3 
Lunt, 343, I. B. Balfour, Anderson; Sheikh Othman, Schweinfurth, 
; Shugra, Schweinfurth, 141; Wadi Shign, west of Bolhaf, 
Schweinfurth, 148. Perim, Farmer ; Coomaraswamy. British Somali- 
land: Bulhar, Drake-Brockman, 667, 671, 672, 673, 674, 680, 682 ; 
without precise locality, Drake-Brockman, 332. 
and if planted in places where the sea is encroaching too far, might 
as varied generic transferences and corresponding combinations as the 
grass described here. In the course of time it has become connected 
with no fewer than nine mostly widely different genera—a veritable 
Odyssey, hence the name—but it may be hoped that it has at last 
reached a safe port. Bentham was the first to suggest an approxl- 
mately correct affinity, when he placed it in his new section Sclero- 
other species of the new genus mentioned above (see Rendle in Cat. 
Afr, . ii, p. 232, and Stapf in Fl. Cap. vol. vii. 
p. 593), whilst Pilger in Engl. Bot. Jahrb. vol. xlii. p. 95 proposed a new 
subgenus Herpodiplachne of Diplachne for the latter. But a complete 
revision of the genus Diplachne has convinced me that Odyssea has 
while at the same time it is easily recognised and remember : 
Diplachne Jaegeri, Pilg. 1.c. p. 94, from Lake Eyasi, East Africa, seems 
to be a third species of Odyssea.—O. STAPF. 
Fic. 1, spikelet with glumes removed ; 2, florets; 3, lower glume ; 4, upper 
glume ; 5, valve ; 6, valvule (5 and 6 in back view, flattened out) ; 7, grain, back 
view ; 8, grain, front view ; 9, section of grain through the middle. All enlarged. 
