II] THE SOIL AND PLANT FOOD 13 



as the nitrates, and the mineral group, inchiding 

 the phosphates, etc. of potassium and other metals- 

 There is good ground for the distinction. The nitrates 

 are derived almost exclusively from organic matter, 

 but the mineral food, on the other hand, comes 

 partly fi-om the rock material of the soil. Further, 

 the nitrates are easily soluble in water, and, there- 

 fore, readily washed away ; they are, besides, liable 

 to other sources of loss, while the mineral food only 

 suffers slight losses. Lastly — and this aspect cannot 

 be overlooked in a technical subject like ours — 

 nitrates and other nitrogenous foods are by far the 

 most expensive when any purchasing has to be done. 



In trying to find out how plant food is made in 

 the soil, investigators have confined themselves almost 

 exclusively to the nitrogenous portion. This re- 

 striction was forced on the earlier workers by the 

 circumstance that our soils stand much in need of 

 nitrogenous manure: very much of the Rothamsted 

 work has been and still is devoted to the study of 

 the nitrogen problem, and a large part of our present 

 knowledge is built up on the foundations laid by 

 Lawes, Gilbert and Warington. We also must be 

 content to accept the restriction, and pass over any 

 changes which the mineral food may undergo, for 

 the very good reason that we know so little about 

 them. 



It was early discovered that the plant residues 



