158 
INVESTIGATIONS ON ROTHAMSTED SOILS. 
We see that in the first 9 inches the leguminous soils are uniformly 
very markedly richer in nitrogen than the wheat-fallow soils. The 
same thing is in nearly every case shown in the second 9 inches, and 
for the most part also in the third 9 inches. 
If we look at the line in the lower part of the table showing the 
nitrogen per acre for the aggregate first three depths (1 to 27 inches), 
the nitrogen accumulation in the leguminous land, as compared with 
the wheat land, assumes large and very striking dimensions, except 
on one plat (Vicia sativa, plat 6, 1883), where there appears to be 
throughout some irregularity of subsoil. Even here, however, the 
accumulation is well seen in the surface soil. 
In some plats the influence of the vegetation appears to be recog- 
nizable lower still, but the natural irregularities of the lower subsoil 
render it impossible here to draw certain conclusions. But the gen- 
eral evidence of accumulation in the upper layers is most clear and 
convincing, and accords admirably with modern knowledge of the life 
history and functions of leguminous crops — knowledge which was but 
dimly toreseen, although its advancing shadow was felt when these 
experiments were started. It was not until 1886 that Hellriegel 
announced the results which gave definite shape and direction to the 
long- vexed problem relating to the sources of the nitrogenous food of 
this highly interesting natural order of plants. 
In the first 27 inches of soil we find that the ten sets of samples 
representing the leguminous plats average 0,604 pounds of total nitro- 
gen per acre, while the three sets of wheat soils average but 5,847 
pounds, showing an average gain of 757 pounds of nitrogen per acre 
under the influence of leguminous vegetation. Sir Henry Gilbert, in 
his lectures, further pointed out that the annual output of nitrogen in 
the crops had been far greater on the leguminous land. During the 
twenty-seven years before the miscellaneous leguminous crops were 
first planted, and when the leguminous land was under red clover, it 
yielded in its clover crops a yearly average of 32 pounds of nitrogen 
pei- acre, as against L5 pounds per acre yielded in the crops of the 
wheat and fallow land, while during the following fourteen years of 
various leguminous herbage the average yearly output of nitrogen in 
crops was 11 pounds per acre for the red-clover plat, 24 pounds for 
the white clover, 75 pounds for the vetches, 58 pounds for the Bokhara 
clover, an<l no Less than L37 pounds for the lucern. During the same 
time the wheat and fallow land yielded only L2 pounds of nitrogen per 
acre per annum in its crops. No nitrogen was nianurially applied on 
any of these plats. 
We nave, therefore, not only soil accumulation in the root residues, 
bill also a great output of organic nitrogen in the crops (varying with 
the species and habits of the individual Leguminosee grown), a Large 
pari of which nitrogen must undoubtedly have been due to the fixa- 
tion of atmospheric nitrogen. 
