422 
The Herbage of Old Gi'ass Lands. 
of the general appearance of each plot. It was specially desired 
that each turf should be cut from, and should be typical of, 
what practical men regard as the best grass-land of each district. 
In no case did I select the turf myself, preferring always to be 
guided by those whom long experience had taught where to 
select the best old herbage in their respective districts. 
In all, 25 turfs were received and planted, 17 from England 
and 8 from Ireland. The English specimens comprised 3 from 
Somerset, 2 each from Dorset, Wilts, and Kent, and 1 each 
from Lincolnshire, Leicestershire, Derbyshire, Staffordshire, 
Buckinghamshire, Gloucestershire, Herefordshire, and Devon- 
shire. The Irish specimens were 1 each from the counties of 
Tipperaiy, Kildare, Mayo, Meath, Cork, Wexford, Dublin, and 
Clare. With one or two exceptions, which are noted, all these 
turfs were taken from old pastures and not from meadows. 
In the month of July the herbage upon each turf was cut, 
but they were not all cut simultaneously, for a reason which 
will at once be made apparent. Directly a turf was mown the 
herbage was, without loss of time, separated into what I may 
call its proximate botanical constituents of gramineous herbage, 
leguminous herbage, and miscellaneous herbage, and each sec- 
tion was at once divided into its specific elements. This work 
is of a tedious and laborious character, particularly the separa- 
tion of the gramineous herbage into its component grasses ; 
but it is the method which has been followed in the Rothamsted 
experiments, and no better system is available. As soon as 
the botanical analysis or separation was completed the various 
sections were weighed on a balance, with metric weights, and 
the results recorded. The separation was performed in the 
shade, and the green weights were taken, as it is in the green 
state that the herbage of pastures is consumed. 
My note-book records (1) the general appearance of each 
plot from time to time, (2) a note of every species of plant 
observed growing upon each plot, (3) the total weight in gi-ams 
of the cut herbage from each turf, (1) the actual weight — when 
it exceeds one gram — of each species of plant found in the 
herbage of each plot, (5) the calculated percentage by weight of 
each species when present in a significant proportion. 
It is impossible to burden this paper with aU these details. 
Accordingly, in the following brief summary of the results 
afforded by each turf, what has been attempted is to give (a) a 
short account of the origin and history of each specimen, (!>) the 
respective percentages by weight of its gramineous, leguminous, 
and miscellaneous herbage, (c) the proportionate weight per 
cent, in which the various species of gi'asses are contained in the 
