OK THE MIXED HERBAGE OF PERMAKEKT MEADOW. 
1401 
The plan of experiment adopted was, to supply not only as much of each of the 
mineral constituents, but also as much nitrogen, as would be contained in one ton of 
hay, and to estimate the amount of increase of produce, and of the constituents con¬ 
tained in that increase. In carrying out the experiments, the constituents were 
supplied in chloride of potassium, sulphate of magnesia, bone-ash and sulphuric acid 
mixed as superphosphate, silicate of soda, silicate of lime, and ammonia-salts. The 
quantities of these supplied contained, as nearly as could be calculated, the full amount 
of potass, magnesia, lime, phosphoric acid, silica, and nitrogen, contained in one ton 
of hay of average composition ; and the quantities of the various substances requisite 
for this supplied at the same time a considerable excess of soda, sulphuric acid, and 
chlorine. 
Stated broadly, the result is that there was, up to 1880 inclusive, that is, over 
16 years, so far as can be calculated, only about three-fourths of a ton instead of 
one ton of annual increase of produce due to the manure. This was so, notwithstanding 
that the nitrogen as well as the mineral constituents of a ton of hay had been 
annually supplied. Of the nitrogen supplied, only about half, but of most of the 
mineral constituents more than half of those supplied, are estimated to have been 
recovered in the increase. As pointed out in Part I., however, as the flora changed 
considerably compared with that without manure, there is some uncertainty in the 
estimation of the produce, and of its constituents, which are to be referred to increase 
due to the manure. Thus, under the influence of the manure, there is a much larger 
proportion and amount of gramineous herbage, which, though it contains a compara¬ 
tively low percentage of nitrogen, is nevertheless greatly influenced by nitrogen 
artificially supplied ; and hence the proportion of that supplied which was actually 
taken up, and contributed to the increase, may be greater than is estimated. But it 
may be observed that a less proportion of some of the more important mineral con¬ 
stituents—potass, for example—was recovered, with this carefully balanced but not 
excessive supply, than when larger quantities of mineral matter and nitrogen were 
applied, and when greater general luxuriance was, in consequence, induced. 
The first year of the experiment being 1865, which was the tenth of the main series, 
we have the results of three complete botanical separations—in 1867, 1872, and 1877. 
The percentage and acreage results of these are given in the following Table, LXXXI., 
p. 1402; as also is the increase in actual yield of the various component species, over 
that on the unmanured plot 3, in the same seasons. 
