E. J. Russell and A. Appleyard 



413 



Composition of the Drainage Water in the Rothamsted Wheat Field 

 before and after the Application of Ammonium Salts. 



Plot 15. October— November, 1880 Plot 7. October— November, 1891 



Conclusions. 



The changes in bacterial numbers and in nitrate content of the soil 

 and in COg content of the soil air have been determined at frequent 

 and regular intervals during several seasons on five different plots of 

 land, and the results have been set out on curves. 



1. There is sufficient resemblance between the curves for bacterial 

 numbers, CO2 (except for a period on cropped land), and nitrate to 

 justify the conclusion that they are all related. 



2. The curve for nitrate, however, is always behind that for bacterial 

 numbers, the lag amounting to two or three weeks. Assuming, as we 

 think we may, that the curves are connected, this would indicate two 

 stages in nitrate production : one related to the bacterial numbers, the 

 other not. We bring evidence against the view that the stages are 

 simply ammonia production and then nitrate production : the division 

 has apparently to be carried further back, and ammonia production 

 divided into two stages. 



3. The biochemical decompositions in the soil are determined in 

 the first instance by the temperature, and do not proceed to any notable 

 extent below 5° C. 



4. As soon as the temperature rises action begins rapidly. But it 

 soon slows down and other factors begin to operate. 



