E. J. Russell axd A. Applkvard 



11 



The minimum amounts of carbon dioxide (0-04, 0-05 and 0-1 "2 pei' 

 cent, in the sand, loam, and peat respectively) were found in February, 

 the maximum (0-18, 0-31, and 0-81 per cent.) in July and August: no 

 spring maximum was observed, but this might easily have been missed 

 in the five weeks that elapsed between the taking of the May and 

 the June samples. Some of the plots were planted and some not : the 

 former contained more carbon dioxide than the latter, even in the 

 summer ; a result directly opposite to that obtained by Wollny. 



Jodidi and Wells adopted Orsat's simpler form of the apparatus, 

 and made a great number of analyses of the soil air from certain plots 

 at Ames, Iowa, over the period April to August, 1910. The mean of 

 all the results showed that at a depth of 7 inches the percentage of 

 oxygen is 20"51, of carbon dioxide 0-25, and of nitrogen 79-24. 



These various results are set out in Table IV and taken in conjunction 

 with our own (Table VI) they establish beyond any reasonable doubt 

 the close similarity between the soil air and the atmospheric air so far 

 as oxygen and nitrogen content are concerned. 



Table IV. Mean composition of soil air. 



These figures are the means of the averages of the various plots. 



