650 
NATURE 
[AUGUST 22, 1912 
the land running wild the vegetation is never re- 
moved, but allowed to die down naturally. Hence 
not only is the nitrogen taken out by the crop 
BrOaDBALK Fietp, RorHamstTep. 
Land allowed to run wild. Nitrogen in Soil, lb. per 
acre. 
In soil to 27 in. | 
| Added Gain in 
= | by | soil per 
1881 7904 | rain | annum 
Broadbalk 5910 8110 | go 92 
returned to the soil, but also a large stock of car- 
bonaceous mattér assimilated from the atmosphere, 
and this carbonaceous matter furnishes a bacterium 
present in the soil, Azotobacter chroococcum, with 
the source of energy which will enable it to fix 
atmospheric nitrogen. Azotobacter is equally present 
in the soil of the unmanured wheat plot; but, as there 
the crop is removed and only a little root and stubble 
left behind, there is but little carbonaceous matter 
for the Azotobacter to work upon, and a correspond- 
ingly small fixation of nitrogen, sufficient only, as we 
have seen, to repair the casual losses by drainage 
and weeding. This plot gives us a clue to the source 
of the vast accumulations of nitrogen in the old prairie 
soils. Vegetation alone, however long continued, 
cannot increase the stock of nitrogen in the soil; 
there is only a circulation of the initial stock removed 
by the plants and then put back when the plant dies 
in situ. But if the conditions are also favourable to 
the development of Azotobacter, this organism derives 
from the carbonaceous part of the plant residues the 
energy it requires for the fixation of nitrogen, and a 
steady addition to the original stock goes on. We 
have found Azotobacter present in all these rich black 
soils, from both South and North America, the Russian 
Steppes, and similar virgin land in all parts of the 
world, and again we also find an abundance of lime, 
one of the other necessary factors for the growth of 
Azotobacter. Virgin soils are not necessarily rich; 
there are miserably poor ones, though they have 
equally carried some sort of vegetation for hundreds, 
indeed thousands, of years. They have remained poor 
because some of the other factors upon which depend 
the development of Azotobacter are lacking. With 
this far-reaching conclusion in sight, we have naturally 
tried at Rothamsted whether we could not bring 
about a similar heaping up of nitrogen in the soil 
by simply adding to it a carbohydrate containing no 
nitrogen, such as starch or sugar. In pots, the ex- 
periment is perfectly successful, and accordingly we 
HoosFigtp Bartey. 
Effects of Sugar (or Starch) on the Amount of 
Produce. Plot 4 O. Complete Minerals. 
Total produce of barley 
= Sugar (or starch) 
Year applied ~ 
Without sugar With sugar 
Ib. lb. 
1906 Spring 2485 2 
1907 * 6 3578 3249 
1908 y< 1820 1404 
1999 ” 2563 | 2261 
1910 Autumn 2082 2502 
tori » 1244 1915 
2 Very small crop, not weighed. 
® Starch applied instead of sugar in 1907. 
NO. 2234, VOL. 89] 
selected one of the plots in the barley field which was 
in a very nitrogen-starved condition, because it had 
been manured for fifty years only with mineral fer- 
tilisers containing no nitrogen, and treated half the 
plot with sugar at the rate of a ton to the acre, the 
other treatment of the two halves of the plot being 
alike. To our surprise, the half receiving sugar gave 
a miserable crop, much below the non-sugar half, 
for four years in succession, and a_ bacteriological 
examination of the soil showed that Azotobacter had - 
not increased in response to the sugar, but that the 
number of merely putrefactive organisms had gone up 
greatly. These facts led Dr..Hutchinson to surmise 
that we had been putting on the sugar at the wrong 
time of year, in early spring or winter, some time 
before the barley was sown, when the soil is cold. 
Now Azotobacter is comparatively inactive at low 
temperatures, and the sugar was probably being 
wholly taken by the Streptothrix, &c., which are less 
affected by cold. As these organisms must also obtain 
nitrogen, they were robbing the barley of the small 
stock available in the soil, and so bringing about the 
observed reduction of crop. A change was accord- 
ingly made in the time of application of the sugar, 
which was put on as soon as the barley had been 
harvested, when the soil still retained its summer 
heat, and the change was immediately followed by 
an increase in the succeeding barley crops, as com- 
pared with the non-sugar plots, that was as marked 
as the deficiency had been previously. This illustrates 
the many pitfalls which attend investigations in agri- 
cultural science. Under laboratory conditions one can 
define the issue sharply, but as soon as the experi- 
ments are extended to the open ground and living 
plant, so many extraneous and unsuspected factors 
come into play that what is popularly called a conflict 
| between theory and practice often becomes apparent. 
We may now take a more complex example from 
the Rothamsted plots to illustrate what I have called 
the conservative systems of farming. One of the 
fields is farmed on a four-course rotation of turnips, 
NITROGEN PER CENT. IN SOIL OF AGDELL FIELD, 
ROTHAMSTED. 
The Plots receive Mineral Manures, but no Nitrogen. 
Fallow Clover 
Roots Roots Roots Roots 
carted off, returned, carted off, returned, 
13/14 | 0/10 15/16 1/12 
——— ae a 
1867 | 01224 | o'1240 0°1327 0'1380 
1874 O'1147 0°1238 O'1241 O'1321 
1883 O'r161 01228 071329 0°1383 
1g09 | O1I59 O'1195 0°1347 0'1498 
1852-1903 
Wheat, average | 31°2 bush.| 32°2 32°2 35'1 
Clover 5 — | 4I'o cwt. 47°7 
Swedes ,, I5I Ocwt. 268°0 160°0 1870 
Barley | 22°1 bush. 28°7 24°5 34°55 
barley, clover, wheat, but over half the field the 
clover is replaced by a year’s bare fallow. Further, 
if we confine our attention to the one plot which 
never gets any nitrogen, but only mineral fertilisers, 
it is again divided at right angles into plots from 
which the turnip crop is wholly removed, and others 
on which it is returned, as so often occurs in practice 
when the turnips are eaten off in situ by sheep. 
The above table shows the average yield on these 
