E. J. Russell and A. Appleyard 403 



The drains on the dunged plot ran much less frequently than those 

 on the unmanured plot, the large amount of organic matter having the 

 effect of lightening the soil and facilitating percolation. 



The sharp drop in nitrate content of the soil during September is 

 difficult to explain solely as a leaching effect. During the whole month 

 only 0-82 inch of drainage water passed through the 20-inch gauge, 

 and yet 24 lbs. of nitrogen as nitrate is lost. Later on the percolation 

 becomes very considerable, and the losses proportionately heavy. We 

 have already dealt with this aspect of the question in our earher papers^. 

 All these losses are readily exphcable as percolation effects excepting 

 only that in September, which seems to indicate that when the nitrate 

 becomes very high in amount some other source of loss may set in. 



The changes on the unmanured plot are smaller : there is, however, 

 a drop in June which is difficult to attribute to percolation. 



Carbon dioxide. The formation of carbon dioxide is effected not 

 only by the soil organisms proper, but also by plants, so that it is a 

 more complex phenomenon than nitrification. The simplest case is 

 seen on the dunged fallow plot where the crop is eliminated, but even 

 this is not quite simple as a disturbing factor is introduced in the 

 cultivation necessary to keep down weeds. Beyond the rise to a 

 maximum in July and a fall to October the COg curve in itself shows 

 very httle ; it is, however, interesting in relation to the curve for bac- 

 terial numbers, because it reflects most of the fluctuations there (Fig. 2). 

 The unmanured fallow is throughout on a lower scale but shows less 

 connection with bacterial numbers : the fluctuations here, however, are 

 probably too small to be significant (Fig. 1). 



The effect of the crop is to be seen in May as one would expect from 

 its rapid growth, and, curiously enough, in August at the time of ripen- 

 ing; in both cases the outpouring of CO2 is sudden and extensive, so 

 that the amounts rise to 1-6 and 1-9%. It has already been pointed 

 out that the bacteria fall to their lowest numbers at these points, and 

 for a period there is no accumulation of nitrate ; whether this inertness 

 is due to the CO2 or to some action on the part of the crop is not shown 

 by field observations. The great increase in production of CO, in the 

 soil at the time of ripening is not easily explained as a respiration effect ; 

 the roots at this stage are usually supposed to have ceased functioning. 

 It may result from a decomposition of dead rootlets, as it is chiefly 

 observed when the conditions are favourable for bacterial activity. 

 The problem requires greater investigation. 



1 See note 2, p. 400. 



