UxiNUNCULACEJE. 



33 



filameut expanding above into an erect basifixed connective, 

 which supports the two adnate vertical cells of an extrorse anther 



M. Lingua. 

 Fig. 60. 

 Dias'ram. 



Ranunculus repens. 



Fig. 61. 

 Flower. 



Fig. 62. 

 Lonsritudinal section of flower. 



dehiscing by two longitudinal clefts.' The carpels are each com- 

 posed of a transversely compressed ovary tapering into a beaked 

 style, recurved outwards. Along the whole length of the inner 

 angle runs a vertical groove, whose margins, thickened and somewhat 

 everted, are covered above with stigmatic papillae. In the inner 

 angle, at a variable distance from the base of the single-celled ovary, 

 is inserted an ascending ovule, whose micropyle looks outwards and 

 downwards." After flowering, the perianth and androceum usuall}- 

 fall and discover a multiple fruit, formed of a variable number of 

 aclienes, each of which encloses a seed containing a minute embryo 

 towards the apex of abundant fleshy albumen. The surface of the 

 acbene is sometimes smooth and sometimes covered with ribs, 

 wrinkles, or even well-developed prickles, as occurs in B. arvensis 

 (figs. 63, 64), muricatus, Plii/onotis, and a certain number of allied 

 species.^ The form and height of the beak or jjersistent style which 



^ The lines of dehiscence are very decidedly 

 exterior in R. Seguieri, and the anther is cer- 

 tainly extrorse, though less markedly so, in R. 

 Lingua, Flammula, acris, arvensis, angidaius, 

 gramineus, parvijlorus, &e. ; the dehiscence is 

 exactly lateral in S. plaiunifoUus and aconiti- 

 folius. R. sceleratus and aquafilis are interme- 

 diate between these two groups, their dehiscence 

 being but slightly extrorse ; but in no case is it 

 introrse. 



" The ovule is always inserted into the inner 

 angle of the carpel, near its organic base. Hence 

 the ovule becomes horizontal, or even slightly 

 drooping, whenever the ovary is much developed 

 in its dorsal and posterior part. Here, as every- 

 where else, a drooping ovule witli the raplie 

 dorsal corresponds to an ascending ovule with 

 the raphe ventral. Moreover, as Bentitam & 

 Hooker (G&n., 6) remark, on the subject of 



VOL. T. 



CyrtorhyncJia Nuttall (ToiiK. & Ge., FI. X. 

 Am., i. 26. — Endl., Gen., n. 4771), which they 

 refer to the genus Ranunndus, an ovule which 

 is absolutely drooping in space, is really ascend- 

 ing in relation to a carpel whicli, pressed on by 

 its neighbours, has its tip turned first outw.irds 

 and then downwards. 



2 In grouping tlie genus R inuncuhis into 

 sections, some use has been made of the nature 

 of the surf\ice of the carpels. Thus Ue Candollk 

 distinguishes Ranuncidastnim (sect. ii. Prodi:, i. 

 27), Tliora (sect. iii. 30), and Ilecafonia (sect. iv. 

 30), by their smooth carpels, while liis Batrachium 

 (sect. i. 26), admitted as a distinct genus l)y 

 several authors (Spacu, Stdt. a Ruff'., vii. IDit), 

 has the pericarp transversely striated and rugose, 

 and \\\s Fchiiii'lla (sect. v. II. — Gfii. l\ichgh>ma 

 SrACn, Suit, a Ruff., vii. \dU—Phi)onolis 

 llEicnn., Coiisp., l',)l) has the car]iels covorcd 



I) 



