22 



other remains of previous crops. This action will probably 

 also explain at least part, if not the whole, of the loss of 

 nitrates where ammonia salts, spring sown, were applied in 

 addition to minerals (Column III.), for the larger crop thus 

 produced would, of course, leave in the soil a larger quantity 

 of nitrogenous matter capable of undergoing nitrification. 



TABLE IX. Nitrogen as Nitrates per Million of Water. 



The plot receiving minerals and 400 Ibs.of ammonia salts 

 sown in autumn (Column IV.), showed a very different rate 

 of loss by drainage. On Oct. 23, the proportion of nitrates 

 in the drainage water was less than where the ammonia 

 salts were spring sown, probably because smaller crops had 

 been produced, and there was therefore less material in the 

 soil to undergo nitrification. On Oct. 27, the ammonia 

 salts for the following year's crop were applied ; and the 

 next time the drains ran the proportion of nitrates had enor- 

 mously increased, being then higher than in any other sample 

 of drainage water examined. All through the winter, until 

 the end of the experiment, the quantity of nitrates, though 

 steadily falling, remained a good deal above that washed 

 from any of the other plots. Judging by the results obtained 

 on March 5, 6, and 7, we may conclude that the same 

 exceptionally heavy loss continued until the spring. When 

 it is remembered that about 70 per cent, of the total drainage 

 of the year occurs during the months of October to March 

 inclusive, it will be seen what a large quantity of nitrate may 

 thus be lost in the course of the winter. 



Column V. of Table IX. is added for comparison with 

 Column III., to illustrate the fact that has been mentioned 



