The Herbage of Old Grass Lands. 
437 
The gi'asses were represented in the following proportions : — 
Lolium perenne 67" 
Dactvlis glomerata 26 
Agrostis sp. 6 >• 100 
Phleum pratense 1 2 
Cynosurus cristatus / 
The legnminous herbage consisted to the extent of over 98 per 
cent, of Trifolium repens, the rest being Trifolium pratense. 
The table (I.) showing the distribution of species, which is 
given on page 438, may be regarded as a summary of the results 
obtained. In it are named all the gramineous species, all the 
leguminous species, and seven of the most noticeable miscel- 
laneous species detected amongst the herbage of the twenty-five 
turfs. Seventeen species of grasses are enumerated, but, if 
Agrostis sp. be regarded as denoting two species and Festuca 
ovina et var. as denoting three species, the number is brought 
up to twenty, as compared with twenty-six species of grasses 
found upon the water meadows, or with twenty species of 
grasses which have been identified on the variously manured 
plots in Rothamsted Park. The table is, of course, strictly 
qualitative, and not in any sense quantitative. For example, 
Briza media was represented by a solitary plant found upon one 
plot only, and Poa annua, a weed grass, though recorded for 
six plots, was never present save in very insignificant quantity. 
The total number of leguminous species represented was only 
four, as compared with seven on the water meadows and ten in 
Rothamsted Park. The list of miscellaneous species might be 
continued to a much greater extent, but the seven which are 
enumerated in the table include the really significant ones, the 
others being in very small quantity and more of botanical than 
of agricultural interest. The table is further useful as being a 
census of species, recording as it does the relative frequency of 
occurrence, in different localities, of each plant mentioned. In 
this respect the premier position is taken by Lolium perenne 
amongst the grasses, and by Trifolium repens amongst the legu- 
minous plants. The greatest number of species of grass which 
occurred on any plot was ten on one of the Kent turfs (No. 2), 
on the Stafford turf (No. 6), and on the Tipperary turf (No. 8) ; 
the least number was three on the Herefordshire turf (No. 20). 
Not one of the turfs was utterly destitute of leguminous plants, 
though on fifteen of them only one species was represented, and 
only on the Wexford turf (No. 16) did all four species occur. 
It is possible, however, to present a quantitative view of the 
results, as is done in table (II.) of percentages by weight of green 
herbage (page 439). This table shows that in fifteen cases the 
