well as of Sprengel in his edition of the St/stema Vege-
tabilium *.
Root fusiform. Stem three feet or more in height in favourable
situations, furrowed, smooth to the touch in the
lower part, roughish among the flowers. Leaves oblong-
lanceolate, wavy, smooth, except a little roughness on the
nerves and veins on the under side ; the base of the lower
ones rounded, approaching more or less to cordate, and
often oblique; root-leaves bluntish, the rest pointed.
Branches upright, forming a panicle of long clusters of
flowers. Whorls not crowded, a few of the lower ones
only subtended by a leaf. Flowers mostly perfect, but occasionally
intermixed, as in other c< L a p a th a with a few of
separate sexes. Calyx never reflexed. Inner valves of the
flower (“ petals ” ) whilst in flower oblong, obtuse, and
slightly toothed; when in seed, somewhat heart-shaped at
the base, dilated and toothed in the lower part, terminated
by a short triangular entire point; the outermost larger
than the others, more dilated and more deeply toothed, and
bearing an ovate tubercle, whilst in the others the midrib
is usually but slightly incrassated. Not rarely, however,
(both in this plant and in R. crispus,) the tubercles on the
three valves are nearly of equal size. Seed (or nut) acutely
triquetrous.
R. pratensis is in some respects intermediate between
R. crispus and K. obtusijolius, t- 1999, but roost allied to
the former, with which it nearly agrees in habit, and in
bein°- of rather stouter growth than R. sanguineus, t. 1533,
and R . acutus, t. 724, the R. Nemolapathum and R. conglo-
meratus of continental botanists. The leaves are somewhat
broader than those of R. crispus, and less curled ; the clusters
are less crowded, and the flower-valves unequal in size
and more distinctly toothed. In R. crispus, although not
entire, as they have been generally described, they are
rather crenulate, or erose, than toothed. In R. obtusifolius
the teeth are usually longer than in R. pratensis, and the
entire terminal part of the valves is mostly oblong or almost
lingulate.
Our plant is not found in Campdera’s Monograph of the
genus.—W. B.
* The Linnsean character of JR. acutus certainly agrees better with this
plant than with the R. acutus of Smith, on which no other author seems
to have found dentate valves. Linnaeus probably applied the name to
more than one species.