98 NATURAL HISTORY OF PLANTS. 



which causes them soon to become exserted. In others, as S. 

 speciosa, the tube is also very long, but the staminal filaments are so 

 short that the anthers remain enclosed. Thus the former species 

 are united gradually to S. acuta, rupestris, etc., of which a genus 

 Br achy siphon has been made and of which, the stamens being also 

 enclosed, the tube of the perianth remains relatively shorter. 



Endonema, formerly classed with Sarcocolla, is distinguished by 

 each of the ovarian cells enclosing two pairs of, instead of two, 

 ovules. Those of the upper pair are ascending with inferior mi- 

 cropyle ; those of the lower pair, descending, with superior micro- 

 pyle. All have a dorsal raphe. Sometimes the lobes of the 

 perianth are valvate, as in the true Endonema, and sometimes they 

 are reduplicate, as in the section Glyschrocolia. The stamens are 

 shorter or longer than the perianth. Endonema is from the same 

 country as Sarcocolla whose organs of vegetation it has. 



This small family, according to Lindley l was verbally established 

 by him in 1820. Jusstetj had left Penœa among the Genera of un- 

 certain place. 2 In 1830, Kunth 3 divided the old genus Penœa of 

 Linn^us and Thunberg into three genera ; Penœa, Sarcocolla and 

 Geissolomeœ. But Endlicher, 4 in 1841, placed the Geissolomeœ in 

 a small distinct group, following the Pcnœaceœ, which consequently, 

 according to him, contained only Penœa and Sarcocolla. In 1846, 

 A. de Jussieu, in a note on the family of Penœaceœ, 5 added to the 

 preceding genera Stylapterus and Brachysiphon, which we cannot 

 retain, and the genus Endonema, to which should be added one of the 

 three species of Sarcocolla admitted by him, and of which A. de 

 Candolle, in 1867, 6 made a genus Glyschrocolia, proposed by End- 

 licher 7 as a section of Sarcocolla. Summing up, the Pcnœaceœ with 

 us number only three genera, comprising some twenty species, all 

 natives of the Cape, all frutescent or suffrutescent, 8 with opposite 

 persistent leaves, regular apetalous isostemonous flowers, gyna?cium, 



1 Introd. 71; Veg. Kingi. (1846), 577, Ord. rally has a square or lozenge shape, in accord- 



209. — Sweet, Hort. Brit. (1827), 488. anee with the arrangement of the leaves. In 



3 Gen. (1789), 419. the medullary cellules and in those of the ver- 



3 I.inncea, v. (1830), 676. tieal parenchyma is found a yellowish or 



4 Enchirid. 213, Ord. 112; Gen. 335. brownish resinous substance,in appearance much 



5 Ann. Sc. Nat. sér. 3, vi. 15. resembling the gum-resins of the Burserœ, and 



6 Prodi: xiv. 483, Ord. 165. also certain Sarcocoh, probably spurious, found 



7 Gen. Suppl. iv. 73. here and there in commerce. 



8 The transverse section of tho wood gene- 



