592 
CLVII. G It AMINE iE (Stapf.) 
[Urochloa. 
Marsh, on the lower Shire, Scott ! between Kongwe and Karonga, 1700-2000 ft., 
Whyte ! Mashonaland : North Mazoe district, Mundy ! Southern Rhodesia : 
Victoria Falls, Rogers , 5719 ! and without precise locality, Walter, 2216 ! 
Var. (?) mosambicensis, Stapf. Lower glume, with or without a fine stiff 
hair springing from a small central tubercle. Upper glume more or less 
pubescent. Valve of lower floret pubescent or shortly (very rarely long) 
fringed. — Panicum mosarribicensis. Hack, in Bolet. Soc. Brot. vi. 140 ; Durand & 
Schinz, Consp. FI. Afr. v. 755 ; K. Schum. in Engl. Pfl. Ost-Afr. C. 104. 
Mozambique Distr. Portuguese East Africa : Mozambique ; mainland 
opposite the Mozambique Island, Carvalho ! between Shupanga and Sena, 
KirJc ! Shupanga, Kirk ! Mazzaro, Kirk ! Nyasaland : Katunga, Scott ! 
Matabeleland : Insiza District, Mundy, 19 ! Victoria Falls, Allen, 267 ! 
Ngamiland : Kwebe, Lugard, 163 ! and without precise locality, Stigand ! 
The status of these specimens is somewhat doubtful. Some of them resemble 
typical examples of U. jpullulans so closely that they might be considered 
replicas of them, but for the hairiness of the spikelets; e.g. Scott’s Katunga 
specimen (6-10 ft. high, upper glume and flexures of lower floret densely 
pubescent) compared with the plant collected by him at Chirimbo, two of 
Lugard’ s inflorescences contrasted with the other two from the same collecting 
(upper glume pubescent, lower floret shortly fringed), Allen’s Victoria Falls 
specimen and Kirk’s from Mazzaro and Shupanga on one hand and Allen’s 
no. 1 from Portuguese East Africa on the other (lower floret shortly fringed) ; 
Carvalho’s and Mundy ’s from North Mazoe (upper glume and flexures of lower 
floret very pubescent), etc. The Insiza specimens have longer spikelets with 
rather long fringes, but they repeat the habit of'Mundy’s North Mazoe typical 
samples. It is impossible to decide from dried material whether these modifica- 
tions are of the nature of unstable fluctuations or represent definite strains 
(and crosses of these) of indiscriminate distribution. The specimens referred to 
Panicum trichojpus in Dyer's Flora Capensis, vii. 392, represent Hackel’s Panicum 
dorsisetum (name only in Durand & Schinz, Consp. FI. Afr. v. 748). They 
differ from typical XJ. jpullulans in having still smaller and mostly narrower 
spikelets with glabrous or very slightly pubescent upper glumes and shortly 
fringed or unfringed lower florets. A very fine minutely tubercle-based hair 
is generally present on the lower glume. 
i l-jcce-'tl .) 
4. U. brachyura A Stapf. Annual, up to over 4 ft. high. Culms 
scantily tufted, geniculately ascending, sparingly branched, 4- or 
more-noded, internodes at length shortly exserted upwards, glabrous 
or the upper more or less pubescent or softly hirsute. Leaf-sheaths 
somewhat loose, pale, striate, densely ciliate along the outer margin, 
otherwise glabrous or sparingly hairy, hairs tubercle-based, nodes 
appressedly and often obscurely silky ; ligules reduced to a densely 
ciliate rim, cilia over 1 lin. long ; blades linear from a widened and 
semi-amplexicaul base, long-tapering to an acute point, 6 to over 
12 in. by 3-6 lin., flat or wavy downwards, soft or slightly rigid and 
stout, pale or yellowish-green, softly hirsute all over and ciliate 
with tubercle-based hairs or almost glabrous, finely scaberulous, 
margins narrowly cartilaginous, scabrid, sometimes crisped, midrib 
very slender, slightly prominent below, primary lateral nerves 
about 4 on each side, very slender or obscure. Inflorescence of few 
to 9 erect or suberect straight and stiff or subflexuous sessile or sub- 
sessile spiciform subsecund racemes ; common axis slender, 1-3 in. 
long, terete below, then semiterete, more or less flattened upwards, 
finely and densely pubescent. Racemes dense, 2- or more or less 
