RUTALES.] 



MELIACEyE. 



463 



Order CLXXIII. MELIACEiE.— Meliads. 



Melia;, Jess. Gen. 263. (1789).— Meliacea, ^«**. Mem. Mus. 3. 436. (1817) ; DC. Prod. 1. 619. (1824) • 

 Adr. de Jiiss. Memolre (1830) ; Ed. pr. Ixxvii. ; Endl. Gen. ccxxv. • - > 



Diagnosis. — Riutal Exogens, with consolidated berned or capsular fruit, deeply mona- 

 delplious stamens f a few wingless seeds, and dotted leaves. 



Trees or shrubs. Leaves alternate, or occasionally somewhat opposite, simple, or 

 pinnate, without stipules. Flowers sometimes imperfect by abortion, usually in loose 



Fig. CCCXXIII. 



masses. Sepals 3, 4, or 5, more or less vmited. Petals the same number, hj-pogynous, 

 conni^^ng at the base, or even cohering, usually having a valvate or imbricated aestiva- 

 tion. Stamens twice as many as the petals ; filaments cohering in a long tube ; anthers 

 sessile %\dthin the orifice of the tube. Disk frequently highly developed, surrounduig the 

 ovary like a cup. Ovary smgle, with the same number of cells as petab, or fewer (3-2), 

 very seldom many more (10-12) cells ; style 1 ; stigmas distinct or combined ; ovules 

 anatropal, semi-anatropal, amphitropal or orthotropal ! 1 or 2 in each cell, very rarely 

 4. Finiit beiTied, di'upaceous or capsular, often, in consequence of abortion, 1 -celled, 

 the valves, if present, having the dissepiments in their middle. Seeds not winged, with 

 or without an aril ; albmnen fleshy (Meliese), or usually absent (Trichihea?). Embryo 

 with leafy or amygdaloid cotyledons, within which the radicle is drawTi back. 



This Order was ill understood matil it was investigated by Adrien de Jussieu, from 

 whose Memoir I borrow the principal part of what follows. It is, no doubt, related to 

 Citronworts, although Canella, winch was considered a case of transition, is removed 

 from it. The inflorescence of Citronworts terminating m dichotomies ^\ith a central 

 and praecocious flower, the union that sometimes occurs between the filaments of 

 Citronworts, the number of stamens often double that of the petals, and the embryo 

 with a short radicle drawn back between thick cotyledons, are all points in which thei'e 

 is an accordance between the two Orders. The occasionally monadelphous stamens of 

 Rueworts indicate an analogy with that Order, which is confirmed by the general 

 tendency in both cases to produce two ovules in each cell of the ovary. The number 

 and the relative position of the parts of the flower show an affinity with Soapworts, the 

 structure of whose seeds is often absolutely the same as that of MeUads ; their 

 accordance in habit is incontestable, and in fact the species of the two Orders are 

 often mixed together in herbaria. Cedrelads are chiefly distuiguished by their 



Fig. CCCXXIII.— Ekebergia Senegalensis. 1. a flower ; 2. the calyx and staminal tube ; 3. a trans- 

 verse section of the ovary ; 4. a ripe fruit ; 5. a vertical section of the latter. 



