414 NATURAL HISTORY OF PLANTS. 
separation of the petals after anthesis ; these are generally shorter, 
and the flower is sometimes only tetramerous. These characters 
do not permit Aruba to be generically separated from Qvassia. 
They have the same fruit, but sometimes of the largest dimensions, 
as is seen in the drupes of Q. cedron (fig. 468). The leaves are 
alternate, compound-pinnate, sometimes trifoliolate ; the flowers are 
collected in racemes, rarely simple, but oftener ramified and com- 
posed of cymes. Some fifteen of them are known.’ 
Very nearlyrelated to Quassia by the section Aruba, Simaruba is only 
essentially distinguished from it by its dicecious or polygamous 
flowers, hemispherical receptacle more or less velvety, and its impari- 
pinnate leaves, with opposite folioles, We must consider as very nearly 
allied genera: Hannoa, of tropical Africa, which has polygamous 
flowers, with subbilabiate quinquefid calyx, an elongated floral recep- 
tacle with ten grooves, and a fruit formed of five or six scarcely 
fleshy drupes ; Samandura, distinguished by its 3—5-merous, herma- 
phrodite flowers, a calyx glandular at the base, and simple, alternate 
leaves, biglandular at the base; J/annia, whose pentamerous, herma- 
phrodite flower has a five-lobed receptacle, and from fifteen to twenty 
stamens, with pinnate leaves ; MHyptiandra, an Australian shrub, 
whose 4—5-merous flowers, solitary, or few in number, in the 
axils of the simple, entire leaves, with a diplostemonous androceum, 
without scaly appendages to the filaments, and coriaceous fruits, 
with seeds slightly albuminous ; Castela, consisting of small Ameri- 
can shrubs, often thorny, with simple, alternate leaves, flowers 
grouped in small axillary cymes, polygamo-diæcious, tetramerous, 
with fleshy 8-lobed receptacle, four styles with free revolute sum- 
mits, and four drupes with crustaceous stone ; /Holacantha, a thorny, 
aphyllous shrub (imperfectly known) of New Mexico, whose dicecious 
flowers are 7—8-merous, and the fruit formed of a variable number 
of drupes, with but slightly albuminous seeds. 
Ailantus (figs. 469-471), with polygamous, pentamerous flowers, 
whose organization is nearly the same as those of Simaruda, is clearly 

B. H., Gen. 308, n. 2.— Zwingera SCHREB., H. B. K., Nov. Gen. et Spec., vi. 18, t. 514.— 
Gen., ii. 802.— Phyllostema Necx., Elem., n. Hook.,, Kew Journ. ii. t. 11 (Simaba).— 
1075.—Homalolepis Turez., in Bull. Mose. Grise. Fl. Brit. W.-Ind., 139 (Simaba).— 
(1848), ii. 575. Tr., in Ann. Se. Nat., sér. 5, xv. 357 (Simaba). 
1 See upon this question Adansonia, viii. 88. — Bot. Mag., t. 497.—Warr., Ann. i. 161, 
2? A.S. H., Pl. Rem. Brés., 126, t. 10, 11; 162; iv. 420; vii. 737 (Simaba). 
Fl. Bras. Mer. i. as t. 14 (Simaba), — 
