ON THE MIXED HERBAGE OE PERMANENT MEADOW. 
1347 
over plot 9 was all in tlie later years. Over the 20 years, however, the excess of 
silica, taken np where the straw was used, was only about one-eighth as much as it 
would contain; and the fact that the excess was all in the later years is consistent 
with the supposition that the silica of the straw would be only very gradually available. 
A similar argument applies to the increased assimilation of nitrogen, which was much 
more in the later than in the earlier years. 
An examination of the comparative botany of the two plots will afford useful 
evidence in aid of the explanation of these results. 
The Table LXXIL, pp. 1348-9, shows the percentage of each species at each separa¬ 
tion, the produce per acre, the increase over that without manure; and, finally, the 
increase or decrease over that on plot 9 with the same manure, excepting the straw. 
As with the ammonia-salts and mineral manure without the straw, so with it, we 
have, but even in a greater degree, a very marked reduction in the number of species 
found in the samples. We have, on the average, three fewer Graminese, two fewer 
Leguminosse, and 18 fewer Miscelianese, in all 23 fewer species, than without manure; 
and there are two fewer Miscelianese than on plot 9 without the straw. The average 
number of species found on plot 13 has been 26, instead of 49 without manure; and 
the range in the several separation-years has been from 30 in the first (1862), to 22 
only in the last (1877). 
With the straw, the average percentage of Graminese in the total herbage is higher, 
that of Leguminosse equally low, and that of Miscelianese lower, than without it. 
With the straw the highest average percentage of Graminese was more than 95 in the 
third separation-year, and it was 91 taking the average of the four years. The highest 
average percentage of Leguminosse was also in the third separation-year, but it did 
not then reach to one-third, and averaged less than one-fifth, of 1 per cent. The 
highest percentage of Miscelianese was 13 5 6, in the second separation-year, the lowest 
4‘29, in the third year, and the average 8*77. 
Compared with the unmanured plot 3, there was an average of more than 4000 lbs. 
per acre per annum excess of Graminese, of nearly 200 lbs. deficiency of Leguminosse, 
and a small deficiency of total Miscelianese also. Compared with plot 9, there were 
between 600 and 700 lbs. average increase of gramineous herbage, practically the same 
average amount of leguminous herbage, but a deficiency of total miscellaneous. 
Looking to the composition of the herbage in more detail, it is to be observed that 
this almost exclusively gramineous produce was characterised by containing a very 
large and increasing percentage of Dactylis glomerata . Next in order, but with less 
than half the average quantity, came Agrostis vulgaris , which was in fairly uniform 
amount throughout, and next Poo, pratensis in less, but in increasing proportion. 
Though yielding less average quantities than the foregoing, Arena elatior and 
Alopecurus pratensis have gained ground in a marked degree ; whilst Poa trivial is, 
instead of increasing, as did Poa pratensis, has in a striking degree diminished, indeed, 
almost disappeared. Avena pubescens, Avena flaveseens, and Festuca pratensis , have 
