370 Report of Experiments on the Growth of Barley, 
Not only, then, do the results point to a satisfactory explanation 
of the loss of nitrogen which has been observed in the use of 
artificial nitrogenous manures, but also of the much greater loss 
when they are applied in the autumn for wheat, than when in 
the spring for barley or for oats. In confirmation of the expla- 
nation on the latter point, may be cited the facts that, not only 
was there on the average much more increase even of wheat, and 
much more nitrogen recovered in the increase, when a given 
amount of it was applied as nitrate of soda in the spring than 
when as ammonia-salts in the autumn, but the difference in 
favour of the spring-sown manure was especially marked after 
unusually wet autumns and winters. 
There is another point to notice in connection with the action 
of nitrate of soda. A given surface of soil has much less 
power to retain either nitrate of soda, or other nitrates, than am- 
monia, and so far their nitrogen is, ceteris paribus, more liable 
to loss by drainage. Yet, when frequently used on the same 
land, such was the effect of the nitrate, or its products of decom- 
position, aided by increased development of root, in causing the 
disintegration, and so increasing the porosity and surface of the 
clay subsoil, that there would appear to have been not only a 
greater retention of moisture in an available form by the subsoil, 
rendering the growing crop more independent of drought, but 
also a greater retention of nitrates than would be anticipated 
considering their solubility, and, hence, a more lasting effect 
from previous applications than would otherwise be expected. 
On the other hand, where, as in the case of the experiments at 
Rothamsted, nitrate of soda has been used in large quantities so 
many years in succession, the surface soil has retained so much 
moisture as to be difficult to work after wet weather. 
The results have shown that a considerably less proportion of 
the nitrogen applied as rape-cake, than as either ammonia-salts or 
nitrate of soda, was recovered in the increase of crop within a 
given period of time, and again considerably less of that applied 
in farmyard manure than in rape-cake. Owing to the slow 
decomposition of the nitrogenous organic matter of these ma- 
nures, their nitrogen is necessarily but slowly available. It 
would appear, however, to be, at the same time, less subject to 
loss by drainage ; and analysis has shown that a large pro- 
portion of their nitrogen is retained by the soil, becoming but 
very gradually available for a considerable length of time. 
Indeed, analysis showed that where farmyard manure had been 
applied for wheat every year for twenty-five years in succession, 
the top 9 inches of soil contained nearly twice as high a per- 
centage of nitrogen as the corresponding layers of any of the 
artificially manured plots, which, though they received much 
less nitrogen annually, as ammonia-salts or nitrate of soda, 
