64 Oil the Bain and Drainage - Waters at Rothamsted. 
acid accordingly. Attention was called to the fact of the 
possible loss of nitrates, and even of soil-nitrogen, which may 
thus occur in ill-drained soils in wet weather. 
It will be obvious that in the case of the artificially-manured 
plots in the experimental wheat-field, the soil will be unusually 
poor in decomposing organic matter, and that, so far, the condi- 
tions may be supposed to be little favourable for the evolution 
of free nitrogen from that source. Nor will the conditions of 
absence of gaseous oxygen, and of saturation by water, which 
are favourable to the reduction of the nitrates, frequently 
prevail. Without, however, denying that each of these actions 
may take place occasionally, or in a limited degree, we are 
disposed to give more weight to the established fact that in 
our mode of experiment and calculation the loss of nitrogen by 
drainage is under-estimated, than that there is material loss in 
other ways. 
In 1866, at the meeting of the British Association held at 
Nottingham, we discussed the results of the determinations of 
nitrogen in the samples of soil collected from the experimental 
wheat-field in 1865, in their bearing on the question of the 
amount of the nitrogen of manure which is unrecovered in 
the increase of crop. We [concluded that although some 
remained in the soil, " a considerably larger proportion would 
remain entirely unaccounted for within the soil to the depth 
under examination than was there traceable, and the probability 
was, that at any rate some of this had passed off into the drains, 
and some into the lower strata of the soil." The investigations 
which have been described in this paper, and which were under- 
taken to settle this point, have fully justified the conclusions 
then drawn. Finally, there is not sufficient evidence to show 
that, under the conditions of the experiments in question, there 
would be any other source of considerable loss. 
Thus, it has been shown, in the case of a field under con- 
tinuous wheat cropping, and receiving nitrogenous manure in 
the form of ammonium-salts or nitrate of sodium, that there was, 
with autumn sowing, even under the most favourable conditions 
as to mineral manure, on the average of a long series of years 
not quite one-third of the supplied nitrogen recovered in the 
hicreaseoi crop ; and there was much less with defective mineral 
supply. Nor, so far as can be judged, would the accumulations 
oi residue within the soil raise the amount recovered to one-half 
of that supplied. There was, on the other hand, a very great loss 
by drainage, which was very much the greater the more unfavour- 
able were the conditions of growth, and the greater the excess of 
nitrogen in the manure. In fact, under comparatively favourable 
conditions, the amount found in the drainage nearly, or quite, 
