36 On the Rain and Drainage - Waters at Rotliamsted. 
nitrates of the soil furnish the chief, if not the onlj, supply of 
nitrogen available to a wheat-crop. 
The production of nitrates in an unmanured soil kept free 
from vegetation has been discussed in Part II. of the present 
paper. It then appeared that the drainage-water collected 
during the last four years from the soil drain-gauges contained 
on an average 10 7 parts of nitrogen, as nitric acid, per million 
of water. The soil of these drain-gauges was ordinary arable 
soil, undisturbed in its condition, and had been maintained for 
ten years without manure or crop, and kept free from weeds. 
When we turn to the nitrogen as nitrates found in the drainage- 
waters from the unmanured plots in Broadbalk field (Table 
XLVII.), we find that the quantity is far smaller than in the 
water from the drain-gauges. The average figures for three 
years given at the foot of the table, show that the mean amount 
of nitrogen is 3'9 per million in the case of Plot 3 & 4, per- 
manently unmanured, 4 3 per million in the case of two plots 
receiving only ash constituents, and 4'5 in the case of the unma- 
nured Plot 16. This much lower proportion of nitrates in the 
drainage-water is doubtless partly owing to the great exhaustioii 
of the nitrogen of the soil by continuous wheat-cropping with- 
out manure, but is chiefly due to the fact that the crop actively 
appropriates the nitrates formed in the soil. So complete is the 
appropriation of nitrates by the wheat-crop, that during the 
time of active growth, and for some time after, no nitric acid, 
or a trace only, can be found in the drainage-water from several 
of the plots in Broadbalk. This is well shown by the analyses 
of Frankland of drainage-waters collected on May 18, 1872 
(Table XLI.). In a later collection on June 11 of the same 
year Frankland found no nitric acid whatever in the waters from 
Plots 3 & 4, 5, 6, 7, 13. Similar results were found in the ana- 
lyses made at Rothamsted of the waters collected during the 
summers of 1878 and 1879. 
In autumn nitric acid again begins to appear in the drainage- 
waters of the plots unmanured with nitrogen, and continues 
steadily to increase if no serious loss by drainage takes place. 
Even, however, with an excessive amount of drainage a certain 
proportion of nitric acid is fairly maintained throughout the 
winter months, the production of nitrates generally keeping 
pace with their removal. The best illustration of the increase 
of nitrates during autumn on land unmanured with nitrogen, is 
afforded by the analyses of the drainage-waters from Plots 3&4, 
5, 16 and 17, in August and October 1881. The maintenance 
of the nitrates during extremely wet winters is shown by the 
analyses of waters during the winters of 1878-i) and 1880-1. 
