20 THE FERTILITY OF THE SOIL [CH. 



these actions, the formation of nitrates as well as the 

 loss of nitrogen, tend to use up the nitrogen com- 

 pounds of the soil, because the nitrates not taken 

 by the plant are speedily washed away and lost in 

 the rivers and the sea. As the original stock was 

 probably never high, it is clear that there must be 

 some reverse process by which the soil gains nitrogen, 

 or the supply would long since have given out. This 

 has long been realised by men of science, and a 

 careful and systematic search begun 30 years ago was 

 ultimately rewarded by the discovery of two ways in 

 which such a gain takes place. 



The old-established cultivator of the land has a 

 great stock of information about the ways of plants ; 

 some of it is disconnected and fragmentary, but it 

 has to be sorted over and examined experimentally 

 by the man of science. One of the old bits of know- 

 ledge handed down from time immemorial, and 

 already traditional when Virgil wrote his Georgics, 

 was that beans, vetches, and lupins improve the land 

 for the next crop. Sow your golden corn, says Virgil, 

 on land where grew the bean, the slender vetch 

 or the fragile stalks of the bitter lupin 1 . When 

 Lawes and Gilbert began their exj)eriments in 1843 

 one of the early problems was to discover the reason 

 for this improvement, and they were able to trace it 

 to the fact that a soil was richer in nitrogen after the 



1 Georgia, Book i, lines 73 et eq. 



