160 Brenchley. — The Weeds of Arable Land. III. 
This hypothesis is upheld by the behaviour of two species of Polygonum , 
P, aviculare and P. Convolvulus. These plants were exceedingly rare among 
seeds in Norfolk and Bedfordshire during two seasons in which plenty of rain 
had fallen, so that the crop had made good growth, rendering competition 
keen. In the West Country both these Polygonums were found with seeds, 
but this was in a very dry season in which a long period of drought 
immediately succeeded the sowing of the crop. As a result little growth 
was made, thus enabling these sensitive weeds to get their footing in an 
unusual habitat. 
A very few plants seem to be even more exclusive in their association, 
the following species being found only with cereals : 
Lapsana communis Plantago media 
Lolium perenne Poa trivialis 
The indications are that many of the weeds will ultimately prove 
to have a very decided preference for or abhorrence of some particular class 
of crop, but that these individual peculiarities are liable to be modified by 
local conditions, although some plants may prove strong enough to master 
the local variations, so remaining consistent in their crop association. 
The question arises as to the effect of impurities in the seeds sown upon 
the weed flora. Broadly speaking, the effect is not well marked. Of all the 
impurities indicated by Borlase 1 as occurring in Leguminous seeds, Geranium 
dissectum is the only one which is confined to such crops during growth. 
Poa annua , which is introduced with the crop seed, is conspicuous by 
its absence in the grown crop. At present it does not seem possible to 
draw any conclusion with regard to this point. In the cereals so few weed 
seeds are sown that their effect is practically negligible. 
Points of Special Interest. 
i. The different orders of flowering plants vary very much in the 
contribution they make to the weed flora of the arable lands. Many 
of them provide just one or two species, but when the large orders are con- 
sidered the differences in representation are very striking. Leguminosae, 
Rosaceae, Umbelliferae, and Chenopodiaceae are very poorly represented, 
both relatively in proportion to the size of the orders, and actually in the 
number of species occurring as weeds, Lamiaceae being very little better. 
Ranunculaceae, Papaveraceae, Geraniaceae, and Euphorbiaceae each pro- 
vide one main genus, but in each case several members of that genus 
are commonly found. 
See p. 156. 
