sometimes four or more. STAMENS none. Ovary adnate, three-an- 
gled, three-celled, with polyspermous central placente. STYLE short 
and thick, three-lobed. St1iemas three, waved or two-lobed. CapsuLE 
membranous three-sided, unequally three-winged, three-celled, many 
_ seeded, opening at the base in slits along the margin of the wings. 
Sreps with a thin transparent reticulate testa, a very cellular embryo, 
no albumen, and a short radicle directed towards the hilum. 
DescripTION OF THE Spectres, Begonia INcaRNAaTA. STEM 
herbaceous, three to four feet high, smooth, swollen at the joints. 
Leaves on rather long foot-stalks, very unequally cordate, acuminate, 
angularly toothed or almost lobed, and also furnished with a number 
of small ciliate serratures of a shining green, with a few scattered hairs 
above, smooth and often reddish underneath. Streuxes lanceolate, 
pellucid. PepuNcLEs two or three times dichotomous, at first appa- 
rently terminal, but becoming at length lateral, nodding. BracTeas 
ovate cordate, deciduous. FLOWERS moneecious, large, rose-coloured. 
Mates: Sepats large, round-cordate. Prras nearly as long, but 
narrow, spathulate. STaMENs connected together at the base. AN- 
THERS oblong-clavate, rather longer than the fore part of the filaments. 
FemaLes: Loses of the calyx smaller than the sepals of the males, 
and not so broad. One PETAL (opposite the convex side of the ovary) 
as large as the divisions of the calyx, and, more or less deeply and 
unequally divided into two; the other petal nearly as long, but much 
narrower. CapsuLE three-winged, one very large, triangular, with a 
sharp or somewhat blunt external angle; the opposite one considerably 
smaller, and very blunt; the lateral one very narrow. 
Poputar AnD GrocrapnicaL Notice. The genus Begonia, nu- 
merous in species, and widely diffused in the hotter regions of both 
hemispheres, with the exception, perhaps, of the continent of Africa, 
has, by the singularity of its structure, much puzzled the most distin- 
guished botanists, both as to the nature of the floral parts, and as 
to the affinities of the genus forming by itself the order Begoniacez 
it was formerly considered as having but a simple floral envelope, and 
as being allied amongst Monochlamydew to Polygonacee. This 
opinion is now, however, generally abandoned. All appear to agree 
in classing it amongst vegetables of a higher order of structure, with 
a double floral envelope; some considering the divisions of the appa- 
rent perigon as a true corolla, the calyx being supposed to be abor- 
tive; and others following Lindley the perigon of the ordinary species 
as a calyx, and the additional divisions observable in some as the 
corolla. The affinities of the order have been severally sought in 
Hydrangee, Umbellifere, Campanulacee, Cucurbitacee, and Ona- 
grariex. 
With regard to the floral envelopes, starting from the principle that 
all such organs are intrinsically of the same nature, viz: transformed 
