on Permanent Meadoio Land. 
247 
varying- decree in wliicli they rospectlvely developed the different 
kinds of plants. 
Allusion has already frequently been made, in a cursory way, 
to the greater development of the Lef/nminoiis herbajfe by purely 
mineral manure!;, and to that of the Graminaceovs plants, or 
natural grasses commonlj' so-called, by characteristically vitro- 
genous manures. In fact, the plots had each so distinctive a 
character in regard to the prevalence of different plants, that the 
experimental ground looked almost as much as if it were devoted 
to trials with different seeds as with different manures. So 
striking and characteristic, indeed, were the effects produced 
in this respect, tliat, in 1857 and 1858, the subject was thought 
of sufficient interest to induce us to request the examination 
of the plots by Professor Henfrey, to which he kindly assented. 
An endeavour was also made in the second year, 1857, to sepa- 
rate, and determine, the proportion of the different plants in care- 
fully averaged and weighed samples, taken from the several plots 
as soon as the grass was cut. Taking advantage of the expe- 
rience gained in this first trial, the separations have been carried 
out more carefully in the case of the produce on some of the 
most important plots in the tliird season, 1858. The results of 
these separations are recorded in detail in Table IX., p. 250, 
and in a summary form in Table X., p. 252 ; and it is the con- 
sideration of those results that will constitute the subject of this 
Third Part of our Report. 
The mode of proceeding in making the separations and esti- 
mations may be shortly explained. As soon as the grass on a 
plot was cut down, samples were taken from many parts of it. 
These were carefully intermixed in such manner as to shake out 
as little seed as possible ; and then, from the whole, a certain 
quantity was weighed out to be further operated upon. Charac- 
teristic specimens of each of the plants in Jlower or seed, or in 
other conditions in which they could be recognised, were then 
selected as types ; and a number of boys were set to pick from 
the weighed sample, all they could find to correspond with these 
types. Tlie remainder consisted chiefly of detaclied foliage, and 
undeveloped stems, which was then separated into four or five 
different lots, according to types selected to the best of our 
judgment. Each weighed sample was thus divided into from 
filteen to twenty different descriptions of herbage. The weight 
of each of the selected portions was afterwards taken — all in an 
equal condition of dryness. The weights so obtained, of the 
respective grasses, or other plants, or parts of plants, in the ori- • 
ginal weighed sample from the plot, were then calculated into their 
percentage relation to tlie collective weight of the whole of the 
separated portions in their partially dried state. It is the results 
