22 THE FERTILITY OF THE SOIL [CH. 



of certain wonderful discoveries that had recently 

 been made. Hellriegel and Wilfarth, therefore, very 

 naturally asked if bacteria could be the active agents 

 here, particularly as they knew that the little swell- 

 ings on the roots of the pea the so-called nodules 

 contained bacteria, and also that some bacteria could 

 take in gaseous nitrogen and use it. To test the 

 matter peas were sown in sterilised sand (i.e. sand 

 baked so as to kill all living organisms), containing 

 mineral food, but no nitrogenous food; these made 

 little or no growth and developed no nodules in the 

 roots. Other peas were also sown in similar sand, 

 but they received a water extract of ordinary arable 

 soil ; these made excellent growth and had a marked 

 development of root nodules. If, however, the ex- 

 tract was first boiled it had no effect in increasing 

 growth. 



These experiments afforded satisfactory evidence 

 that the pea could form an association with certain 

 bacteria which should be self-supporting so far as 

 nitrogen was concerned in that it could draw on the 

 immense stores of free nitrogen in the air. The 

 proof was made more rigorous by other and later 

 workers, and the proposition is now one of the most 

 definitely established in modern science. 



Thus peas, vetches, lupins, and, we can add, beans, 

 clover, lucerne, sainfoin, in short all the tribe of the 

 leguminosae, take in stores of nitrogen from the 



