E. J. Russell and A. Appleyakd 389 



By comparing the curves for fallow and for cropped land it is pos- 

 sible to ascertain the effect of the growing crop. So far as the accumu- 

 lation of nitrate is concerned the effect is markedly depressing. Part 

 of this, however, is due to the nitrate taken by the crop, but after 

 allowance is made for this there still appears to be a depression. The 

 curves do not show how this is caused, whether it is due to the com- 

 petition of the plant roots for moisture, or for dissolved oxygen, or 

 whether some definite toxin is concerned. The effect is counteracted 

 for a time by a shower of rain, but it soon sets in again. 



We therefore regard the gro^\ang crop as a fourth factor and on the 

 whole a detrimental factor in the biochemical relationships. 



Probably also the amount of COg present in the soil air is another 

 factor, though here the evidence is less clear because the effect of 

 CO2 only appears when large amounts are present, and this only happens 

 on cropped ground. 



But these also do not account for the two periods already men- 

 tioned, nor do any other of the factors we have studied. 



So far we have dealt with biochemical activity as a whole. We 

 must now discuss the problem a little more closely so as to deal with 

 the separate quantities concerned ; the bacterial numbers, the CO2 

 content of soil air and the nitrate content of the soil. 



It is generally recognised that the biochemical decompositions in 

 the soil are mainly brought about by bacteria, and this view is sup- 

 ported by the fact that a distinct connection can be traced between 

 the bacterial numbers and the production of CO2. We observed this in 

 our previous paper ; this year's results emphasise it. On the Broadbalk 

 dunged fallow, for instance (Fig. 2), where there is no comphcation arising 

 from the presence of a crop, the curves both for bacterial numbers and 

 for CO2 rise and then fall in May, rise in July, fall in August and Sep- 

 tember, rise and fall in October, and then rise in November. There is, 

 however, a discrepancy in June when the numbers fall while the CO2 

 rises. Great Harpenden Field (Fig. 5) shows a similar kind of agreement 

 in late March and April, and also from October onwards. The agreement 

 continues so long as the CO2 is only small in amount, but it breaks 

 down when the CO2 becomes high as on the cropped ground between 

 May and August, i.e. during the active growdng period ; the numbers are 

 then depressed and show no rise corresponding with the COg peak. 



As they stand the curves for bacterial numbers are not clearly 

 related to those for nitrate. Closer examination shows a similarity but 

 some displacement, the rise in nitrate coming after the rise in bacterial 



Joum. of Agric. Sci. vni 26 



