EXPERIMENTS OX THE MIXED HERBAGE OF PERMANENT MEADOW. 365 
and the increase of mineral matter removed is not much more than half as much as 
would be contained in 1 ton of hay. 
Whilst it is clear from the above comparisons, that the annual supply of not only the 
mineral constituents, but the nitrogen also, of 1 ton of hay, yielded less than two- 
thirds instead of 1 ton of increase of produce, it is at the same time pretty certain 
that the differences in the amounts of hay, nitrogen, and mineral matter, obtained, 
without and with the manure, do not accurately represent the proportion of the con¬ 
stituents supplied which contributed to the result; for, one effect of the manure was 
very materially to modify the character of the herbage. The percentage in the pro¬ 
duce, and the amount per acre, of gramineous herbage, were much increased; both the 
percentage, and the actual amounts, of leguminous herbage were much reduced ; the 
percentage of the miscellaneous herbage was also reduced, though the amount of it 
per acre was increased. Thus, whilst the amount of leguminous herbage, which is to 
a great extent independent of direct supplies of nitrogen by manure, was reduced, the 
increase of produce consisted almost exclusively of gramineous herbage, the increased 
growth of which would doubtless depend—mainly, at any rate—for the nitrogen it 
required, on that which was artifically supplied ; and, therefore, probably a larger 
amount of that supplied contributed to the increase than the table shows. 
As in the case of the nitrogen, so also in that of each of the mineral constituents, 
there was always very much less, and sometimes more than one-half less, increased 
amount in the produce removed, than was supplied in the manure ; and in the case of 
some important constituents—potass, for example—a less proportion of that supplied 
was recovered as increase than when larger amounts of both the mineral and the 
nitrogenous manures were applied, and when, consequently, greater luxuriance of the 
grasses was induced. And, notwithstanding potass is very little subject to loss by 
drainage, and there would therefore be an annually increasing residue of it within the 
soil, there is even less of it taken up during the second than during the first half of 
the period of experiment. There is, nevertheless, a higher, but a decreasingly higher, 
percentage of potass in the manured, than in the unmanured produce. There is also a 
higher percentage of soda, of phosphoric acid, of sulphuric acid, and of chlorine, but a 
lower percentage of lime, magnesia, and silica. 
These results suggest several points of interest. 
It has already been shown, in experiments in which comparatively large quantities 
of ammonia-salts or nitrate of soda were used, how much greater was the increase 
of produce, how much more mixed was the herbage, and how much better it was 
developed, when a given amount of nitrogen was applied as nitrate of soda than when 
as ammonia-salts ; the difference in favour of the nitrate in these respects being much 
the greater when these nitrogenous manures were respectively used alone—that is, 
without the conjunction of mineral manures. On plot 18—with nearly one-fifth less 
nitrogen annually applied as ammonia-salts than on either plots 16, 19, and 20 as nitrate 
of soda, and therefore with less forced luxuriance, but, it is true, at the same time, with 
