\4o 



this suggestion. All that can be said is that the plants 

 which appeared to have the la ger root systems were in 

 several cases injured more quickly. In all cases the 

 time required for injury to become manifest is subject 

 to much variation, as is indicated by the detailed data 

 given on page A • Probably these variations are 

 due to incomplete removal of the original soil atmosphere 

 at the beginning of an experiment. The removal of this 

 original atmosphere must be accomplished by sweeping it 

 out with the nitrogen (or other gas) which is to replace 

 it. This replacement cannot be expected to be entirely 

 complete unless an impossibly la^rge amount of nitrogen 

 is run through. Accordingly a certain amount of oxy- 

 gen will be left behind and must be used up by respira- 

 tion before injury will occur. The actual time of ap- 

 pearance of injury will depend upon (1) the amount of 

 oxygen left in the pot at the beginning of the experiment, 

 (2) the rate at which this oxygen i3 used up, and (3) 

 the degree to which the oxygen content in different 

 parts of the pot is equalized by diffusion or in some 

 other way. 



Whether ofc not the injury to Coleus be regarded 

 as due to a disturbance of root respiration, it is inter- 

 esting that the local injury manifests itself in the 

 economy of the plant as an interference with water- 

 absorption by the roots. This follows both from the main 



symptomf of injury, namely wilting. and from the observed 



