74 



DEAINAGE WATER 



Deherain's experiments were commenced in 1892. His method 

 of observation was to use 20 large water-tight cases 2 metres square 

 and 1 metre deep, each holding 4 cubic metres, or about 5 tons of soil. 

 These were filled with soil and subsoil. Access was provided to one 

 side by means of a sunk path, so that the drainage water could be col- 

 lected from the bottom of the cases. The results obtained during the 

 first year were probably untrustworthy because of the increased nitri- 

 fication due to the trituration and aeration of the soil in filling the 

 cases Eeports as to the results are published annually in the 

 " Annales Agronomiques ". The Grignon soil is rich in organic matter 

 and capable, if moist, of yielding large quantities of nitrates. 



In the year March, 1896 to March. 1897, the results which follow 

 were obtained. 1 



The effect of vegetation in checking the loss of nitrates is clearly 

 evident in these results. This it does in several ways : (i) By 

 actually absorbing the nitrates ; (ii) by lessening the amount of 

 drainage by increasing the evaporation ; and (iii) in checking nitrifi- 

 cation by rendering the soil too dry. This last effect is most marked 

 when the period of most active growth corresponds with a dry portion 

 of the year and is least evident when rain is abundant during that 

 period. [Compare the results with maize (with late development) 

 the crop of which is estimated to contain 156 Ib. nitrogen per acre ; 

 this added to 20 Ib., the amount lost in drainage, gives 176 Ib. 



1 Jour. Chem. Soc., 1897. Abstracts, ii. 591. 



