228 



NITRIFICATIOX 



the other hand, the rapid decHne in the production of Plots 

 5, 6, 7, and 8, following the discontinuance of the manure, is 

 more consistent with the existence of a residue of nitrate 

 rather than of slowly-decaying organic material derived from 

 the haulm of the potatoes. 



Soil samples down to a dej^th of six times 9 inches were 

 taken from some of these plots in February 1903, after the first 

 barley crop had been removed. The following table shows the 

 average figures olitained. 



Table LXXXI. — Nitrogen as Nitrates per million in dry Soil of former 

 Potato Plots. 



In this case the samples were taken in the early spring after 

 a winter of fair rainfall, which had distriljuted the nitrates 

 throughout the soil ; the total, however, present in the first six 

 depths roughly corresponds to the variations in the Ijarley crop 

 of 1903 which followed. 



One striking fact in connection with these and similar 

 determinations, is the absence of any lateral diftusion of the 

 nitrates in the subsoil water beneath the plots. The Broadbalk 

 wheat plots, for example, are comjDaratively narrow, l.)eing about 

 7 yards in breadth, separated by paths of 4 feet in width. Yet, 

 as is seen in Table LXXVIIL, Plot 5, which receives no 

 nitrogen, shows no trace of the influx of nitrates by diffusion 

 froju the much richer subsoil water IdcIow the immediately 

 contiguous Plot 6, even down to the depth of 9 feet. Just 



