pended: seeds 1 -2 ; stems very short, ccespitose, woody, imbricated with
spatulate entire exstipulate leaves: scapes, or Jlgwering stems, sparsely
leafy : spike dense, cylindrical.—P etrophytum, Nutt. mss.
13. S. ceespitosa (Nutt.! mss.): “ low, shrubby ; leaves rosulate on the
very short tufted branches, small, spatulate-oblong,- entire, silky-villous;
those of the scape scattered and much smaller ; flowers white, in a dense
spike ; calyx silky-villous; the segments ovate, acute.
“ On high shelving rocks in the Rocky Mountains, towards the sources of
the Platte. July.—A singular dwarf alpine plant, with scarcely the habit of
Spiraea. Leaves, about one-third of an inch long. Scapes 3-5 inches high ;
the cylindrical spikes an inch or two in length, obtuse. Bracts as long as the
flowers. Carpels linear, a little curved at the apex, dehiscent along the inside
and 2-cleft. The taste of the plant scarcely perceptible.” Nuttall.
S. hypericifolia (Linn.) was erroneously said by Linnaeus to come from North
America. The plant so called by Muhlenberg in his own herbarium is a variety
of S. salicifolia, although we believe a specimen of the true plant, sent by Muhlenberg,
exists in the herbarium of Willdenow, probably a cultivated plant. Pursh
has committed some error in stating S. hypericifolia to grow in dry swamps of
Canada and New York.
S. crenata (Linn.) must also be excluded from our flora.
S. sorbifolia (Linn.) is doubtless wrongly given by Pursh as a native of the
North West Coast.
6. GILLENIA. Mwnch, sufpl. p. 286 ; Nutt. gen. 1. p. 307.
Calyx tubular-campanulate, the orifice somewhat contracted, 5-toothed;
the teeth glandular-ciliate, erect. Petals 5, linear-lanceolate, very long,
somewhat unequal, inserted in the orifice of the calyx. Stamens 10-15,
mostly included, 5 of them sometimes shorter. Carpels 5, distinct or at first
connate, follicular or 2-valved: styles filiform, terminal: stigmas subcapi-
tate. Seeds 2-4 from near the base of each carpel, ascending, oval; the
testa rather erustaceous.—Perennial herbs, with trifoliolate stipulate sub-
sessile leaves : leaflets membranaceous, doubly serrate and incised, petiolu-
late. Flowers (rose-color) axillary and terminal, paniculate-corymbose:
peduncles and pedicels elongated.—Roots emetic and cathartic, also said to
be tonic.—Indian-physic. Bowman’s Root.
1. G. trifoliata (Mcench, 1. c .) : stipules linear-setaceous, entire ; leaflets
ovate-oblong, acuminate.—Nutt.! gen. 1. c .; DC.! prodr. 2. p. 546; Dar-
lingt. Jl. Cest.p. 300. Spiraea trifoliata, Linn. !. spec. 1. p. 489 ; Michx. !
fi. 1. p. 294 ; Bot. mag. t. 489 ; W illd.! spec. 2. p. 1063; Mill. ic. t.
252; Bigel. med. bot. 3. p. 11, t. 4 1 Bart. veg. mat. med. 1 .1. 5.
Canada and Western part of NewYork! to the upper districts of Georgia!
in shady woods : scarcely found west of the Alleghany Mountains. (South
Western part of Missouri, Dr. Engelmann.) June-July.—Stem 2-3 feet
high. Flowers large, nearly white. The seeds are intensely bitter to the
taste.
2. G. stipulacea (Nutt.): stipules ovate, foliaceous, doubly incised;,
leaves lanceolate, deeply incised.—Nutt.,.' 1. c .; E ll.! sk. 1. p. 562; DC.!
l. c. Spiraea stipulata, Willd.! enum. 1. p. 542. S. stipulacea, Pursh, Jl.
l.p . 343; Bart. 1. c. 1. t. 6 ; Camb. in ann. sci. nat. 1. p. 387, t. 28.
Western States ! to Arkansas! and Louisiana! Western part of Pennsylvania
! and NewYork (not east of the Alleghany Mountains), and in the
mountains of the Southern States to Alabama ! June.—Radicle and lower
leaves almost pinnatifid. Peduncles few-flowered.
T ribe II. DRYADEiE.
Potentillece & Sanguisorbem, Juss.
Calyx valvate or often imbricate in aestivation. Stamens sometimes
definite. Carpels (achenia) 1-seeded and indehiscent, either
few or solitary, or numerous and then collected into a head on a
conical or hemispherical torus : ovules solitary, rarely 2, suspended
or ascending (sometimes amphitropous): styles lateral or terminal.—
Herbs, shrubs, or rarely trees.
The suborder or tribe Sanguisorbese of Jussieu, De Candolle, Lindley, &c.
appears to comprise genera that are not very nearly related, and is circumscribed
by no constant character. The indurated calyx-tube contracted at the orifice is
perhaps the best character, but this does not well apply to Alchemilla, Adenos-
toma, or Cercocarpus. Poterium and Cliffortia only have unisexual flowers.
Alchemilla has sometimes four ovaries, and Sibbaldia, Chamserhodos, Wald-
steinia, and Dalibarda have often no greater number. If the few and definite
stamens be assumed as the leading character of the tribe, Poterium and Cliffortia
would be excluded, while Horkelia, Sibbaldia, and Chamserhodos would be included.
If the absence of petals be deemed the most important character, we
must include Cercocarpus and exclude Purshia, while Agrimonia, Aremonia, and
Adenostoma (in which the calyx-tube is indurated, and the ovaries one or two,)
will be rejected from the tribe. The division we propose rests upon more important
and constant, although less obvious, characters, and appears to produce
a more natural arrangement.—It appears to us that the genus Cliffortia is incorrectly
described. In the few species that we have examined, the seed is certainly
suspended, and the radicle superior.
Series I. Seed attached to the extremity o f the cavity o f the ovary
farthest from the insertion o f the style. ■ Radicle inferior.
Subtribe 1. E udryade®.—Calyx campanulate or turbinate, or rather
flat, valvate in aestivation. Stamens numerous. Carpels numerous, dry:
style terminal. Seed erect. Radicle inferior.—Herbs or sometimes shrubby
plants. Flowers perfect.*
7. DRYAS. L in n .; Lam. ill. t. 443; Gcertn. fr . t. 74.
Calyx concave at the base, 8-9-parted ; the segments nearly equal and in
a single series. Petals 8-9, large. Stamens numerous. Achenia numerous,
aggregated on the dry receptacle, caudate with the very long persistent
* Coluria, R. Br. and Cowania, Don., also belong to this section.