EANUNGULACE^. 



49 



the order is that each carpel originally contains two ovules, of which 

 one alone attains its full development, and appears suspended beside 

 its aborted fellow, with the raphe internal and the micropyle turned 

 upwards and outwards/ CaUlanthemum has, therefore, but one seed 

 in each achene. The plants are herbaceous perennials, with alternate 

 compound leaves and terminal flowers. As yet but two species are 

 known; one European- and the other Asiatic' 



The Canadian plant Hi/drastis,' which we refer with some doubt 

 to this group,' has regular flowers, usually" hermaphrodite; the 

 perianth, simple and very caducous,'' consists of only three petaloid 

 leaves. Above this the receptacle, of the same form as 

 in Ranmiculus, bears numerous stamens and then carpels. 

 The stamens are free, each consisting of a filament 

 dilated above, and a basifixed anther with two cells de- 

 hiscing by nearly lateral clefts.' Each carpel is composed 

 of a unilocular ovary, tapering above into a style, whose 

 apex is dilated into two lateral papillose and fringed lips. 

 Half-way up the inner angle of the ovary (fig. 88) the 

 placenta forms two vertical projections, each supporting 

 an ovule. These ovules are at first horizontal and placed 

 back to back ; but as they grow, one becomes ascending 

 with the micropyle usually downwards and outwards, Hydrastis cmia- 

 the other descending with the micropyle upwards and y"^'L 

 inwards. The fruit consists of a variable number of Carpei. 

 berries united into a head, and buried in the mass we Longitudinal 

 find seeds with thick coats containing a small embryo 

 at the apex of abundant fleshy albumen. This plant, the only one 



' We explained for the first time in our 

 Memoire stir la Famille des Renonculacces 

 [Adansonia, iv. 23) how in C. rutafoHum there 

 are first of all two ascending ovules ; then how 

 the one of them, which has become superior, 

 compresses the other and forces it down, in- 

 creasing gradually at its chalazal end, so that its 

 micropyle remains above and turned outwards. 

 Figs. 86 and 87 represent two phases of the evo- 

 lution of these ovules. That which becomes fertile 

 has two coats. In C. acaule the carpels are 

 stipitate, and the stigniatic papillai arc borne on 

 the summit of the ovary. In C ridrefolium the 

 ovary tapei's above into a style papillose at the 

 summit. 



- Reicht?., Icon., iii. 25. — Gren. & Godr., 

 Fl. Fr., i. 17. 



* Camr., in .lAOQtTEM., Fb//., 5, t. 3. — Don., 



VOL. 1. 



in RoTL., Hhnal., iii. 45, 53.— HoOK. & Tii., 

 Fl. Ind., i. 26. 



* Hydrastis canadensis L., Spec, 781. — J., 

 Gen., 232.— Micux., Am. Bor., i. 317.— DC, 

 Prodr., i. 23.— Spacit, Suit, a Buff., vii. 3S3.— 

 Endl., Gen., n. 4777.— Hook., Bot. Mag., t. 

 3019, 3232.— A. Gray, Gen. III., t. 18.— 5. H., 

 Gen., 7, n. 16. — H. Hn., Adansonia, iv. 25, 

 53. — Warneria canadensis Mill., Icon., ii. 190, 

 t. 285. 



* Its habit and flowers bring it somewhat near 

 Actrea. Most authors make it a Hellebore. 



•• Sometimes there are llowors without ii gyniB- 

 ceum. 



' Hence we are ignorant of its rcstivation, which 

 can hardly be observed except on a living plant. 



'^ Nearer, however, the internal tlian tho ex- 

 ternal fitce. 



E 



