42 



PHYSIOLOGICAL CONCLUSIONS. 



There can be almost no doubt that the injuries 

 observed with Coleus and Heliotropium were actually 

 caused by deficient oxygen in the soil. The harmlessneso 

 of the technique of sealing and the suitability of the 

 experimental conditions for the growth of the plants 

 are attested by the satisfactory growth of many controls 

 and by the recovery of several injured plants af-ter the 

 admission of oxygen. Any theoretical possibility of 

 error due to a specific poisonous actiQn of the kind of 

 nitrogen used or of any impurity in it is removed by 

 the agreement between the effects of commercial nitro- 

 gen and of nitrogen prepared from air# Any other form 

 of accidental injury is excluded by the complete agree- 

 ment of the several independent tests. 



It is conceivable that the injury caused by 

 the removal of oxygen may be either a direct response 

 of the roots themselves to oxygen deficiency or an in- 

 direct effect through some change produced in the soil* 

 The exclusion of oxygen from the soil probably has im- 

 portant effects on the quantity and character of the soil 

 micro-organisms and it is possible that such alterations 

 of the micro-flora may react upon the plant roots* How- 

 ever, any modification of the soil micro-flora or any 

 similar changes which might be responsible for secondary 

 effects on the roots would be expected to be different 



in different soils, The original floras of any two soils 



