Brachiaria.] 
CL VII. GRAMlNEiE (Stapf). 
557 
Bl. 19 (name) ; A. Rich. Tent. FI. Abyss, ii. 364 ; Steucl. Syn. PL 
Glum. i. 63 ; Engl. Hochgebirgsfl. Trop. Afr. 120 ; Durand & Schinz, 
Consp. FI. Afr. v. 765 ; K. Schum. in Engl. Pfl. Ost-Afr. C. 101 ; 
Chiov. in Ann. Istit. Bot. Roma, vii. 63. P. nilagiricum, Steud. 
l.c. 62. P. villosum , Hook. f. FI. Brit. Ind. vii. 34, and in Trimen, 
Handb. FI. Ceyl. v. 139 (partly) ; not of Lam. (?). 
Nile Land. Abyssinia : Tigre ; shady grassy places near Adoa, Schimper, 
289 ! Shire plateau, Schimper, 1833 ! Metgalo, 6700 ft., Schimper, 1018 ! 
Lower Antitscho, 5500 ft., Schimper , 1022! Saraen; Jaja, Schimper, gfiS. ! 
and without precise locality, Schimper, 177 ! 237 ! South Abyssinia; Djan- 
bure (Djaribule ?), among herbs, 11000 ft., Riva, 81. Uganda: Ruwenzori, 
Stuhlmann ! British East Africa : near Nairobi, Miss Johnstone , 43 ! 
Also in India (Western Ghauts) and Ceylon. 
All the African specimens examined have glabrous spikelets, but those 
from India have them either glabrous or pubescent to a varying degree. In 
the Flora of British India, l.c., Panicum semiundulatum is given as a synonym 
of P. villosum , but this also includes P. coccospermum, Steud. (P. vestitum , 
Nees ex Steud.), a species of Northern India where it ranges from the Chenab 
to Sikkim, the Khasia Hills and Upper Burma. Lamarck’s description of 
P. villosum leaves it somewhat doubtful which of the two species was meant, 
though it points rather to P. coccospermum. On the other hand the source 
of the originals of P. villosum — namely, Sonnerat’s collections — suggests 
P. nilagiricum. The question can only be decided by the examination of 
the type in the Lamarckian Herbarium. Chiovenda, in Ann. Istit. Bot. Roma, 
viii. 299, accepts Hooker’s identification of P. villosum, Lam., with P. nila- 
giricum, Steud., and describes (p. 300) a var. Erythrceum, which is said to 
be characterised by its smaller stature and smaller and glabrous spikelets. 
It is almost certainly the pJant described here as B. semiundulata. He quotes 
it from the following districts of Eritrea : Amasen, Pappi, 4780. Sarae, 
6230 ft., Pappi, 22, 496. Ocule Cusai, Pappi, 1447, 2014, and Scimezana, 
Mount Matara, 9100 ft., Pappi, 928. 
45. B. distichophylra, Stapf. Annual, ft. high. Culms very 
slender, terete, geniculate, usually ascending from a decumbent 
or prostrate and rooting base, often much branched below, the 
culms and branches gathered in more or less dense bunches, 5- to 
many-noded, the uppermost internode usually very long and like the 
preceding internodes (or at least their upper parts) pubescent, rarely 
glabrous. Leaf -sheaths more or less herbaceous, rather tight, those 
supporting a branch slipping oh the internode and rolling round the 
branch, finely striate, more or less pubescent, sometimes densely so, 
or finely hirsute, rarely glabrous ; ligules reduced to a narrow ciliolate 
rim ; blades lanceolate or linear-lanceolate from a suddenly con- 
tracted rounded base, gradually tapering from very low down to an 
acute point, 1-2 in. by 2-3 lin., rarely longer, rather stiff and somewhat 
succulent, green, glabrous or more or less pubescent, margins markedly 
cartilaginous and usually crisped downwards, spinulously ciliolate to 
serrulate, primary lateral nerves 3-4 on each side, very fine and 
differentiated from the numerous and close secondary nerves only 
below, midrib very slender. Panicle at length long-exserted, secund, 
1J-3 in. long, of 4-8 obliquely spreading or almost horizontal distant 
solitary slender spike-like mostly simple racemes ; common axis very 
