316 MR. BENTHAM ON TROPICAL LEGUMINOSE. 
of Barter’s specimens are none of them quite expanded, yet they have every appearance of having 
attained their full size, the lower ones of the racemes being often partially open, whilst those from 
still lower down have already fallen away. 
XVIII. COPAIFERA, Linn. ; Benth. et Hook. f. Gen. Pl. 585. 
Two African trees have been published as new genera, which I should be inclined to 
reduce to Copaifera : they differ strikingly from the majority of the American species in 
their leaves reduced to a single pair of leaflets; but we have one Cuban species, C. hy- 
meneifolia, Moric., in which that is the case, besides the doubtful n. 3150 of Blanchet 
from Brazil, which may not be a congener, being only known in unripe fruit. One of 
these African genera, Guibourtia of Bennett, has every character of Copaifera, except 
that the bracteoles are persistent and about one-fourth the length of the calyx, whilst 
in all the other species they are very deciduous and often very minute: but I have never 
found the persistence and size of bracteoles a good generic character, excepting when 
they valvately enclose the bud, forming as it were an outer calyx. The other genus, 
Gorskia of Bolle, has a thinner pod than other Copaiferæ ; but in the American species 
the pod is variable in thickness and quite flat in some, very convex in others; and this, 
again, appears to me to be a specific, not a generic distinetion. A third tropical-African 
species of true Copaifera has the foliage, minute bracteoles, and flat pod of the West-In- 
dian C. hymeneifolia; but the seed is almost totally enveloped in a scarlet arillus, which, 
in the Cuban species, is thick, obliquely truncate, and under the seed, scarcely embracing 
its base. To these three I provisionally add the widely spread Ironwood tree of Dr. 
Kirk, which, on account of its remarkable seed with corrugated cotyledons and numerous 
immersed resinous vesicles, he had proposed as a distinct genus under the name of Colo- 
phospermum. I have also since received it from Dr. Welwitsch, who had also proposed 
it as a new genus. Unfortunately the flowers have not been detected by either of these 
travellers, and there is nothing in the foliage or pod to distinguish it from the thin- 
podded Copaiferas except the cotyledons; and the corrugated cotyledons observed in 
Macrolobium acaciefolium, Benth., and not in the other species of the same genus the 
nearest allied to it, show that that is not to be regarded as a generic character, unless 
~ accompanied by floral differences. We must therefore wait till we have seen the flowers 
to determine whether Colophospermum of Kirk is to be adopted as a genus or to merge 
into Copaifera. 
The bifoliolate species of Copaifera known to me may be thus distinguished :— 
1. C. HYMENÆIFOLIA, Moric. Pl. Nouv. Amer.i.t.1. Foliolis unijugis falcato-ovatis 
acuminatis penniveniis coriaceis nitidis, floribus pedicellatis, legumine plano co- 
riaceo, arillo crasso sub semine oblique truncato. 
Hab. Cuba (Ramon de la Sagra) ; near Monte Verde in eastern Cuba (Wright, n. 1189). 
2. OC. COLEOSPERMA, sp.n.?  Foliolis unijugis faleato-ovatis acuminatis penniveniis 
coriaceis nitidis, floribus pedicellatis, legumine plano coriaceo, arillo coccineo semen 
involvente. | 
Hab. East tropical Africa. Highlands of the Batoka country, where it forms a large tree, in some parts 
