52 EXPERIMENTS UPON WHEAT 



it had the year before. The crop shows every sign of nitrogen 

 starvation, and amounts on the average to only 15*8 bushels 

 of grain, as compared with 14*9 bushels on Plot 5 which has 

 received minerals without any nitrogen every year since 1852. 

 On the Rothamsted soil, then, we may conclude that the effect 

 of sulphate of ammonia applied to a cereal crop is confined to 

 the season of its application. In the seasons when the 

 ammonium-salts are applied the crop is but little short of that 

 on Plot 7, where minerals are used every year with the same 

 amount of ammonium-salts, thus showing that the previous 

 mineral manuring is carried forward and has an effect in 

 seasons beyond the year of its application. 



Much of our knowledge of the process of nitrification, by 

 which not only ammonium-salts but other compounds of 

 nitrogen, such as are contained in dung, are converted into 

 nitrates, was worked out in the Rothamsted Laboratory by Mr 

 Warington. From the continued analyses that have been made 

 of the water flowing from the drains beneath the Broadbalk 

 wheat plots, we learn that not only may readily nitrifying 

 manures suffer great losses through nitrates forming and being 

 washed out when a crop does not occupy the ground, but that 

 the same causes lead to continuous loss of nitrogen from all 

 cultivated land. This loss is at its highest when heavy rain 

 falls after the land has been broken up after harvest ; then the 

 conditions occur which are most favourable to nitrification, 

 i.e., warmth, moisture, aeration, and stirring of the soil. Thus 

 analyses of the soil show that, despite the fact that much larger 

 amounts of nitrogen are applied to Plots 7 to 18 than are re- 

 moved in the crop, the soil is not getting any richer in nitrogen ; 

 and even on Plots 2 and 19, where organic compounds of 

 nitrogen are used, the accumulation of nitrogen is far less than 

 the difference between the nitrogen applied and that removed 

 Avould indicate. 



Table XVIII. gives an estimate of the nitrogen per acre 

 supplied in the manure and recovered in the crop over a fifty- 

 year period, 1844-1893, together with the nitrogen contained 



