TRËMANBRACEJi. 69 



distinctly arvauged at all ages in one verticel, are grouped in paii-s 

 enveloped by each petal, its edges being folded inwards. The non- 

 articulate anthers have two cells or four cellules arranged in two rows. 

 The gynseceum, fruit, and arillate seeds are the same as in Tremandra ; 

 they have as many as four ovules^ in each cell and are generally 

 destitute of disk. Some twenty species have been described,^ 

 glabrous or glandular with alternate opposite or verticillate leaves. 



This small group ^ has often beeu considered as representing the 

 regular form of the Pohjgalaccœ * ; and this was the opinion of E. 

 Brown. Others connect it rather with LasiopetaJfe^ which the true 

 Trcmandras really resemble much in aspect and foliage and by their 

 stellate haii's, but are very clearly distinguished from them by the 

 proefloration of the corolla and the organisation of the androceum and 

 gynaîceum. There is also believed to be an affinity between the 

 Trcmandras and the Chciranfhcras ■' of the group Pittosporece. The 

 TremandracccB seem to us placed between the Pol/jgalacece on one 

 side, having the same gynteceum and nearly the same ancboceum and 

 from which they are separated by the regularity of the flowers ; and 

 the Linaceœ on the other, having their regular corolla, diplostemous 

 androceum, capsular fruit, the same dii-ection of the regions of the 

 ovule, and from which they are separated by their mode of proeflora- 

 tion, the fewer number of ovary cells, and the difference in the 

 consistence of the albumen. All the Tremandraceœ described, 

 twenty in number, are extra-tropical Australian ; they have no 

 known property. Platytheca veriicillata (fig. 98-103) and several 

 Tetrathecas are valued in oui- green-houses as pretty ornamental plants. 



' In the speciea from the south-west of iv. 241 ; vii. 241. 



Australia ; those from the east have hut one or ' Tremandi-eœ E. Br. Gen. Scm. (1814), 544 ; 



two. Beside the prolongation of the region of 3Iisc. Word's, ed. Benn. i. 15. — Endl. Gen. 



the chalaza, the ovule presents a slight thicken- 1076, Ord. 232.-1)0. Prodi: i. 34.3, Ord. 19.— 



ing of the exostomc often capped by a small B. H. Oeii. 133, Ord. 232. — Tremandiaccœ 



obturator (as in the FAiplinrbiac'œ). The ovules Lindl. Vcg. KUh,d. 384, Ord. 132. 



may be nearly collateral, three in number. ■* It constitutes "with Poliitjalaceœ, the class 



The seeds are covered with hairs in the Eastern Pulyijalineœ of Endlicuer. 



species. ^ They are doubtless more apparent than real, 



'^ Labill. pi. Nouv.SoU. i. 95, t. 122, 123. — the organisation of the gynseceum being quite 



Reichb. Ic. Exot. t. 78. — Eudg. in Trans. Linn. different, and depending upon the analogy of the 



Soc. viii. t. 11. — Endl. in Hneg. Enum. 7. — fonn and colour of the perianth . The androceum 



Hook. Icon. t. 268. — Hook. f. Fl. Tasm. t. 7. — of the Cheirantheras is also very different from 



Steetz, in PL Pniss. i. 212. — Benth. Fl. that of the Tremnndracece. According to 



Austral, i. 11^. — Lindl. ra Mitch, thr. Exp. i\. Agardh [Theor. Si/st. 190), these are "more 



206; Sw. Riv. App. 38; m Bot. Reg. (1844), t. yavieoi Bertyacecc {Eiiphorbiaceœ)." 

 67.— Walp. Rep. i. 249 ; v. 68 ; Ann. ii. 87 ; 



