92 NATURAL HISTORY OF PLANTS. 



exstipulate leaves articulated at the base. The flowers are solitar}', 

 and terminate the short branches' which bear below them a few 

 alternate, more or less sepaloid, bracts. 



Many other lUhbcrtiaK, which, like the one we have just studied, 

 grow in Australia, present the same general organi- 

 zation, but with some diff'erences in habit and 

 flower. The stems do not climb, being suffrutescent," 

 or herbaceous.' The leaves may be narrow, like 

 those of cei-tain Heaths,^ or dilated below into an ' 

 imperfect sheath. The carpels contain a variable 

 number of ovules,* and are themselves sometimes ten 

 in number (five superposed to the sepals), or even 

 indefinite. In some species the gyna?ceum consists of a 

 single carpel.® But in all the stamens and the external 

 staminodes, if present, are arranged in a circle round 

 the carpels, an arrangement which calls to mind 

 the name " Cyclandra"'' given to all this section of 

 the genus llihbert'ia. 

 The genus Trimorphandrcf has been proposed for a cyclandrous 

 llihljcrtia, of which the outer stamens are short and sterile, as 

 in most of the preceding plants ; but some of the inner fertile 

 stamens are longer than the others — a fact which exists in a 



» n. perfoliata, HuG., in PL Prtiss., i. 266 

 {Candullea perfoliata Leum.), often cultivated 

 in our conservatories, has a Hower of the same 

 construction as //. volubilis, with the outer sta- 

 mens sterile, and with live carpels, each contain- 

 iug from two to four ascending ovules. The 

 raphe is at first exterior, but as the ovules grow 

 the raphes of the adjacent ovules turn towards 

 each other. Besides the fact that the leaves 

 should be noted for their sessile auriculatc blades, 

 we must especially notice that in this species the 

 solitary terminal flowers are on long peduncles, 

 but what has been termed usurpation takes 

 place ; tlie axillary branch being rapidly developed 

 to form a pscudo stem, while the llower becomes 

 very distinctly k'af-opi>oscd. This occurs in several 

 other species, though not w> decidedly. When 

 there are four ovules in two vertical rows the 

 lower pair are much the older. Long before an- 

 thesis they have each an arillury ring round the 

 umbilicus, while the others show no trace of it. 



•' This is the ca'^e in most of our cultivated 

 Hj)ecies except 11. volidiilix. 



" Like (jur culliv.iltd //. t/rosauhirittfoliti, its 

 habit haji been compared to that of J'utentUla. 



* Especially in certain species of Pleurandra 

 cultivated in our conservatories. Sevenil liave 

 the aspect of certain Salsolaceif, while others 

 possess a whitish down recalling that of the Sun- 

 flowers. 



* As in Trisema from a couple to half a score 

 may be counted, but rarely more than this. Their 

 raphes are more or less turned towards one 

 another. 



" For instance, JI. monogyna H. Hr. (ex. DC, 

 Prodr., i. 74) which, with the androceum and 

 perianth of the preceding s|)ecie8, )>08se8scs but 

 one carj)el with one or two ascending ovules in 

 its ovary. This we at one time considered the 

 type of a spt^cial section, very near Trisema, under 

 the name "Jlaplot/t/ne" (Adaiisonia, vi. 280). 

 Hut this cannot Iw maintained as a distinct sec- 

 tion if, following Hkntham (/7. Austr., i. 37), we 

 make //. monogyna only a variety of 11. diffusa 

 R. Hk. (ex. DC. Syst. \. 129). 



7 F. .Mi'Ki.LEK, ex. H. H., Gen. 11, n. 13 (4). 

 — Oihroliusia TvKtz., Hull. Mouc, xxii. (1849), 

 ii. 3. 



" T.pulchella Wii. Si V,\i., luliull. f>uc. Hot. 

 xi. 190; Aiiu. 6c. A'al.. sor. 5, ii. IW. 



