68 INVESTIGATIONS ON ROTHAMSTED SOILS. 
case of four of the other plats, with the results for the first 27 inches. 
We may also here consider the rape-cake plat. It is convenient to 
tabulate the results separately. 
Table 34. — Broadbalk wheat soils, plats 3, 4, 2a, 2b, and Id— Nitrogen as 
nitrates ( " nitric,'" nitrogen) in samjrfes collected in October, 1893. 
Depth. 
Plat 3, un- 
manured 
5u years. 
Plat 4, un- 
manured 
since 1852. 
Plat 2a, 14 
tons farm- 
yard ma- 
nure, 9 
years. 
Plat 2b, 14 
tons farm- 
yard ma- 
nure, 50 
years. 
Plat 19, 
rape-cake 
plat. 
Second 9 inches 
Third 9 inches 
1 to 27 inches 
Pounds. 
9.64 
9.22 
2.74 
21.60 
Pounds. 
7.96 
7.03 
2.12 
17.11 
Pounds. 
23. 67 
22.65 
10.27 
56.59 
Pounds. 
10.53 
45.36 
12.25 
68.14 
Pounds. 
22.06 
24. 36 
14.18 
60.60 
We see that even on the land wholly unmanured for fifty years, 
notwithstanding the recent removal of the fiftieth continuous wheat 
crop, the uppermost 27 inches contained 21.60 pounds of nitric nitro- 
gen per acre, equivalent to 1^ hundredweight per acre of commercial 
nitrate of soda. Of this the greater part is in the upper 18 inches. 
Unfortunately this autumnal store of nitrates is destined for the 
most part to be washed away before the young wheat plant is old 
enough to appropriate much of it, and it is mainly on the nitrates 
formed in the spring and early summer that the crop lives. Of course 
in a fairly dry winter much of the autumn nitrates may be retained, 
and when this is the case the crop is benefited by food that in a wet 
winter would be wasted. It is, however, remarkably interesting to 
find that on a soil absolutely unmanured and continuously cropped 
for fifty years there should still be found within 18 inches of the sur- 
face, from two to three months after harvest, as much nitrate as 
would be annually carried oft 7 in a year's crop. 
Ii will be seen later that the quantity found in the surface soil is 
pract ically the same as at the same time of year in 1881, twelve years 
earlier, and the quantity found in the first 27 ifoches is not very dif- 
ferent; and if we allow for the fact that the soil had in 1881 been very 
heavily rained on between harvest and the time of sampling the gen- 
eral similarity of the results is increased. 
Plat 4, though unmanured since 1852, was in its earlier years man- 
ured wil h superphosphate and ammonium sulphate. Although so many 
years have elapsed, plat 4 has (see Table 1) on the average yielded 
distinctly larger crops than plat 3, its superiority in yield of grain 
having persisted longer than in that of straw. In 1893, which was 
a very bad harvest year owing to summer drought, plat 4 had yielded 
2 bushels per acre more grain than plat 3, and it is noteworthy that 
we find in it some 41 pounds less of nitric nitrogen in the first 27 
Inches I ban we And in plat 3. 
One is naturally struck with the fact, that, in the first 9 inches, plat 
21) the richer land — should show less than half the nitric nitrogen 
