99 



land now represents the result of something more than 20 years 

 of prairie conditions in England, and as samples of soil had been 

 taken at starting it affords an opportunity of gauging the rate at 

 which fertility is accumulating. A very similar experiment was 

 also made with a portion of the Geescroft field, which had carried 

 beans from 1847 to 1878 and clover from 1883 to 1885 ; after the 

 second cutting of clover in 1 885 the field was fenced off and has 

 been left untouched ever since 



Both fields show a marked gain of carbon and nitrogen down 

 to the third depth of 27 inches, the increase in the lower depths 

 being due to the roots which have decayed in that stratum 



If the total amounts of nitrogen in the Broadbalk soils be cal- 

 culated on the assumption that the weights of the soil layers were 

 the same in 1904 as in 1882, the total gain of nitrogen per acre 

 would amount to 2,200 lbs., which is at the rate of more than 100 

 lbs. per acre per annum. So great an accumulation of nitrogen is 

 manifestly impossible to account for in the present state of our 

 knowledge . . . The Geescroft field shows a similar though 

 smaller increase in the proportion of carbon and nitrogen down 

 to the depth of 27 inches ; considering the surface soil only the 

 difference in the amount of nitrogen accumulated by the two fields 

 amounts to about 350 lbs. per acre. The fact that the increase on 

 Geescroft is smaller than on Broadbalk is of considerable interest, 

 because after the Geescroft sample had been taken in 1883 clover 

 was grown for three years before the land was allowed to run 

 wild. Moreover the soil was sampled again in 1885, after three 

 years' growth of clover, and showed in the surface soil an increase 

 of nitrogen . . . The Geescroft field had in fact some start of 

 the Broadbalk field, why did it not maintain its lead ? The answer 

 is probably to be found in the botanical composition of the wild 

 herbage which has taken possession of these two bits of waste. 



It should be remembered that previously the Broadbalk land 

 had been growing wheat, the Geescroft field beans until they 

 would grow no longer, then a good crop of clover. At the pre- 

 sent time the vegetation on the Broadbalk waste contains a fair 

 proportion of leguminous plants, chiefly meadow vetchling, while 

 this class of plants is and has been for many years, since the 

 dying out of the clover, absent from the Geescroft waste. It is 

 impossible to refrain from correlating the absence of leguminous 

 herbage on these old bean and clover plots with the well-known 

 fact that land becomes " sick" of the leguminous crops in a way 



that never happens with the other farm crops The absence 



of leguminous herbage collecting nitrogen from the atmosphere 

 would explain why the Geescroft field has gained nitrogen less 

 rapidly than has the Broadbalk field with its more mixed herbage. 



Another question however arises : how comes it that the Gees- 

 croft land, with no plants growing on it which are capable of 

 fixing free nitrogen, has yet gained an enormuus quantity of ni- 

 trogen during the twenty years under review, a quantity which at 

 the lowest reckoning amounts to about 25 lbs. per acre per year? 

 The nitrogen brought down in the rain would account for perhaps 



