134 LIVING ORGANISMS OF THE SOIL [chap. 



pots of sand are watered nearly up to saturation with a 

 culture solution made up as on p. 45, but omitting the 

 sodium nitrate, and the seedling plants are carefully 

 introduced. The pots must now be either kept in a case 

 well protected from dust, or the surface of the sand 

 must be covered over by a thin layer of cotton-wool, in 

 order to keep away dust which may contain the organism. 

 After a few weeks' growth all the plants will begin to 

 show the want of nitrogen by a cessation of growth and 

 the sickly yellow colour of the leaves. To one pot 

 is then added a few cubic centimetres of a soil extract 

 made by shaking up soil with rather more than an equal 

 bulk of water, and filtering ; to a check pot is added 

 the same quantity of infusion that has been boiled. 

 Similarly, another pot is watered with an infusion of a 

 nodule taken from a leguminous plant of the kind 

 growing in the pot, and as before the check pot is given 

 the same infusion, boiled. The checks will continue to 

 get more sickly looking and will shortly die, whereupon 

 it can be ascertained that they possess no nodules on 

 their roots ; but the plants inoculated either from soil or 

 another nodule develop nodules, recover their health, 

 and will grow into perfect plants if the supply of mineral 

 constituents be sufficiently renewed. The experiment 

 should be duplicated, because the check pots are liable 

 to get accidently inoculated unless dust is most rigor- 

 ously excluded. 



This discovery of the fixation of nitrogen by the 

 bacteria associated with leguminous plants at once threw 

 light on a number of facts which had been difficult of 

 explanation before. For example, farmers had always 

 known that in some way the growth of clover and similar 

 plants was beneficial to the soil ; even in the time of the 

 Romans, Pliny, Virgil, and other writers on agriculture 

 had instructed the husbandman to prepare his land for 



