II] THE SOIL AND PLANT FOOD 15 



because the organisms would have to gi'ow. To test 

 this hypothesis they added a little chloroform vapour 

 and found that nitrification was stopped entirely : it 

 could, however, be started again by adding a little 

 turbid extract of fresh soil after the chloroform was 

 removed. They concluded, therefore, that nitrifica- 

 tion was the work of "organised ferments." More 

 rigid proof was afforded by Waring-ton and later on 

 by Winogradsky. 



More recent experiments render it highly impro- 

 bable that any chemical or physical process going on 

 in the soil gives rise to nitrates, and we may take it 

 that their production is entirely bacterial. Warington 

 showed that the process takes place in two stages ; 

 the ammonia is first converted into nitrites by one 

 organism, and the nitrite is then changed to nitrates 

 by another organism. Nobody has yet succeeded in 

 finding any third stage between ammonia and nitrites 

 although one might be expected on chemical grounds. 

 There is practically no waste of ammonia during the 

 process, and the conversion is almost if not entirely 

 complete, but its mechanism is not at all understood 

 and it cannot be reproduced artificially. 



The organisms which alone can bring it about are 

 utterly unlike any others and completely baffled the 

 earlier investigators. Bacteriologists usually grow 

 their organisms on gelatine or some similar medium. 

 But this plan invariably failed to bring out the 



