E. J. Russell and A. Appleyard 409 



Broadbalk Unmanured Cropped (Fig. 3). 



The bacterial numbers fluctuate as on the fallow land: starting 

 from the winter level of about 7 millions per gram they rise in April 

 to 16, then fall to less than 5 in June; they rise and remain about 

 10 millions with some fluctuations, and rise in October, finally falHng 

 to the winter level in November. 



The CO2 curve differs considerably. The winter level is 0-2 % : a 

 rise sets in in March to 0-6 % at the end of April, then comes a drop in 

 May, and a rise again to 0-6 % beyond which there is no further rise 

 till July, when the amount rapidly rises to over 1-8 % followed by 

 a sudden drop to 0-2 % in September and October, and a small rise in 

 November. 



The nitrate curve rises and falls with the CO2 curve, the maximum 

 amount, 55 lbs. per acre of nitric nitrogen, being attained in October, 

 The curve also resembles that for bacterial numbers, but it is about 

 three weeks to a month behind. 



The difference in nitrate content between cropped and fallow land 

 is much less on the unmanured than on the dunged plots, and is never 

 more than 26 lbs. per acre in the top 18 inches. The crop contained 

 21 lbs., so that the cropped land seems to have contained as much 

 nitrate as the fallowed land (Fig. 7). 



Broadbalk Dunged Cropped (Fig. 4). 



The bacterial numbers fluctuate more than on the fallow plot but 

 in the same general way. The winter level is, as on the fallow plot, 

 about 22 millions, but in April the numbers run rapidly up to 35 miUions, 

 followed by a drop to 10 milhons in June: then comes a rise to 

 26 milhons in July : a drop in August and early September, but a sharp 

 rise to 31 milhons in October, finally a fall to the winter level. It is 

 remarkable that the March numbers rise so much higher than those of 

 the fallow plots. 



From April onwards there is a close agreement between bacterial 

 numbers and rainfall curves excepting only in early May and early 

 August, the times when the COg is at a maximum. These curves 

 represent the amount of rain falling during the seven days preceding the 

 observations. 



The CO2 curve begins at a low winter level of 0-4 %, rises steadily 

 then finally sharply to 0-8 % in April : then comes a fall in May, 

 followed by a rapid and considerable rise to 1-6 %: then comes a fall 



