CONNARACEA). 9 
number of free follicles, tapering at the base, then swelling out, 
and tipped by a little reflexed apiculus. Each follicle opens at 
maturity along its ventral angle, we may then easily distinguish 
the rather fleshy pericarp from the woody endocarp, which is a little 
shorter ventrally than the rest of the pericarp.’ Hence it gapes on 
this side and parts from the contained seed a little above its micro- 
pyle. ‘The seed (fig. 13), now free in the endocarp,’incloses in its 
coats a copious, nearly horny albumen, in whose axis is a long green 
embryo with flattened cotyledons and a superior radicle. The whole 
of the outer surface of the seed consists of a fleshy tissue, which, 
as in Magnolia, represents the external coat thus modified through- 
out; it may be viewed as an aril, generalized in Manotes, but 
specialized in Connarus and its allies. Three species of Manotes are 
known, all natives of the west of tropical Africa.’ 
In Tricholobus' (fig. 14) we find the habit and foliage of Connarus, 
with flowers whose perianth and androceum resemble those of Manotes; 
the five sepals are valvate; the five longer alter- 
nating petals are imbricated or twisted in the ” 
bud; and the monadelphous androceum consists 
of ten stamens, whose filaments are free above, 
and bear introrse two-celled anthers dehiscing 
longitudinally. The five stamens superposed to 
the petals are the shorter, and may even become 
altogether sterile. But the gyneceum never at 
any age consists of more than one carpel, whose 
free one-celled ovary is surmounted by a style of 
variable length, dilated at the tip into a stig- 
matiferous head. The fruit is a sessile or stipi- 
tate pod,’ surrounded at the base by the non- 
accrescent calyx, and containing within a pericarp 
of variable consistency an ascending seed,’ which possesses a 
somewhat lateral, irregularly-lobed aril, and a thick, fleshy, ex- 
albuminous embryo, with its radicle superior. 
Jeholnhus 5 ps es 
Fre. 14. 
Fruit, right valve removed. 
1 The woody endocarp sends a long hard tail 
into the stalk of the follicle. 
2 This it is which Pxrancuon described as 
an aril, also mistaking the lower hard con- 
tracted part of the endocarp for a funicle (see 
Adansonia, loc. cit., 246). 
3 Baxer, loc. cit., 459. 
4 Bu. Mus. Bot. ‘Tuga. ~Bat., i. 236. —B.H., 
Gen., 433, n. 9. 
5 This is the only name which can be used to 
describe it, as it opens by two longitudinal clefts 
into two valves, which are altogether free from 
each other and only adhere to the receptacle 
by their bases. One of these valves has been 
detached in fig. 14, where we only see its 
cicatrix. 
® Its attachment may be altogether basilar as 
in TZ’, cochinchinensis H. By. But, as in Con- 
