84 LXXXIV. APOCYNACEA (STAPF). | Carpodinus. 
arches ; petiole 1-1} lin. long. Flowers axillary, rarely terminal, solitary 
or sometimes in pairs, subsessile. Calyx 1 lin. long, very minutely 
rusty-pubescent ; sepals 4, ovate, subobtuse, ciliolate. Corolla white, 
fragrant, glabrous except the delicately rusty-downy outer base of the 
lobes, slender; tube slightly widened below the mouth, 8-9 lin. long; 
lobes linear, acute, 12 lin. long, 1$ lin. wide. Ovary finely pubescent 
except at the glabrous base; style slender, 6—7 lin. long, glabrous above. 
—Stapf in De Wild. & Durand, Contrib. Fl. Congo in Ann. Mus. Congo, 
Bot. sér. 2, fasc. i. 34; Hallier f. Kautschuklianen in Jahrb. Hamburg. 
Wissensch, Anstalt. xvii. (1899), 3. Beih. 112; De Wild. & Durand, 
Reliq. Dewevr. in Ann. Mus. Congo, Bot. sér. 3, fase. ii. 149. C. gracilis, 
Stapf, H.cc. 303 and 35 respectively (the fruit according to Hallier f. 
lic.}; Dewévre in De Wild. & Durand, Reliq. Dewevr. I.c. (Dewévre’s 
note). 
Upper Guinea. Cameroons: near Batanga, Dinklage, 1161! (not 1162, as 
Hallier quotes). 
Lower Guinea. Lower Congo: Stanley Pool, Dewévre, 709! Kimuenza, 
Dewevre, 516! 
I suggested in my paper on the Apocynacee of the Congo Free State (Ann, 
Mus. Congo, Bot. sér. 2, fase. i. 36) that some detached fruits accompanying 
Dewévre’s 516/! might belong to C. gracilis. Hallier (l.c.) pointed out, however, that 
this is not the case and suggested that the fruits might be those of C. ligustrifolia, 
1 have since seen a specimen of C. gracilis, collected by Gillet at Kisantu, with 
flowers and fruits still attached to the branch. These fruits are different from 
those which I described as belonging to C. gracilis, and there can be no doubt 
that my assumption was erroneous. Dewévre, in his note referring to 516 (sic) 
quoted by De Wild. & Durand in Relig. Dewevr. l.c., describes some fruits, which 
are evidently the same, but being subsequently mixed up with his 516/! misled 
me. They very likely belong to C. ligustrifolia, the more so, as there is also a 
fragment of a branch with 516/! the leaves of which agree very well with those 
of C. ligustrifolia. The fruits are more or less shortly conic with a wide base, much 
resembling those of C. turbinata, but smaller. Hiallier f. (1.c.) also refers a speci- 
men collected by Pogge (1153 !) near Mukenge in the Congo Free State to C. ligus- 
trifolia, \t has very similar fruits, but is otherwise so defective that its identity with 
C. ligustrifolia is doubtful. Another specimen collected by Laurent at Lusambo! is 
quoted by Hallier under this species, but (consisting of a barren twig) is very doubtful, 
20. C. gracilis, Stapf in Kew Bulletin, 1898, 303 (excl. the fruit). 
A dwarf, sometimes prostrate shrub, with or without very slender 
terminal tendrils; branches slender, reddish-brown, glabrous or 
sparsely hairy when young; lenticels very minute. Leaves opposite, 
ovate-oblong or ovate-lanceolate, gradually and obtusely acuminate 
(acumen 2~4 lin. long), rounded at the base, 13-24 in. long, 7-12 lin. 
broad, thinly coriaceous, scantily hairy below, chiefly along the midrib, 
which is slightly raised above, or glabrescent; lateral nerves about 12 
on each side, like the tertiary nerves and the veins very faint, 
connected close to the margin by flat faint arches; petiole 2-24 lin. 
long, hispidulous. Flowers pedicelled in terminal or axillary few- 
flowered peduncled cymes, or solitary ; peduncie up to 1 in. long, very 
slender, like the pedicels more or less hispidulous, these up to 3 lin. 
