45 



drawn through the soil. 



It is probable that this occurrence of injury 

 with the soil atmospheres/of lower oxygen content, even 

 when these were changed frequently, is due to insuffi- 

 cient diffusion of the oxygen in the soil itself. The 

 place where oxygen must be supplied is at the roots, not 

 merely within the pot. It is probable that the internal 

 distribution of the oxygen through the soil is regulated 

 by the same diffusional principles which have been de- 

 scribed as controlling the exchange of oxygen between 

 field soils and the general atmosphere. In the experi- 

 ments with atmospheres of low oxygen content the oxygen 

 may have disappeared altogether in parts of the soil 

 although still available in other parts or in the sealed 

 space above the soil. For satisfactory determination 

 of these questions it would be necessary to experiment 

 with moving atmospheres or with special devices for 

 maintaining a substantially uniform gas composition 

 in all parts of the soil. 



The hypothesis that injury is due merely to 

 interference with the respiration of the root proto- 

 plasm is supported by a general impression derived from 

 the experimental results to the effect that injury is 

 likely to be manifested a little more quickly with 

 plants of large root systems than with plants the roots 

 of which are few. The available data do not have suf- 

 ficient quantitative accuracy to permit actual test of 



