THE NATURAL HISTORY OF BRITISH GRASSES. 57 
to the farm in general by being allowed in watercourses, and 
no farm can be kept in good order without efforts are used 
to destroy them. 
The Poa aquatica and P. fluitans may now also be referred 
to as having been subjects of some interesting experiments 
in the garden of the Royal Agricultural College. Here, 
some two years since, we sowed a plot of each of these 
grasses which soon came up very well but did not flower 
until the second year, when, to our astonishment, the crops 
of the two beds were entirely alike in all their details of 
leaves and flowers ; the new plants however being so entirely 
distinct from either of the originals, and from every other 
known British grass, as to present a matter of interest for 
further investigation; and in order to this we have this year 
endeavoured to collect wild seed of the two forms in ques- 
tion, but as regards the Poa aquatica entirely without success, 
as though a wide district has been carefully examined, in no 
instance have we succeeded in finding any fertile seeds, 
Plants therefore of the two forms have been obtained from 
the Cirencester Canal and planted in our experimental plots ; 
and if it should be ultimately proved by repeated and varied 
experiments that a change of habitat from a residence in 
water to one on dry land is capable of producing so great 
an alteration in specific details, it will afford another inte- 
resting proof of the curious physiological facts which are to 
be elicited from experiments in the cultivation of grasses. 
Briza—panicle lax ; locuste of from six to eight very 
tumid florets; glumes equal, rounded, and with the 
whole flower quite smooth. 
Briza media—quaking grass—is so called from the rest- 
lessness of its pendulous flowers, which are comparatively 
heavy, and balanced on delicate rounded pedicels. Though 
a beautiful species, it is of no use agriculturally ; however, 
as it grows for the most part in poor, stiff, undrained clays, 
it may always serve by its presence or absence as an indica- 
tion of condition. If present in quantity, we may predicate 
