266 



hollow. Fruit usually capsular, 3-celled, many-seeded ; occasionally berried (the dissepi- 

 ments generally central, proceeding from the axis of the valves, at last usually separate 

 from the latter, and of a different texture. R. Br.) Seeds roundish, or angular, with or 

 without an arillus {albumen floury, its substance radiating, and deficient near the hilum. 

 Jt. Br.) ; embryo enclosed within a peculiar membrane {vitellits, R. Br. Prodr. ; membrane 



of the amnios, ibid, in King's Voyage, 21), with which it does not cohere Aromatic 



tropical herbaceous plants. Rhizoma creeping, often jointed. Stem formed of the cohering 

 bases of the leaves, never branching. Leaves simple, sheathing, their lamina often sepa- 

 rated from the sheath by a taper neck, and having a single midrib, from which very nume- 

 rous, simple, crowded veins diverge at an acute angle. Inflorescence either a dense spike, 

 or a raceme, or a sort of panicle, terminal or radical. Flowers arising from among spatha- 

 ceous membranous bracteae, in which they usually lie in pairs. 



Affinities. Formerly Scitaminese and Marantaceae were united in one 

 tribe called Canneaj, and this is even still followed by some botanists ; hence 

 it is certain that they are at least more nearly related to each other than to 

 any thing; else, and that whatever is the affinity of the one will be that of the 

 other. Taking the vegetation into account, these two tribes are exceedingly 

 nearly allied to Musacese, in which is found the same kind of leaf, the veins 

 of which are closely set, and diverge from the midrib to the margin, being 

 connected by very weak and imperfect intermediate veins ; the leaves have 

 also the same distinct petiole, often with a thickened rounded space at the 

 apex ; Musacese are, however, pent- or hexandrous, with a calyx and corolla of 

 the same texture. Irideae are the next order with which Scitamineae may be 

 compared, agreeing in their superior flowers, which have sometimes an 

 approach to the irregularity of Alpinia and the like, and also in the triple 

 number of their stamens; lout while these organs are all developed in Iri- 

 deae, two are abortive or deformed in both Scitamineae and Marantaceae. 

 Bromeliaceae have been identified with them of old, but their resemblance 

 consists chiefly in the distinction of calyx and corolla, and their inferior 

 ovarium. To Orchideae they are related in consequence of the reduction of 

 their three stamens to one by the abortion of two ; but the cohesion of the 

 stamens and style in the latter, and the want of any distinction between 

 calyx and corolla, sufficiently separate them, besides which the series which 

 produces the stamens in Orchideae answers to the sterile stamens or inner 

 limb of the corolla in Scitamineae. For the differences between Scitamineae 

 and Marantaceae, see the latter. There is a fine volume consecrated to 

 plants of these two tribes by Mr. Roscoe, who first remodelled the genera 

 and reduced them within certain limits. Between the embryo and the albu- 

 men is interposed a fleshy body enveloping the former : this has been called 

 a process of the rostellum by Correa, a cotyledon by Smith, a vitellus by 

 Gaertner and Brown, a central indurated portion of the albumen by Richard. 

 It is now known to be the innermost integument of the ovulum, unabsorbed 

 during the advance of this body to maturity. 



Independently of the presence of this vitellus, the most remarkable part 

 of the structure of vScitauiineae consists in the number of divisions of the 

 floral envelopes, which consist of a tubular calyx, and of two more series 

 instead of one. Mr. Brown, struck with this unusual deviation from the 

 ordinary organisation of Monocotyledons, was disposed to consider the calyx 

 an accessory part {Prodr. 305); but M. Lestiboudois' explanation appears 

 more satisfactory. According to this botanist (as quoted in Ach. Richard's 

 Nouv. Elem. 439), Scitamineae are really hexandrous, like the nearly-related 

 Musaceae; but of their stamens the outer series is petaloid, and forms the 

 inner limb of the corolla, and of the inner series of stamens the central one 

 only developes, the lateral ones appearing in the form of rudimentary scales. 

 This notion of M. Lestiboudois is confirmed by Marantaceae, in which the 

 inner stamens (even that which is aiitheriferous) become petaloid like the 



