76 The Food of our Agricultural Crops. 
manures had lost upon tlie acre 506 lb. more nitrogen than the 
unmanured land ; this would be equal to the annual loss of 
25 - 3 lb. The analysis of the hay removed from the two plots 
showed that that of the mineral plot contained 23'6 lb. more 
nitrogen than the hay of the unmanured plot, a quantity corre- 
sponding very closely with the greater loss of nitrogen in the 
mineral manured soil. Too much reliance must not, however, be 
placed upon so complicated an operation as soil analysis, though 
the evidence points to the conclusion that the larger amount of 
nitrogen taken off the land by the leguminous herbage on the 
minei'al plot was derived from the soil, and not from the atmo- 
sphere. 
Although we found it impossible to grow red clover year 
after year upon the same arable soil of the farm, whatever might 
be the amount or composition of the manure applied, we found 
no difficulty in growing other leguminous plants upon the 
clover-sick land — the importance of which fact is very obvious, 
as it enables us to continue removing large quantities of nitrogen 
in leguminous crops year after year, instead of having to wait for 
some years until the land will again grow red clover. 
It is evident, if the soil of our arable fields is the source of 
nitrogen, and the manures which we apply to them do not con- 
tain any, that, sooner or later, the stock of nitrogen in our soils 
will be greatly diminished ; even to an extent which may 
eventually be measured by analysis. It is, unfortunately, the 
case that leguminous crops, such as lucerne, sainfoin, &c, which 
carry off the largest amount of nitrogen, are also those whose 
roots penetrate the deepest down into the subsoil ; also, that 
the subsoil of my experimental fields varies in character so much, 
even where samples are taken with the utmost care, and at no 
great distance apart, that at present any comparative measure- 
ments of the loss of nitrogen in the soil are not to be trusted, 
and the experiments must be carried on for some years more 
before we can expect to give a satisfactory answer to this 
important question. 
The general results of our experiments appear to show that 
while mineral manures, especially those containing potash, 
produce a large increase in leguminous crops, neither salts of 
ammonia, nitrate of soda, nor rape-cake, which furnish organic 
nitrogen, appear to have any decided influence upon them. This 
indicates a very distinct difference in the source from which the 
cereal and leguminous crops obtain their nitrogen ; for, in the 
same field, and within a few feet of each other, nitrate of soda 
alone applied to the continuous barley crop has produced 
;;n average of 33 bushels per acre for thirty-six years in 
