PENNSYLVANIA FIELD EXPERIMENTS 439 



As already stated, a study of the present nitrogen content of the 

 soil of Agdell field will probably furnish more satisfactory informa- 

 tion than can be secured from any other source at this time. 



It is very evident that the loss of nitrogen in drainage water 

 usually exceeds the addition in rainfall; and, unless there are 

 sources of nitrogen other than can be found by the analysis of the 

 legume plants (tops, roots, and tubercles), we must make provision 

 to supply a sufficient excess of nitrogen in farm manure, crop resi- 

 dues, or otherwise, to meet the needs of large crops and to overcome 

 the loss in drainage from rich land. 



In Bulletin 221 of the New Jersey Agricultural Experiment 

 Station, issued July, 1909, Voorhees and Lipman report in detail 

 the results of ten years' investigations with twenty culture experi- 

 ments (in triplicate) in which corn, oats, wheat, and timothy were 

 grown in rotations in 60 cylinders, each 4 feet long and 23^ inches 

 in diameter, set in the earth and open at both ends, so as to approach 

 natural conditions for drainage. Cow manure, fresh and leached, 

 and cow dung (solid excrement), fresh and leached, were used with 

 and without sodium nitrate, ammonium sulfate, and dried blood, 

 in various combinations. 



At the beginning of the experiment, in 1898, the surface soil 

 (8 inches deep) contained 155.47 grams of nitrogen in each cylinder. 

 The amounts of nitrogen applied during the ten years varied from 

 38.25 grams in the leached dung to 58.31 grams in the fresh manure 

 and sodium nitrate combined. The total amounts of nitrogen 

 removed in the sixteen crops harvested during the ten years varied 

 from 21.88 to 36.70 grams; and the total loss of nitrogen, other 

 than that contained in the crops removed, varied from 25.12 to 

 39.38 grams. Thus, in these long-continued and very carefully 

 conducted experiments the absolute chemical control shows loss 

 of nitrogen by leaching far in excess of possible additions by rain- 

 fall, azotobacter, etc. 



After a full consideration of the data accumulated in these 

 experiments with respect to their bearing upon the question of 

 denitrification, the authors make the following statements: 



"We must conclude, therefore, that at least with cow manure, used at the 

 rate of sixteen tons per annum for a period of ten years, no destruction of ni- 

 trates takes place. In view of the long duration of the experiment, and of the 



