ON THE MIXED HERBAGE OF PERMANENT MEADOW. 
1375 
several of the mineral constituents taken up declined in the later years in about the 
same proportion on plot 4-2 as on plot 5, and the quantity of potass (which was not 
supplied) did so even in a much greater degree, it is obvious that there was still very 
considerable deficiency of some of the mineral constituents required. 
The following Table (LXXVII., pp. 1376-7) gives the botanical results in the same 
form as usual; and it further shows the amount of yield of each species on plot 4-2 
compared with that on plot 5 without the “ superphosphate.” 
There is an average of 19 fewer total species on 4—2 with the ammonia-salts and 
superphosphate of lime, than without manure. The average number of grasses was 
reduced by three, that of Leguminosse by two, and that of Miscellanese by 14. 
There were even, on the average, one grass, one Leguminosse, and one Miscellaneous 
species, less than on plot 5 with the ammonia-salts without superphosphate. The 
average number of species found on the plot was 30, reducing from one separation- 
year to another, as follows : 35, 30, 28, 26. 
The produce showed a higher proportion of gramineous herbage from one separation- 
year to another, and higher than with the ammonia-salts alone. 
Leguminosse were scarcely represented; Lathyrus pmtensis being the only legu¬ 
minous plant which has come into the samples in each separation-year, and then in only 
insignificant proportion. 
Of Miscellaneous species only five came into the list of those yielding more than 1 
per cent, in any one year, whilst there were seven such with the ammonia-salts alone. 
The results relating to the Grand nese are very striking. As with the ammonia-salts 
alone, so now with ammonia-salts and superphosphate of lime, Festuca ovina is not 
only, on the average, the most prominent grass, but it has enormously increased in 
predominance, yielding more than 50 per cent, to the produce on both plots in the 
last separation-year. On both plots, again, Agrostis vulgaris is the second in predomi¬ 
nance ; and it also increased, but in a much less degree, than Festuca ovina. It 
increased more in percentage, but not in actual weight, on plot 5 with the ammonia- 
salts alone, than where they were used in conjunction with superphosphate of lime; 
whilst the Festuca ovina increased the most under the latter condition. In the later, 
wetter seasons, the Agrostis has apparently gained in relative predominance. The 
only other grasses which yielded even moderate average percentages to the herbage on 
plot 4-2 were H&lcus lanatus, Alopecurus pratensis , and Poa trivialis, each of which 
however, declined very considerably in the later years ; whilst Avena pubescens , Lolium 
perenne, and Avena flavescens, which were fairly represented in the first separation- 
year, contributed scarcely anything in the last. Besides these, A7ithoxanthum odo- 
ratum , Avena elatior , Poa pratensis , and Dactylis glomerata yielded, upon the whole, 
small but fluctuating percentages. 
In the last separation-year more than 82 per cent, of the produce on plot 5, with 
the ammonia-salts alone, consisted of Festuca ovina and Agrostis vulgaris together, and 
nearly 80 per cent, did so on plot 4-2 ; but, as already said, the produce on the latter 
8 N 2 
