FLORA NIGRITIANA. 287 
may be traced in a slight degree, and very irregularly, in the 
leaflets of the Sorindeia Madagascariensis, and more distinctly 
in the simple leaves of a species from Penang (Wallich, n. 8505) 
and Malacca (Griffith) apparently referable to the same genus, 
as well as in the Dupuisia juglandifolia mentioned below, but 
not in any other terebinthaceous plant I am acquainted with. 
A species of Sorindeia, from Congo, is alluded to by Brown, 
and has been named S. Africana by De Candolle, but being as 
yet undescribed, I have no means of judging whether it be 
different from the above S. heterophylla or not. 
l. Dupuisia? Jongifolia, Hook. fil.; foliolis 15-18 anguste ob- 
longis coriaceis glabris supra glaucis subtus elevato-penni- 
nervibus rete venularum inconspicua, panicula mascula ampla 
floribunda ferrugineo-tomentella, disco staminifero rufo-hirto. 
— Sierra Leone, on the borders of marshes, Vogel. 
Arbor excelsa. Foliorum petiolus cum rhachi bipedalis et longior. 
Foliola 6-8 poll. longa, 14-2 poll. lata, inferiora sublanceo- 
lata, basi rotundato-cuneata, apice breviter acuminata; ul- 
tima subzqualiter oblonga basi longe acutata ; omnia subtus 
siccitate leviter rubro-fusca; petioluli breves, crassi. Pani- 
cule semipedales ad pedales, e ramis ortæ ad axillas foliorum 
delapsorum, pyramidato-ramosissime, temento minuto ferru- 
ginem. Flores in specimine omnes masculi, brevissime pedi- 
cellati, subfasciculati, ebracteati, nutantes, duplo fere majores 
lis Sorindeie Madagascariensis. Calyces breves, lati, ferru- 
ginei, dentibus 5 brevibus distantibus. Petala 1} lin. 
longa, crassa, glabra, sstivatione valvata, per anthesin pa- 
tentia. Discus planiusculus dense hirsutus. Stamina 5, 
petalis breviora; antherz filamento longiores. 
The original species, Dupuisia juglandifolia, A. Rich., from 
Senegal, differs from the above by its broader leaflets, of a 
thinner texture, with the venation of Sorindeia, by its smaller 
flowers and smoother panicle, &e. Both are evidently nearly 
allied to Sorindeia, and satisfactory specimens are wanting to 
ascertain with certainty whether the two genera ought or not 
to be united. In the mean time the number of the stamens, 
