1398 
MESSRS. J. B. LAWES, J. H. GILBERT, AND M. T. MASTERS, 
So far as the Miscellanea are concerned, the general result on both plots would seem 
to be, as was the case with the Graminese, to approximate more nearly to the condition 
of plot 16. With the longer continuance of the treatment on plot 16, and with an 
increasing competition on the part of the grasses and Leguminosse, the Miscellanese 
collectively declined to a lower point on it than on either of the two plots more 
recently brought under experiment. The previously most prominent species on plot 
16, Conopodium, Rumex , and Achillea , on it declined, but, so far, they have either 
increased or remained stationary on the newer plots. 
Referring to the actual yield per acre on the three plots, though there is considerable 
fluctuation in the amount of produce contributed by the three main groups of plants, 
plot 16 has, on the average, not only of the two complete separation-years, but also 
including those of partial separation, yielded considerably more total gramineous 
herbage, more leguminous herbage, and about the same amount of miscellaneous. 
It would be of little use to compare the actual produce of individual species on the 
three plots with that without manure, though the results are given for reference in 
the table. The comparison of the produce of the different groups and species on each 
of the two new plots with that on 16, is of much more interest. The most striking 
points indicated are that, although there was a much greater amount of both Agrostis 
vulgaris and Festuca ovina on both plots 19 and 20 than on plot 16 in the first 
separation-year (1872, which was also the first year of the experiments on 19 and 20), 
there was, in the second separation-year (1877), on both a less amount, especially of 
Agrostis vulgaris. The fact is, that whilst both these grasses had increased considerably 
from the first to the second of these two separation-years on the older plot 16, each 
greatly diminished from its higher original amount on plots 19 and 20. Alopecurus 
pratensis again showed on both the new plots, and in both seasons, a great deficiency 
compared with plot 16, although on both it showed an actual and considerable 
increase. Holcus lanatus, on the other hand, was in much greater, and in more 
rapidly increasing, amount on the two plots 19 and 20 than on 16 ; and it increased 
very much the more largely on the nitrate of potass plot 20, where in the second 
separation-year it contributed more than a third of the total gramineous herbage. 
Avena jlavescens also, especially in the first year, yielded considerably less on plots 
19 and 20 than on the older plot 16, though on both it rather increased in actual 
amount. Dactylis glomerata, Lolium perenne, and Avena pubescens , were also each 
more or less in deficiency on plots 19 and 20. Cynosurus cristatus and Anthoxanthum 
odoratum were, on the other hand, in some excess. It was, in fact, in the poorer 
grasses— Festuca ovina , Agrostis vulgaris, and Holcus lanatus —that the two newer 
plots showed the greatest excess over plot 16 in the first of the two separation-years. 
In the second separation-year, however, there was either a deficiency, or a less excess 
of these, as compared with their yield on plot 16; whilst those which were in the 
greatest deficiency in the first separation-year on the newer plots compared with the 
older one, were the better grasses Alopecurus pratensis and Avena Jlavescens, each of 
