204 REPORT—1857. 
the road dirt with which the tops of our stone walls are frequently capped, 
having a decidedly creeping habit which, if shown as a tendency “in light 
sandy pastures near the sea” which is given by Hooker as the habitat of the 
F’. rubra, may account for the difference. 
As respects the varieties F’, ovina, tenuifolia and duriuseula, it may be 
remarked that poor uplands present the first, the bushes and hedgerows 
around these the second, and meadows examples of the latter; but seldom 
are they greatly intermixed, which, perhaps, may be taken as an argument 
that these forms are but varieties induced by different circumstances. From 
long observation and experiment I can only so consider them; and had Ia 
choice of names for the typical form, I] should choose that of duriuscula, as 
the departure seems to be from that type, of which F’, ovina is a mountain 
form, and F’. rubra a seaside or arenaceous one. 
b. Festuca loliacea varieties. 
Six years since, I sowed the seeds of the three forms as below, and in the 
following order. 
Festuca loliacea. Festuca pratensis. Festuca elatior. 
1. 2. 3. 
These plots the first year of flowering presented appearances as under :— 
Ist. Festuca loliacet.—Most of the plants of the true spicate type, but 
sparingly mixed with paniculate flowers: the herbage of which was of the 
rich green which characterizes FP’. loliacea. 
Qnd. Festuca pratensis.—All true, but with a tendency to a rigidity of 
leafage. 
3rd. Festuca elatior, scarcely distinguishable from (2). 
In three years great changes had been wrought as under :— 
Ist. F. loliacea.—No spicate flowers. 
Qnd. F. pratensis—More rigid and larger, in fact none of the true mea- 
dow type. 
3rd. F. elatior—A little larger, but otherwise not distinguishable from 
2). 
Tn the fifth year the F. elatior prevailed in all the beds. 
These plots are destroyed, as in 1855 the same experiments were recom- 
menced in another part of the garden, the plots, however, being placed at a 
distance apart ; and the present year they were plainly observed to be taking 
the same course as the others. 
Here then, I think it satisfactorily proved by experiment that these three 
forms are all of them referable to a single species, as the changes indicated 
have taken place in individuals; they, however, maintain their distinctive 
characters under the following circumstances. 
In meadows by the sides of rivers subject to occasional floods, as the Isis 
at Oxford, or irrigated meadows, as on the banks of the Churn at Cirencester, 
F. loliacea is constant in its characters, and is a most valuable grass for hay 
or pasture. 
In rich meadow flats, as in the Vale of Berkeley—the celebrated country 
of the “‘ double Gloucester cheese,’—the /’. pratensis is a common and yalu- 
able denizen, and any meadow where it maintains its character may be consi- 
dered as of good quality. 
— or 
