210 CROPS GROWN IN ROTATION 



the loss of nitrogen to the soil during the rotation is greater, 

 amounting to over 24 lb. per acre per annum, the increase 

 being almost wholly due to the much larger Swede crop 

 which is obtained by the help of the mineral manures. 

 Coming to the plot which receives a nitrogenous manure for 

 the Swede crop, we find the average removal of nitrogen 

 becomes 40 lb. per acre per annum, a slightly greater quan- 

 tity than is supplied by the manure, so that the net loss 

 is about 5 lb. of nitrogen per acre per annum, approximately 

 the amount annually restored by the rain. Thus, if we con- 

 sider this plot alone, an almost exact balance is obtained 

 l)etween the nitrogen supjDlied and the nitrogen removed, so 

 that the fertility of the land should be closely maintained. 

 There are, hoAvever, other sources of loss which the above 

 figures do not take into account — losses by the removal of 

 weeds, losses by the washing away of nitrates, especially 

 during the bare fallow, and losses due to the decomposition 

 of nitrogenous materials in the soil with the evolution of their 

 nitrogen as gas, " denitrification " so called. Possible sources 

 of gain are the absorption from the atmosphere of ammonia 

 otlier than the ammonia washed doAvn in the rain, and the 

 fixation of atmospheric nitrogen by soil bacteria which do not 

 require the co-operation of a leguminous plant. It is difiicult 

 to decide whether the fertility of this plot is i-eally falling off 

 or not, so great is the effect of seasons in causing fluctuations 

 in yield which cannot be gauged ; the last four root crops 

 have actually been greater than the first four, the wheat has 

 been somewhat less, Avhile the barley has given in the latter 

 years less than half the crop of the earlier ones. The seasons 

 in the latter years were, however, against the l^arley crop, so 

 that one can come to no very definite conclusion as to whether 

 the recuperative agencies indicated above are sufficient to com 

 pen sate for the unestimated but inevital^le losses. 



Turning now to the plots on which clover or beans are 

 grown, it becomes still more difficult to estimate the gain or 

 loss of nitrogen to the land, since the leguminous crop gains an 



