On the Plant Communities of Farm Land. 
BY 
R. G. STAPLEDON, M.A., 
Botanist , Agr. Dept., University of Wales , Aberystwyth. 
Introduction. 
I T is proposed in the present paper to give some account of observations 
made over a period of eight years on the weeds found under various 
•farm crops, including land put down to grass for various short periods and 
fields that have ‘ run down ’ to more or less permanent grass without the 
addition of seeds, or with a very inadequate sowing. Most of the work 
was carried out on the Cotswolds (36 o'- 7 oo ') and in Mid-Wales (350'- 1 300') ; 
a few observations have also been made near Holsworthy, Devonshire, and 
near Greenhithe, Kent. 
Method of Study. Endeavour has been made to study the weed 
flora in terms of whole communities and not as isolated species ; and to 
supplement careful observations by a more accurate method of procedure. 
Two statistical methods have been employed throughout the whole period of 
the work. 
(a) Specific Frequency. This method is used on arable land and 
occasionally on grass-land. It is practically identical to that employed 
by Rhaunkiaer, which has been explained by Smith (8) in his review of 
that authors ‘ Life Forms and Statistical Methods \ The method has been 
described by the present writer elsewhere (10 and 11 ) ; its essence is to 
record the species found on a large number of small unit-areas, without 
paying any attention to the abundance of individual species per each unit- 
area. A mesh 6 " x 6" is used and dropped at chance on the ground ; 
the names of all the species occurring in each reading are carefully noted. 
The area to be examined is traversed across two diagonals and from 50 
to 200 readings taken. The results given are recorded against each 
species in terms of the number of its occurrences per 100 readings. 
For purposes of publication it is, however, found more convenient to 
tabulate the frequencies on a scale of 10 ; species with frequencies below 1 
are marked o, if occasional ; s, if scarce or solitary ; and r, if of quite 
exceptional occurrence. 
(b) Percentage Frequency. This method, originally devised by 
Armstrong ( 1 ), is advantageously employed upon grass-land. It has been 
explained elsewhere (11). Small unit-areas (turfs 6" x 6") are removed 
(Annals of Botany, Vol. XXX. No. CXVII. January, 1916.] 
M 
