On the Rain and Drainage- Waters at Rothamsted. 
69 
from 31 lbs. with the most liberal mineral manure, to 43'2 lbs. 
with the ammonium-salts continuously used alone. 
25. Reckoned over thirty years, not quite one-third of the 
nitrogen supplied by manure was recovered in the increase of 
«rop under favourable conditions as to mineral manure and 
growth, and very much less when there was a deficiency of 
potash and phosphoric acid, and defective growth accordingly. 
26. With 400 lbs. of ammonium-salts, and the most liberal 
mineral manure, there was the maximum amount of nitrogen 
recovered in the crop, and the minimum amount in the drainage ; 
but with the ammonium-salts used alone, there was the minimum 
amount in the crop, and the maximum amount in the drainage. 
27. Only with the smallest quantity of ammonia applied was 
the amount of nitrogen in the total crop and drainage together, 
more than was supplied in the manure ; in all other cases there 
■was a greater or less deficiency. Besides the nitrogen supplied 
in manure (which was not entirely recovered in the crop and 
•drainage), it is estimated that on the average about 30 lbs. would 
be contributed per acre per annum by the soil and by rain, 
and condensation of combined nitrogen from the atmosphere ; 
perhaps more by the soil in the earlier, than in the later years. 
28. Analyses of the soils of the different plots, made at dif- 
ferent periods, show that the amount of nitrogen was consider- 
ably reduced where no nitrogenous manure was applied ; but 
^'here nitrogenous manures were applied, the amount remained 
stationary, or slightly increased, or diminished, according to 
the condition of the soil as to mineral constituents, and to the 
amount of growth. In fact, the difference in the amount of 
nitrogen in the soils of the plots with ammonia applied, com- 
pared with that where none was applied, bore a close relation 
to the amount of growth, and was mainly due to the residue of 
the crops. 
29. The amount of nitrogen in the crops, and estimated to 
be lost in the drainage, together with the excess in the soil 
where it was supplied in manure, is not sufficient to account 
for the whole of that so applied, and that available from other 
sources ; but there is evidence that, reckoned according to the 
composition of the collected drainage-waters, the estimates of 
the loss of nitrogen by drainage are too low. 
30. When farmyard-manure is largely used, there is some- 
times considerable loss of nitrogen, due to the decomposition of 
nitrogenous organic matter, and the evolution of free nitrogen ; 
or when the soil is saturated with water, or imperfectly aerated, 
there may be destruction of nitric acid and evolution of free 
nitrogen. It is believed that, under the conditions of the arti- 
ficially manured plots in the Experimental Wheat-field, there 
