Sanffuisorha.] ROSACEiE. 



197 



Sp. PL p. 643. Mich. Am. v. 1. p. 287. Pursh, FL Am. v. 1. p. 325. Eyigl. BoL t. 1335. 

 BigeL FL Bost ed. 2. p. 189. EllioU, Carol, v. I. p. 526. Torrey, FL of Un. St. v. I. p. 

 473. De Cand. Prodr. v. 2. p. 587. — /?. hirsuta. Torre?/. — 5. parviflora; racemis gracilibus, 

 pedicellis longioribus, florlbus remotior ib us nuiioribus. — A. parviflora. Ait HorU Kew. ed. 

 1. V. 2. p. 130 ? et aliorum ? — A. striata. Mich. Am. v. 1. p. 287 ? et aliorum ? 



Had. Throughout Canada, as far as Lake Winipog;. Dr. RicJinrdson^ DntmmoJid. /3. and y. Laho 

 Huron. Dr. Todd. — What I take to he a small flowered variety of this is not uncommon in Upper Caxiada, 

 and this, I strongly suspect, is the A. parviflora of Ait., and probably the A. striata, imperfectly deticrihed 

 by Michaux, and which he says inhabits Canada; but he observes that the flowers are white. According to 

 ray specimens from Mr. Elliott, his A. maveolens is this var., but whether it be the same with that of Pursli, 

 I am not able to determine. 



Trie. IV. Sanguisorbeje. Juss. 



18. ALCHEMILLA. Touni. 



Cal^ tubulosus, tubo apice subcontracto, limbo 8-partIto, lobls altcrnis (bracteolis) min- 

 oribus interdum minimis dentiformibus. Pet. 0. Stam. 1-4}. Carpdla 1-2 lateriiliter 

 stylum filiformem apice capitatum gerentia, demum sicca indm'ata monosperma. Semen 

 inversum. — Herbse, foliis palmaiim lobatis aut sectis, floribus corymbosis parvis. Genus 

 medium, inter Sanguisorbeas et Dryadeas. DC, 



1. A. vulgaris; foliis reniformibus plicato-concavis 9-loI)is serratis, caule petiolisque gla- 

 briusculis, floribus dichotomis corjmbosis. DC. — Linn. Sp. PL p. 178. Fngl. Bot. t. 597. 

 De Cand. Prodr. v. 2. p. 589. 



Hab. Labrador. Dr. Morrison, Rev. B. Kohlmeistcr. — This species is known to inhabit Greenland, but, 



1 beheve, no other part of the continent of America than that 1 have now mentioned. A. uljnna, too, is 

 found in Greenland ; but in America it is only doubtfully mentioned by Pursh as having been seen on high 

 mountains in Vermont, and New Hampshire. 



19. SANGUISORBA. Linn. 



Flores hermaphroditi. Cal. 4-fidus extus basi bisquamosus. Pet. 0. Stam. 4. Carpella 



2 intra calycis tubum inclusa, stylo apice penicillifoniii superata, in achenia sicca indehisc. 



l-sperma conversa. Semen inversum. — Herbse perenneSy foliis impari-pinnatisy floribus ?w 

 spicas densissimas ovatas cylindraceasve aggregatis. DC. 



1, S. media; spicis elliptico-cylindraceis (atropurpureis) staminibus calyce subduplo 

 longioribus, caljcibus bracteis foliisque longioribus, foIioL's corduto-ovutlii obtusis grosse 

 dentato-serratis. — LiJin. Sp. PL p. 594. Pur shy FL Am. i\ 1. p. 117. Torreyy Ft. of Un. 

 St v.l.p. 176. Sm. EngL FL v. I. p. 219? De Cand. Prodr. v. 2. p. 594.—^. minor. 



Hab. Canada. {Linn<mis,) Douglas, MSS. Columbia River, on the North-West coast of America, Mr. 

 Menzies, Dr. Scouler. /3. At Port Wentworth. Mr. Menzies.—Thh is assuredly very nearly allied to our 

 Europiean S. officinalis; and was described by Linnaeus as a native of Canada, (from which country, how- 

 ever, I have never seen a specimen,) and dLitinguished by its longer cylindrical spikes, and lonjjor stamens. 

 In these characters, our plants from the North-West coast entirely agree; and I have a beautiful drawing, 

 made on the spot by Mr. Menzies, in which these distinctions are faithfully represented. The spike is a very 

 deep purple colour. At Fort Vancouver, on the Colnmbifi, the leaves attain a length of twelve or fourteen 

 inches, and the stem a heijrht of two feet. At Port Wentworth, they scarcely reach the height of a span.— 

 The style is single, according to Mr. Menzies's figure; the stigma a globose tuft of hairs; and there are thice 

 bracteas at the base of the calyx. 



