336 Report of Experiments on the Growth of Barley, 
nitrate of soda. It is obvious, that the nitrogen supplied by the 
dung is retained by the soil in a condition not only much less 
rapidly available to growing crops, but also much less liable to 
loss by drainage. Still, there is a large amount of the nitrogen 
supplied in the dung not yet satisfactorily accounted for. 
The Table shows that at each period of collection there was less 
nitrogen in the drainage-water from the plot the whole of which 
has been unmanured since 1851, and part for a number of years 
previously, than from either of the plots artificially manured 
during the same period. There was, in every case, rather more 
from Plot 5, which received mineral manure alone in 1852, and 
each year since ; but mineral manure in each, with ammonia-salts, 
or nitrogenous organic matter, or both, in 7 out of the 8 preceding 
years. There was, further, in each case, more nitrogen in the drain- 
age-water when, to the mineral manure, ammonia-salts = 41 lbs. 
of nitrogen was added ; with one slight exception again more 
when 82 lbs. were employed ; and more still with 123 lbs. nitrogen 
supplied. 
That is to say, with each increased supply of nitrogen by manure, 
as ammonia- salts, there was an increased loss of nitrogen as nitric 
acid in the drainage-ioater. 
It must be borne in mind that, in the experiments on wheat 
here referred to, the ammonia-salts were always sown broadcast 
in the autumn, and ploughed or harrowed in before sowing the 
seed ; and it is seen that the amount of nitrogen as nitric acid in 
the drainage- water is much greater on the three occasions of winter 
collection, that is, soon after the manures were sown, and when 
there was no growth, than on either of the two occasions of 
spring collection, that is, after the washing out by the winter 
rains, and when active growth had set in. 
The nitrate of soda is, however, always sown as a top-dressing 
about the middle of March. Accordingly, there was, in each 
case of winter collection, much less nitrogen as nitric acid in the 
drainage from the nitrated plot (9), than in that from Plot 7, 
which received the same amount of nitrogen as ammonia-salts 
applied in the autumn. On the other hand, in both cases of spring 
collection — that is, after the sowing of the nitrate— the amount of 
nitrogen as nitric acid was much greater in the drainage from the 
nitrated plot, than in that from the plot which had received 
the same amount of nitrogen as ammonia-salts in the autumn. 
In one case, indeed, April 21, 1868, the nitrate having been 
applied on March 18, the quantity of nitrogen as nitric acid in 
the drainage from the nitrated plot amounted to 5*83 parts per 
100,000 parts of water. Assuming (which, however, was pro- 
bably not the case) that an inch of rain passed as drainage of 
that strength, this would represent a loss of about 13 lbs. of nitro- 
