiry respecting the inOuonce of drainage in practical 
agriculture, it is clearly of primary importance to ascertain what 
proportion of the rainfall passes in each season through known 
depths of soil ; we shall therefore in the first place describe the 
experiments relating to this part of the subject. 
The amount of drainage-water passing through any soil 
depends — 1. On the amount of the rainfall. 2. On the 
physical condition of the soil, its permeability, and water-hold- 
ing power. 3. On the amount of evaporation taking place, 
which latter is determined by the temperature of the soil and air, 
and by the capillary power of the soil, and is greatly increased 
when a crop is growing on the surface. 
The experiments we are about to describe were made to 
ascertain the amount of natural drainage through the Rothamsted 
soil when kept bare of vegetation ; the effect of a crop is thus 
for the present excluded. The drainage-waters obtained have 
been analysed ; their composition will be found to illustrate 
in a striking manner the loss of plant food which an uncropped 
soil may suffer from the percolation of rain-water. 
1. The Drain-Gauges. — The drain-gauges which have been 
constructed are three in number ; they consist of rectangular 
plots of soil, each 6 feet by 7 feet 3 inches, having thus the 
same surface as that of the large rain-gauge, namely 1-lOOOth of 
an acre. The depth of the soil varies. In the first gauge the 
depth is 20 inches ; in the second it is 40 inches ; and in the 
third GO inches. 
In order to obtain a natural drainage, it was of primary 
importance that the soil should be in a perfectly natural condition 
of consolidation, neither more porous nor more condensed than 
ordinary field soil. To accomplish this object a deep trench 
was dug along the front of each intended gauge ; the mass 
of soil was then gradually undermined at the depth previously 
determined, and plates of cast-iron, 8 inches wide and perforated 
with holes, were introduced to support the soil as the work 
proceeded. This perforated iron bottom was finally strengthened 
by transverse iron girders, and the ends of the plates and girders 
supported by brickwork on three sides of the intended gauge. 
The soil being now supported from beneath, trenches were made 
one by one on the three remaining sides of the block of soil to 
be isolated ; walls of brick, laid in cement, 4i inches thick, were 
built against the soil, and the trenches were again filled in with 
earth. The mass of soil was in this manner built in on all 
sides with brick and cement. The surrounding walls were 
