AUGUST 22, 1912| 
NATURE 
649 
consumption of the turnips and hay. Thus but a 
small proportion of the nitrogen taken out of the soil 
by the crop left the farm; the rest was returned and 
used over again, although considerable losses of 
gaseous nitrogen occurred during the making of the | 
dung. Both losses, however, were more than replaced | 
by the nitrogen which the clover crop gathered from 
the atmosphere during its growth. At any rate, we 
find that under such a conservative system of farming 
the productivity of the land remained pretty constant 
at about a level of twenty bushels to the acre from 
the time of Queen Elizabeth down to the beginning 
of the nineteenth century. This conservative farming 
about 1840 began to give place to the third stage in 
the development—intensive farming, rendered possible 
by the discovery of artificial fertilisers and the cheap 
freights which brought foreign fertility in the shape 
of cheap feeding stuffs to the soil of this country. 
By these means the average production of the land of | 
the British Isles has been .raised from the twenty- 
bushce! level to something over thirty bushels, and the 
most intensive farmers reach an average level at least 
25 per cent. higher. In their case the soil has be- 
come practically a manufacturing medium transform- 
ing the nitrogen and other fertilising materials added 
to it into crops, giving nothing to those crops from its 
original stock, and indeed up to a certain point gain- 
ing rather than losing fertility with each year’s culti- 
vation. The inner history of these three stages in 
agriculture may be followed by a consideration of 
certain experimental plots at Rothamsted. We may 
begin with the experimental wheatfield which is now 
EXPERIMENTS ON WHEAT, BROADBALK FIELD, 
ROTHAMSTED. 
Average Produce of Grain, first 8 years (1844-51) and 
the successive 10-year periods 1852-1911. 
| Averages over 
' j 
Plot Manure 8 10 | 10 | zo | ro | xo | 10 60 
years,| years,| years.| years, years,! years, years,| years 
1844- | 1852- | 1862- | 1872-  1882—| 1892- | tg02- | 1852 
185x | 1861 | 1871 | 188 |} 1897 | tgor | rgtr | 1911 
= Bush.| Bush | Bush.| Bush. Bush. | Bush. Bush.| Bush. 
Farmyard 4 5 5 5 : 
{ acnyant 28'0 wala 28°7 ) 382 | 392 | 35°r | 35°5 
3 Unmanured 17-2 15°99 | 14°5 | 10°4 | 126 | 12°35 10°9 12°8 
| | 
carrying its sixty-ninth successive crop of wheat. One 
of the plots has been without manure throughout the 
whole of that period. The production, which fell 
steadily for the first ten years, has since that time 
remained so constant that the slow falling off which 
we still believe to be taking place is disguised by the 
fluctuations due to season. The average yield is about 
twelve bushels to the acre, almost exactly the average 
yield of the wheat lands of the whole world. Unfor- 
tunately samples of soil were not taken at the very 
outset, but if we begin with the earliest analyses that 
were available in 1865 and draw up a balance-sheet 
for the nitrogen, we shall find that the removal in the 
crop is almost exactly balanced by the small amount 
that comes down in the rain and the decrease that 
has taken place in the amount of nitrogen in the 
soil. 
not brought into account; some is washed away by 
drainage water every year, and a further small un- | 
As these 
must 
work 
estimated amount gets removed as weeds. 
losses do not appear in the balance-sheet we 
conclude that some recuperative action is at 
keeping up the stock, though the process is not suffi- 
cient wholly to make up for the removals in the crop. 
The results of this plot show two principles at work 
NO. 2234, VoL. 89] 
There are, however, other losses of nitrogen | 
the tendency of the Jand under an unchanging system 
of farming to reach a position of equilibrium when 
the only variations in the crop are those brought about 
by seasons; and, secondly, that regeneration of the 
nitrogen stock in the soil is possible by natural causes 
alone. 
We may now turn to one of the other plots which 
receives an excess of farmyard manure each year, the 
manure supplying about 200 Ib. of nitrogen per acre, 
whereas the crop only takes away about 50_ Ib. 
Naturally the land in this case increased in fertility, 
but after twenty or thirty years another position of 
BroappaLtK WuHeEat FIELD. 
Nitrogen in Soil, lb. per acre. 
Added | Removed | 
In soil, In soil, Ra ee eos on Unaccounted 
1865 1904 : ai for 
39 years | manure rain crop 
Plot 3.—Unmanured. 
: — | ——— 
2850 2290 | — 560 ao || ay || eile) | —110 
Plot 2.—Farmyard Manure. 
_ ; | 
4470 4970 | +500 7800 | 150 | 1990 | —5460 
equilibrium was attained at a level of about 36 bushels 
per acre, after which, despite the continued additions 
of manure, the crop again did not vary except as the 
result of exceptionally favourable seasons. If we now 
consider a similar balance-sheet for this plot, we find 
that the additions of nitrogen are balanced neither 
by the removals in the crop nor by the accumulation 
of nitrogen in the soil; indeed half of the nitrogen 
applied is unaccounted for. The soil has been getting 
no richer for the last twenty or thirty years, and the 
greater part of the nitrogen is wasted, doubtless be- 
cause bacterial action sets the nitrogen free as gas. 
Here, then, we see another principle illustrated, that 
in very rich land the wasteful agencies are so speeded 
up as to prevent any continued accumulation of fer- 
tility out of the unused residues of the manures put 
on. Higher fertility means a higher level of waste, 
and this explains the rapidity with which the very 
rich virgin soils lose their fertility when they are put 
under arable cultivation. In this Rothamsted plot, the 
soil of which still contains less nitrogen than the less 
rich virgin soils of the prairies, three times as much 
nitrogen are wasted every year as is converted into 
crop, and the same or an even greater rate of wastage 
must attend the conversion of the rich virgin soils 
into land growing a succession of cereal crops. 
We may now turn to another plot on the same field 
to illustrate the recuperative actions of which I have 
spoken. This is a part of the field that has been 
running wild since 1881, when the wheat it carried 
was not harvested but allowed to seed itself. A very 
few years sufficed to eliminate the wheat, which was 
unable to maintain itself against the competition of _ 
the weeds, and the land now carries a miscellaneous 
vegetation consisting mostly of grass. A soil sample 
was taken at starting, and when compared with 
another sample taken twenty-three years later showed 
that in the interval the land had gained nitrogen at 
the enormous rate of 92 Ib. per acre per annum. 
Making every allowance for possible errors in sampling 
and analysis, the accumulation of nitrogen is in 
marked contrast to its steady depletion in the equally 
unmanured arable land alongside. Now, the differ- 
ence between the two plots lies in the fact that on 
