ROSACEA. 



367 



tliree. From Geuni, however, we cannot separate Colnria, because 

 of its articulated style, nor Waldsteinia on account of the reduced 

 number of its carpels, since we do not distinguish llorkelia or Sihhaldia 

 and Ivesia respectively, presenting as they do these characters, from 



Fig. 433. 

 Flower. 



Oeum Waldsteinia. 



Fig. 434. 

 Longitudinal section of flower. 



Potenfilhi. In this genus Gcirm we must also retain Siylipus vernus,^ 

 a small-flowered North American plant, in which the calycle dis- 

 appears entirely, or is only represented by five very minute glands 

 alternating with the sepals. The same thing occurs in some of the 

 known species'^ of the section Waldsteinia. Thus marked out, the 

 genus Geuni includes about thirty-five species,'* from the cold and 

 temperate regions of all parts of the world, but more abundant in 

 the northern hemisphere. 



The genus Brijas" (fig. 435), which has often given the group now 

 under consideration the tribal name oi Dryadea,^ has altogether the 

 sexual oro-ans" and fruit of the section Sieversia of Gei/ni ; but the 



or three branches, as many as there are carpels. 

 P]ach ovary is articulated with the top of this 

 branch and the style Itself is articulated at its 

 base, as in Coluria. The ovule is that of Geum 

 proper. 



1 Rafin., Neogen. (1825), 3, ex Tore. & Gk., 

 Fl. N. Amer., i. 422. — Geum vernum ToER. & 

 Ge., loc. cit. 



2 DC, loc. cil— ToEE. & Ge., Fl. N. Amer., 

 i. 426.— A. Geay, Man. of Bot., ed. v. 153.— 

 CnAPM., Fl. S. Unit.- States, 123. — Walp., 

 Rep., ii. 46.— Hook., JcoM., t. 76; Bot. Mag., 

 t. 1567, 2595. 



3 DC, Prodr., ii. 550, 555.— Geen. & Goim., 

 Fl.de Fr.,i.ZlO.—Boiss., Voy. ^.s/>.,t.58.— TouE. 

 & Ge., Fl. N. Amer., i. 420.— A. Geay, Man. <if 

 Bot., ed. V. 151; VI. Fendler., 40.— ChapiM., 

 Fl. S. U)iU.- States, 123.— C Gay, Fl. Chil., ii. 

 276. — Wedd., Chlor. And., ii. 235.— HooK. F., 

 Fl. Antarrf., ii. 262; Handh. of N.-Zcal. FL, 



55. — Benth., Fl. Austral., ii. 427. — Haet., 

 Thes. Cap., t. 18.— Hakv. & Sond., Fl. Cap., 

 ii. 289.~WALr., i2ep., ii. 46; v. 656 ; Ann., i. 

 972.— Bot. Reg., t. 1088, 1318. 



■» L., Geu., n. 637.— .1., (?eH., 338.— Gj-etn., 

 Fnwt., i. 352, t. 74. — Lamk., Diet., ii. 329 ; 

 SuppL, li. 525; III, t. 443.— Nestl., Pot., 

 16.— DC, Prod)-., ii. 549. — Spach, Suit, a 

 Biijfon, i. 477.— Endl., Ge«., n. 6389.— B. I!., 

 Gen., 618, n. 42. — Chamcedrgs Clus., Hi^t., 

 ii. 351, ex Adans., Fam. des PL, ii. 295. 



^ Ventl., Tail., iii. 319. — Endl., op. cit., 

 1241. — Fudryadeoi ToEU. & Ge., FL N. Amer., 

 i. 426. 



" D. octopetala L. (Spec, 717) has very 

 numerous stamens, as in most species of Geum. 

 The tihimcnts, inllexcd in the bud, are inserted 

 on tlie margin of the glandular coloured disk 

 lining the receptacle. I'he carjiels are very nume- 

 rous, and like those of Sii'rersia (tig. 435). Tlie 



