712 TRANSACTIONS OF SUB-SECTION K. 
Broadbath, Plot 2. Nitrogen, lbs. per acre. 
In Soil, | InSoil, | Gain in | Added in | aoe by | Removed | Unaccounted 
1865 1893 28 Years | Manure | Seed. in Crop for 
4,343 Ibs. | 4,976 Ibs. | 633 Ibs. | 5,600 Ibs. 167 Ibs. | 1,361 Ibs. | —3,773 Ibs. 
Under these conditions the losses are enormous, amounting to about 
one-half of the nitrogen added. 
C. When land is carrying a natural vegetation which is not removed 
there is a gain of nitrogen. A portion of the Broadbalk wheat-field has 
been allowed to run wild since 1881; a rough natural vegetation, consisting 
mainly of grasses, but also containing about 25 per cent. of leguminous 
herbage, has established itself and is never removed, but allowed to fall and 
decay. Analyses of the soil in 1904 showed a considerable accumulation of 
nitrogen, 
Nitrogen, lbs. per acre, top 9 inches. 
Tn Soil, 
In Soil, | Added by Rain | | 
1881 1904" | (28 Years) | Gam 
2,769 Ibs. | 3,711 lbs. 88 lbs. | 854 Ibs. 
! 
There is also a considerable gain in the second and third nine inches. 
A similar piece of land on which the vegetation is exclusively grassy 
and contains no leguminous plants shows a smaller but still marked gain 
of nitrogen. This gain may be attributed to bacterial agencies (No. 2). Not 
only are there leguminous plants in the plot, but azotobacter is also present, 
and is enabled to fix large quantities of nitrogen because it obtains the 
necessary carbohydrate from the annual fall of vegetation. Fixation is 
small on the unmanured arable plot (A) because only a small root residue 
of carbonaceous matter is left in the land every year. It is to the activity 
of azotobacter when thus supplied with carbohydrate by the annual fall of 
vegetation that we may attribute the accumulation of nitrogen in virgin 
soils. The higher plants alone, however long they occupied the land, could 
only restore what they had previously taken from the soil, and thus could 
never originate the vast stores of nitrogen that are found in such virgin 
soils as the black steppe soils of Manitoba and the North-West. This con- 
clusion is strengthened by the fact that such steppe soils are always well 
supplied with calcium carbonate, a necessary factor in the action of 
azotobacter. The organism itself has also been isolated from all such soils. 
From the practical point of view we may conclude that the cultivation 
of successive cereal crops will rapidly reduce the stock of nitrogen originally 
in the soil, not only by the amounts withdrawn in the crop, but also by the 
oxidising actions which are set up in the cultivated land. By a more 
conservative system of farming, in which leguminous crops are introduced 
into the rotation and a certain amount of carbonaceous matter is returned 
to the soil, the recuperative agencies become sufficient to repair the losses 
of nitrogen and maintain a moderate level of fertility without the intro- 
duction of any extraneous source of nitrogen. Such, indeed, was the state 
of things in Europe previous to the discovery of artificial fertilisers, e.g., the 
average production of wheat in England in the early years of the nineteenth 
century was about 20 bushels per acre on farms which were self-supporting 
as regards the elements of fertility, and this level has been maintained for 
a long time. 
This conclusion may be illustrated from the results given by the Agdell 
Tield at Rothamsted, which is farmed on a four-course rotation, with clover 
