THE GERM LIFE OF THE SOIL 135 



of nitrates in the corn land cannot be attributed to failure 

 of the plants to remove it. Grass appears to have a decidedly 

 depressing effect on the process of nitrate formation, and 

 this may be one reason why grass is generally a detriment 

 to the growth of young orchards. 



172. Depths at which nitrate formation takes place. — 

 It is probable that the processes by which nitrates are formed 

 are, in humid regions, confined largely to the furrow slice 

 of soil. Nitrates found below that point have probably 

 been, in large measure, washed down from above. The sub- 

 soil in such a region is not a verjr favorable medium for these 

 processes. In arid and semi-arid regions, however, the case 

 is different. Here the distinction between surface soil and 

 subsoil is not so marked, and owing to the rich and porous 

 nature of these subsoils nitrification may proceed at con- 

 siderable depths. 



173. Loss of nitrates in drainage. — It has already been 

 shown that there is a large removal of nitrates in drainage 

 water (§ 121). As nitrogen is the most expensive of fer- 

 tilizer constituents every effort should be made to prevent 

 this loss. A very effective way to do so is to have a crop 

 growing on the land during all of the growing season. A 

 comparison of the loss from the planted and unplanted soil, 

 in the paragraph referred to, will show how effective a crop 

 is as a means of preventing loss of nitrates in drainage 

 water. 



Hall states that nitrates formed during the summer or 

 the autumn of one year are practically all removed from the 

 soil of the Rothamsted fields before the crops of the following 

 year have advanced sufficiently to use them. 



174. Denitrification. — After nitrates have been formed 

 by the processes that have just been described, there are 

 other bacteria or some of the same bacteria acting under 

 different conditions that attack the nitrates and convert 



