99 
eighty years of cultivation, does not contain aa much nitrogen to ;i depth of 3 feel as 
would be contained in twenty good crops of corn. 
Another large area, evidently comprising several hundred thousand acres, does 
not contain sufficient potassium in the plowed soil to make twenty good crops of 
corn, and the subsoil is still more deficient in potassium than the top soil. 
By chemical analyses we have found that one of these extensive soil types con- 
tains more than six times as much phosphorus as another; one contains five times as 
much potassium in the top soil as another, with a still greater difference in the sub- 
soils; one type contains from ten to sixty times as much nitrogen as another. The 
principal typesof soil in central and northern Illinois contain from two to three 
times as much plant food and produce two to three times as much corn as the prin- 
cipal types in southern Illinois. These are not mere theories; they are absolute 
facts. Itased upon chemical analysis of the soil, upon pot cultures carried on under 
controlled conditions, upon actual field experiments, and upon regular crop yields in 
ordinary farm practice. 
Plant food in same Illinois surface soils [pounds per acre). 
Elements of plant food. 
Black prai- 
l rie (Wis- 
coEMEJa gia- 
eiation). 
Red clay 
hills (uri- 
glaciated ). 
Gray prai- 
rie | lower 
Illinoisan 
glacia- 
tiou). 
Peaty swamps i re- 
een'tly drained). 
Nitrogen 6.200 ! 1.000 - - 
Phosphorus 1,600 1.000 600 
Potassium 8, 800 , 5. 600 4, 200 
67,000 
2.000 
1,200 
Crop yields in soil experiments. 
Plant food applied. 
Corn. 
Wheat. 
Wheat. Corn. 
Corn 
fodder. 
Bushels 
None « ' 
Nitrogen 
Phosphorus 
1 \ itassium 
Nitrogen, phosphorus 
Nitrogen, potassium 
Phosphorus, potassium 
Nitrogen, phosphorus, potassium 
Grams. Bushels. Pounds. 
o 
1.000 

1,200 
o 1 
2,000 
36 
3,600 

1.400 
40 
3,500 
38 
3,100 
60 
4,400 
a A very common yield. No experiment, on this type. 
To the old, worn, nnglaciated hill soil of southern Illinois, whose chemical compo- 
sition shows it to be markedly deficient in nitrogen, we added both phosphorus and 
potassium and obtained practically the same yield as where no plant food was applied, 
but when nitrogen was added the yield of wheat was increased from only 3 grams to 
from 26 to 34 grams per pot 
To the principal type of soil in the lower Illinoisan glaciation, whose analysis shows 
that its phosphorus content is only one-third of a normal fertile soil, we added both 
nitrogen and potassium and produced no increase whatever, but when phosphorus 
was added the yield of wheat was increased from 10 grams to from 14 to 27 grams 
per pot. 
To the peaty swamp soils representing some hundred thousand acres in north- 
central Illinois whose composition shows that it contains less than one-fifth as much 
potassium as the best soils in the corn belt, we added both nitrogen and phosphorus 
and obtained practically the same yield of corn as where no plant food was added, 
the total yield per acre amounting to only a 'ton or less of corn fodder with practically 
no ear corn, and yet where we applied potassium to that soil we obtained about two 
tons of corn stover and from 36 to 60 bushels per acre of good corn, and following 
up these results, the farmers who own and manage those lands are already profitably 
using carloads of potassium salts upon those soils, not highly manufactured so-called 
complete commercial fertilizers, but crude potassium salts direct from the German 
mines and in quantities sufficient for a good crop of corn. It will be observed that 
after the most needed element has been applied the other elements added may pro- 
duce more or less increase in the crop. 
Are all these results produced by the physical effect of these materials? Does 
nitrogenous material produce this physical effect in one soil, phosphatic in another, 
