40 



The Atiiiosplicrc of the Soil 



lowering the mercury reservoir. The first 20-30 c.c. is rejected aud tlie 

 next 25 c.c. is collected over mercury in thick-walled test tubes, which 

 are then placed in small crucibles and transported in a rack to the 

 laboratory for analysis. To prevent the rack from being blown over 

 by winds it is held firmly in the ground by iron spikes passing through 

 the base pieces. Only one sample is collected at each point. Successive 

 samples vary slightly in composition (Table X) but a fairly large 

 volume of air of tolerably uniform composition can if desired be with- 

 drawn from the same hole. 



Table X. Percentage composition of successive 30 c.c. samples of sod 

 air draum from the same hole. 



As a rule samples of air from 8-12 holes on each plot are drawn 

 and analysed separately, and the mean value i.s taken to represent 

 fairly accurately the composition of the soil air. These mean values 

 are given in Table VI and plotted in the various Figures 1 to 6. 



Samples were drawn from all the plots on the same day so that the 

 values are strictly comparable. The variation from place to place is 

 fairly large, especially on the plot which has received annually 14 tons 

 of farmyard manure, but on the unmanured plot it is comparatively 

 narrow. 



