55 



THE COMK'SITIOB OF THE SOIL AIR. 



it was pointed out in the Introduction that 

 there were two ways of experimental attack on the 

 problems of soil aeration and root respiration; first, 

 "by analysis of actual soil air; second, by experimental 

 study of the respiratory behavior of roots. The errors 

 inherent in the sampling of the soil air are considerable, 

 especially when it is desired to draw it from consider- 

 able depths in undisturbed soil or from other situa- 

 tions where anaerobiosis of roots is to be expected, if 

 anywhere. lor this reason the analysis of soil air does 

 not seem the most promising line of attack on the problem 

 and was not employed in the present investigation. It 

 will be interesting, nevertheless, to review briefly 

 the available data as to its composition. This data in- 

 volves two things, the results of actual analyses and 

 more general considerations (experimental and theoretical) 

 concerning the causes which modify the soil air and deter- 

 mine its composition. 



The first extensive investigation of the composi- 

 tion of the soil air was that of Boussingault and 

 Lewy. In analyses of air from fourteen soils these in- 

 vestigators found oxygen contents between 10.35 and 

 20.03 percent, carbon dioxide contents between 0.72 



H Inn. ehim. phys. 47: 5-50 (1853); 

 Boussingault,- Agronomie, vol. 2, pp. 68ff. (1855) 



