io6 SCIENCE AND SOIL 



for 9900 years, or as long as from 8000 years before Christ to 1903 

 years after Christ. In other words, on the absolute mathematical 

 basis there is less of reason for applying potassium to normal 

 soils as plant food than there is for applying magnesium and cal- 

 cium for the sake of adding them also as plant food (see page 561). 



The use of potassium on soils actually deficient in that element 

 (as peaty swamp soils) , and on some other soils as a soil stimulant, 

 is discussed in the following pages. 



In brief, it may be said, of the plant-food elements supplied by the 

 soil, that nitrogen and phosphorus are in one class; that potassium, 

 magnesium, and calcium are far removed in a second class; and that 

 iron is distinctly in a third class; while the nitrogen of the air, so 

 far as concerns the supply for permanent agriculture, should be 

 classed with carbon, hydrogen, and oxygen. The place of sulfur 

 is not so easily determined. Measured by the soil's supply, sulfur 

 would be classed with phosphorus and nitrogen; or measured by the 

 crop demands, it would be classed with iron; but, if both supply 

 and demand are considered, it must be classed with potassium, 

 magnesium, and calcium. It is, however, known with certainty 

 that more or less sulfur is carried into the air with the products of 

 combustion and of decay,' and some sulfates are also carried into 

 the atmosphere in the dust from evaporated ocean spray. Long- 

 continued investigations at Rothamsted and elsewhere have shown 

 that as an average rainfall brings to the soil, chiefly in the form of 

 sulfates, about 7 pounds of sulfur per acre per annum, or i pound 

 more than would be required for a zoo-bushel crop of corn (p. 57). 



NOTE. In New Hampshire Agricultural Experiment Station Bulletin 142 

 (December, 1909), Morse and Curry report 37,400 pounds of total potassium 

 in two million pounds of surface soil of the uplands, and 50,000 pounds in two 

 million of surface soil of the lowlands, in the vicinity of Durham, these amounts 

 representing averages of ten and fifteen respective soil analyses of clays and 

 clay loams. The summary of this bulletin contains the conclusions that the 

 potassium in these soils is soluble enough to supply potassium for heavy crops 

 of grass without artificial reenforcement, and that additional potassium when 

 supplied in commercial fertilizers does not affect the yield or the composition 

 of the grasses. 



