1298 
MESSRS. J. B. LAWES, J. H. GILBERT, AND M. T. MASTERS, 
actual amount grown is much greater in both sets of plants than would be gathered 
from their percentage relations alone. 
Turning to plot 12, the reduction in the actual yield of the grasses is seen to be 
very considerable. It is also more uniform from the first to the third separation-year 
than on plot 3, and the amount is much the same in the fourth as in the third, and 
only about three-fifths as much as in the first. Of leguminous herbage there was, on 
the average, more yielded on plot 12 than on plot 3. The fluctuations in amount from 
year to year were great, and not accordant in order with those on plot 3. Of miscel¬ 
laneous herbage, again, there was, on the average, more on plot 12 than on plot 3, and 
there were also considerable fluctuations from year to year; hut these were not so dis¬ 
cordant in their order with those on plot 3, as in the case of the Leguminosae. On both 
plots there was a considerable increase in the amount of miscellaneous herbage in the 
second year compared with the first, and on both there was much less in the third and 
fourth than in the first year. In the last two separation-years there was more on 
plot 12 than on plot 3. 
The general result is, then, that on botli plots there was a tendency to considerable 
reduction in the amount of both the Graminese and the Miscellanese in the later years, 
with less marked, if any, reduction in the much smaller actual quantity of leguminous 
herbage. The resultant is a considerable reduction in the yield of the total mixed 
herbage. 
It is obvious that, considering the fluctuations from year to year according to season, 
the results obtained in four individual years at stated intervals might be misleading 
in considering whether there has or has not been a real increase or decrease in the 
growth with the continuance of the unmanured condition. But it may be here 
remarked that the decided indications of decrease in growth manifested in the tables 
under consideration are fully confirmed by the amounts of produce of hay, of nitrogen, 
and of mineral matter, obtained in each of the first 20 years of the experiments (see 
Part I., p. 300, and context). 
It will be observed that the actual quantities yielded are, on the average, con¬ 
siderably greater on plot 12 than on plot 3, and, as has been pointed out at p. 301, 
Part I., there is reason to believe that the soil of, at any rate, a part of plot 12 was 
es made ground,” which had a greater depth of mould than in the case of the plots 
generally. For this reason the results obtained on plot 3, instead of the mean results 
of the two unmanured plots, will be taken as the standard by which to compare those 
obtained on the manured land. 
We now turn to the percentage, and the actual amounts, of the different species 
occurring on the two plots. It must be remembered that the first year of complete 
separation (1862) was the seventh year of the experiments, and hence the distribution 
of plants then found cannot be taken as indicating the exact state of the flora at the 
commencement of the experiments. At the same time it must be borne in mind that 
