E. J. Russell and A. Appleyard 407 



which the quantity fluctuates till November and December when it 

 falls. There is no accumulation during the autumn as on the dunged 

 fallow : 60 to 70 lbs. seems to be the maximum obtainable and once 

 this is reached httle further action takes place. 



We are much surprised at this high value for so very exhausted a 

 soil, but we have obtained similar ones before. At no time this year 

 did we find less than 30 lbs. of nitric nitrogen per acre in the top 

 18 inches, and 15 lbs. in the top 9 inches; i.e. 7 or 8 parts per milhon 

 of soil. In the preceding paper we show that the similar Hoos Field 

 unmanured plots contained on the fallow part up to 13 parts of nitric 

 nitrogen per million, or 54 lbs. per acre in the top 18 inches, and on the 

 cropped part up to 9 parts per million, or 23 lbs. in the top 9 inches. 

 Now we know from the Broadbalk Plot 10 that the unmanured crop 

 responds to more nitrate, yet it leaves all this quantity untouched. Of 

 course the roots could hardly be expected to exhaust the soil because, 

 as the grass plot experiments show, diffusion is an exceedingly slow 

 process, and the roots only search through a relatively small part of 

 the soil. But one would look for a larger utiUsation than this. 



That exhausted soils can still go on producing nitrate is shown by 

 the amount of nitrate in the drainage water from the drain gauges 

 which after 46 years of bare fallow still yield some 35 lbs. of nitric 

 nitrogen per acre from 20 inches of soil ; this can only be part of the 

 nitric nitrogen present because the amount of percolation is never 

 sufficient to wash it all out. It is possible that the method brings out 

 something besides nitrate, but such comparisons as have been made in 

 our laboratories with colorimetric methods are against this view. And, 

 moreover, light soils give the low values that would be expected. 



In comparing the COg and the nitrate with one another it must be 

 remembered that a period of non-production of CO2 must appear as a 

 falhng line because of the perpetual loss by diffusion. Further, this 

 perpetual loss makes all the upward parts of the CO2 curve less steep 

 than they would otherwise be. The fluctuations in nitrate are about 

 a month later than those in bacterial numbers. 



Broadbalk Dunged Falloiv (Fig. 2). 



This is part of the plot which has received 14 tons of dung annually 

 since 1843 ; it contains in comparison with the unmanured : 



Dunged Unmanured 

 Organic matter (loss on ignition) ... 10-0 % 4-3 % 



Nitrogen 0-20 0-10 



