on the very steep N. side, where it comes lower down.” — In 1835, the Rev. 
Andrew Bloxam found it in the same locality, in abundance, and kindly com- 
municated a plant to me, which is now (January 25, 1839) growing in the Oxford 
Garden, and from which plant the drawing for the accompanying plate was made. 
Perennial. — Flowers in June and July. 
Root somewhat woody, tapering, with numerous fibres. Stem 
about a foot high, upright, round, reddish, striated, more or less 
hairy, leafy, branched in a corymbose manner, and many-flowered. 
Leaves pinnate in a lyrate manner, hairy but not hoary, deep 
green ; the radical ones largest, on long leafstalks ; their leaflets 7 ; 
those of the upper leaves 5, or 3, roundish, or somewhat inversely 
egg-shaped, veined, unequally cut and serrated. Stipulas of the 
root-leaves strap-shaped, pointed ; those of the stem-leaves romboid, 
roundish. Flowers in a forked corymbose panicle. Calyx downy 
at the base, enlarging after flowering, when it becomes brown and 
membranous. Petals white, roundish, or somewhat inversely heart- 
shaped, much larger than the calyx. Styles reddish. Seedy neither 
hairy nor wrinkled, on a very hairy or bristly receptacle. 
This plant is a native of several parts of Europe and Siberia, on shady alpine 
Tocks. It is of the very rarest occurrence in Britain, its onlj known locality 
being that iccorded above. 
The Natural Order Rosacea is composed of polypetalous, di- 
cotyledonous herbs or shrubs, with alternate leaves, which are either 
simple or compound, and which are almost universally furnished 
with 2 stipulas at their base. The calyx is 4- or 5-lobed, some- 
times having bracteolce on its tu' e equal in number to the lobes, 
and alternate with them (see figs. 1, 2, & 4.), valvate or imbricate 
in the bud, with the disk surrounding the orifice, having the fifth 
or odd lobe next the axis. The corolla is 4- or 5-petaled ; the 
petals perigynous and equal, with short claws. The stamens, which 
are indefinite, arise from the calyx, just within the petals, and are 
curved inwards in sestivation. The filaments are free; the anthers 
innate, 2-celled, and burst lengthwise. The ovaries (see fig. 4.) 
are several, superior, mostly free, rarely cohering either with the 
calyx or among themselves, 1-celled and 1-seeded. The ovule is 
usually suspended, seldom erect. The styles (see figs. 5 & 6.) are 
lateral, near the apex of the ovaries, with simple stigmas, emargi- 
nate on one side. The fruit is either 1-seeded nuts or akenia. 
The seeds are pendent, rarely ascending. The embryo is straight, 
with a taper short radicle, pointing towards the hilum ; the cotyle- 
dons flat and entire ; without albumen. 
Mr. Don divides this order into three tribes, r.amely, 1. Drya'de.e ; 2. Nku- 
ra'de/e ; and 3. Ro'se/e. Two only of these tribes are applicable to the plants 
of Britain. 
The first, or Drya'dm, contains Dry'as, t. 248 Geum, t. 3. — Rubus. — 
Fragrlria, t. 242. — PotentiUa, t. 3 13. — Tormentilla. — Comarum, t. 197. — 
Sibaldia.— and Agrimouia, t. 88. — The third, or Rosea, has ouly the genus 
Rosa. 
Spiree'a, 1. 133, belongs to the Natural Order Spin aa'cea of De Canpoi.le. 
The plants of this order are distinguished from those of Rosacea by their de- 
hiscent carpels, and by their styles being terminal, not lateral, as in that order. 
Alcftemilla, t. 280 ; and Sanguisorba, t. 269, belong to Dr. Lindley’s 
Natural Order Sanquiso , rdea. This order differs from RosA'cs/tin the plants 
which compose it having apetalous flowers, with an indurated calyx, and only 
one nut or carpel. 
