48 



Tlie plant food in tJie soil 



The crops on Pots 10 and 11 ought of course to 

 weigh the same, and so should the crops on Pots 8 and 9. 

 The differences arise from the error of the experi- 

 ment. In all experimental work, however carefully 

 carried out or however skilful the operator, there is 

 some error. 



There is clearly an increase in crop as a result of 

 storing the surface soil in a moist condition, showing 

 that additional plant food has been made since these pots 

 were put up. On the other hand it does not appear that 

 much plant food has been made in the subsoil during 

 this time. Further evidence on this point is given by 

 an experiment similar to that in Fig. 22, but where 

 mustard is grown in subsoil kept moist, but uncropped 

 for some time, and in siibsoU previously cropped with 

 rye. The results in grams were : — 



