ON THE MIXED HERBAGE OF PERMANENT MEADOW. 
1403 
The first separation was in the third year of the experiment, when 40 species were 
found in the sample, there being in that year in the sample from the unmanured 
plot (3) six fewer than the average, namely, only 43. In the second separation-year 
(1872) there were 44 species in the sample from plot 18, and 49 in that from plot 3. 
In the third year, or the thirteenth of the experiment, there were only 39 in the 
sample from plot 18, but 52 in that from plot 3. There was thus a deficiency in the 
number of species on the manured plot compared with the unmanured, in the first 
separation-year of three, in the second of five, and in the third of 13. This, as the 
actual number of species in each separation-year shows, does not represent anything 
like a corresponding decline in the actual number of species on the manured plot; the 
difference between the number on it and that on the unmanured plot being, in fact, 
due to a considerable increase in the later years in the number on the unmanured, 
doubtless owing to the decreasing intensity of the struggle without any manure. 
The increase thus resulting was especially among the Miscellanese; of which there 
were in the samples from the unmanured produce 24 in the first, 28 in the second, 
and 31 in the third, of the three separation-years ; whilst in those from plot 18, the 
numbers ranged 21, 22, and 21; that is practically showing no change. Upon the 
whole, then, with somewhat intensified struggle on the manured plot, there are fewer 
species than without manure; though, with the competition comparatively limited, 
there is fluctuation according to season, rather than any marked diminution, from one 
separation-year to another, in the actual number of species coming into the samples. 
It will be observed that, with comparatively limited manuring, the nitrogen 
supplied as ammonia-salts, and comparatively limited luxuriance and struggle, there 
were only nine grasses which yielded more than 1 per cent, to the produce in one or 
other of the separation-years. With the less intensity of struggle without manure, 
there were, however, 12 ; and on plot 16, with more nitrogen and more mineral matter 
supplied, but the nitrogen applied as nitrate of soda, and with much more active 
luxuriance than on plot 18 with the ammonia-salts, there were, nevertheless, 11. 
This is, no doubt, explained by the fact frequently before referred to, that the nitrogen 
of ammonia-salts distributes much less rapidly and freely than that of nitrate of soda ; 
and, therefore, even with less luxuriance, and consequently less active competition, the 
growth of a more limited number is favoured. 
The percentage of grasses in the total produce, nevertheless, increased considerably 
from one separation-year to another. The increase was, however, almost exclusively 
due to the increased proportion of the poor and comparatively surface-rooting Festuca 
ovina, Agrostis vulgaris, Holcus lanatus , and Anthoxanthum odoratum ; which col¬ 
lectively contributed in 1867, 32 ; in 1872, 60; and in 1877, 67 per cent, to the total 
herbage. Besides these, Lolium perenne, Arena jlavescens and pubescens, Poa trivialis, 
and Dactyhs glomerata, contributed comparatively small but fluctuating amounts, 
without very obvious tendency either to increase or to decrease; though Poa trivialis, 
Arena pubescens, and Dactylis glomerata^ showed the greater tendency to decrease. 
