OK THE MIXED HERBAGE OF PERMANENT MEADOW. 
1187 
when the crops were fed by sheep, observations were made as to what species were 
preferred and what were discarded by the animals. 
Instructions were also given carefully to look for, and to note, throughout the whole 
enquiry, any characters of growth, normal or induced, above-ground or under-ground, 
by virtue of which it was probable that any one species dominated over others, or in 
consequence of which the plants dominated over had succumbed. 
In 1862, 20 samples of the mixed herbage, of 10 lbs. each, were submitted to 
botanical separation ; and the work occupied Mr. Sutherland, several boys, and 
occasionally Mr. Willis, for about four months ; and Mr. Willis and the boys for 
about two months more. The results so obtained were published in Vol. xxiv., Part I. 
of the Journal of the Royal Agricultural Society of England (1863). But, even in 
the case of these separations, which were so much more elaborate than the earlier ones, 
the quantity of undetermined stem, leaf, and shedded flowers and seeds varied from 
under 5 to nearly 28 per cent, of the total sample. These at first undetermined 
residues have, however, since been much farther worked down ; and the results given 
in the present paper relating to the produce of 1862 are, accordingly, considerably 
amended as compared with those originally published as above referred to. 
In 1867, 20 samples of 10 lbs. each and one of 20 lbs. were operated upon. The 
herbage was generally more stemmy and riper than that of 1862. The separations 
were, therefore, less difficult; but they were carried further, about five months being 
devoted to the work ; and the proportions of undetermined residue were less than in 
even the revised separations of the produce of 1862. 
In 1872 most of the herbage was finer than usual, larger samples were taken than 
previously, the amount of matter left undetermined was very small, and the time 
devoted to the separations was much longer than formerly. Thus, in all, 23 samples 
were operated upon: 14 of 15 lbs., and 9 of 20 lbs. each; and the period occu¬ 
pied in the analysis of the samples of that year was between 10 and 11 months. 
Nearly three months were afterwards expended in the revision of the results relating 
to the produce of 1862, as above referred to; when the previously undetermined residues 
were on the average reduced to qdbout one-third of the original amounts, and in many 
cases much lower. 
In 1877, 24 samples of 12 J lbs. each were taken. In the work of separation, besides 
some new hands, two, and sometimes three, who had taken part in the detailed sepa¬ 
rations of 1872, and in partial ones in 1874, 1875, and 1876, were engaged. The 
work was more completely done than on any previous occasion, no undetermined residue 
whatever being left; and the time occupied was not quite five months. 
Besides the complete separations into iu dividual species, in samples from every plot 
in the four seasons of 1862, 1867, 1872, and 1877, partial separations, as above referred 
to, that is into three main divisions of— 
Gramineous herbage, 
Leguminous herbage, 
Miscellaneous herbage, 
