E. J. RUSSELT, AND A. APPLRYARD 7 



It is hardly likely on physical grounds that these atmospheres are 

 abruptly parted at the surface of the film; it is more probable that 

 the free air changes in composition at the surface of the particles 

 where a thin layer of it is to some degree in ecjuilibrium with the 

 dissolved air. The stillness of the soil air is favourable to the formation 

 of a stratum different in composition from the bulk and merging 

 insensibly into it. 



The very small amount of oxygen in the dissolved gas is evidence 

 that the rate of consumjition of oxygen in the solution is greater than 

 the rate at which fresh supplies come in from the soil air, a fact of great 

 biochemical significance. But still more important for our present 

 purpose is the fact of the existence of this atmosphere almost devoid 

 of oxygen. 



We are accixstomed to think of a drained cultivated soil as being 

 under essentially aerobic conditions, and the analyses of the free air 

 show that this view is correct. But the existence of this second 

 atmosphere enables an organism that wants anaerobic conditions to 

 find them by submerging itself into the medium in which this atmosphere 

 is dissolved, especially if at the same time it associates itself with an 

 aerobic form capable of taking up any oxygen that becomes dissolved. 

 Thus alongside of the aerobic life in the soil there is the ijossibility of 

 anaerobic life, and we can no longer dismiss a possible soil change as 

 unlikely simply on the grounds that it requires anaerobic conditions. 

 In the present paper we confine ourselves to the free air in the soil 

 but hope to deal with the dissolved air later on. 



The free, air in the soil. 



For the first examination of the free air of the soil we have to turn, 

 as in many other agricultural studies, to the papers of Boiissingault. 

 In 1853 he published^ the results of analyses of 36 samples of soil gas 

 taken at a depth of 30-40 cms. At that time Bunsen's classical memoir 

 had not been published nor had gas analysis methods been worked out, 

 so that he was compelled to fix a pipe in the soil (thus causing con- 

 siderable disturbance) and periodically to aspirate a large volume 

 (2J to 10 litres) of soil air through baryta water and weigh the carbonate 

 formed. The method must have been cumbersome to work ; neverthe- 

 less the results are fairly close to ours, the air obtained from soils 



' Boussingault and Lewy, 'Memoire sur la composition de I'air confine dans la tone 

 vegetale,' Annates de Chimie et de Physique, 185.3, 37, 5-50. 



