OF FERNANDO PO. 9 
Hab. In Clarence Peak, alt. 9000 ped. (fl. Dec.) 
Herba diffusa, 1-2-pedalis, corolla rubra. 
A common Abyssinian plant, cultivated as horse-forage according to Dr. 
Roth (Herb. Hook.). The Fernando Po specimens differ in having larger 
stipules (?" long.) than the Abyssinian, in which, however, they are very 
large for the genus (nearly 2"): in habit, foliage, inflorescence and flowers 
they are identical. 
3. Trifolium Simense, Fresen ; A. Rich., l. c. 171. 
Hab. In Clarence Peak, alt. 8500 ped. (fl. Dec.) 
Herba 1}’ alt., corolla violacea. 
Very similar, as A. Richard indicates, to T. subrotundum, but well 
distinguished by the calyx, very narrow leaflets, and short petioles. The 
stipules terminate in a long filiform appendage, both in the Fernando Po 
and Abyssinian specimens. 
Rosacea. 
1. Rubus apetalus, Poir. Var. glabrior petalis parvis instructa. 
Hab. In Clarence Peak, alt. 7000 ped. (fl. Nov.) 
Scandens 12-15-pedalis. Petala alba, valde caduca. 
This is clearly a form of R. apetalus of Bourbon, of which I have ex- 
' amined a Bourbon specimen gathered by Carmichael. Another form of it, 
sometimes bearing petals, and otherwise differing in having glabrous carpels, 
is found in Madagascar. A third, always petaliferous, is the R. exsuccus, 
Steud., of Abyssinia, which is described by A. Richard (Fl. Abyss. i. 256) 
as having the fruit entirely dry, but of which Dr. Roth remarks, ** Berries 
eatable” (MS. in Hb. Hook.). Dr. Kirke,who has gathered the latter species 
in the Shira Mountains (Livingstone's Exped., 1860), says, ‘ Fruit good, 
exactly like the bramble, but small." The Fernando Po specimens are 
are not in fruit. This species has not been found in South Africa. 
UNMBELLIFERE. 
l. Sanicula Europea, L. 
Hab. Yn Fernando Po, alt. 4000 ped. (fi. Nov.) 
I should suspect some error in the low elevation assigned by Mr. Mann 
to this plant, were his specimens not so carefully ticketed in other cases that 
I have every reason to put confidence in this. A. Richard gives S. Euro- 
pea, var. Capensis, as a native of Abyssinia ; he does not, however, say what 
the characters of that variety are, and adds that his specimens are identical 
with Parisian. This plant has a very wide range in the mountainous re- 
gions of both Americas and Asia, and is also found in South Africa. 
2. AGROCHARIS GRACILIS, H.f.  Caule elongato ramoso foliisque 
hispidulo-pilosis, foliis gracile petiolatis bipinnatisectis segmentis 
lanceolatis acutis incisis, pedunculis elongatis sub apices patentim his- 
pidis pilis flexuosis, floribus dense congestis. 
Hab. In Clarence Peak, alt. 7000 ped. (fl. Dec.) 
Herba gracilis, 4-pedalis, A. melananthe (Abyssmiæ) quam maxime 
affinis, differt caule gracili elongato ramoso, foliis magis pilosis, capi- 
