38 Proceedings of the Biological Society of Washington. 



Islands), J. M. Hildebrandt, June-Aug., 1875 (Kew. Berlin); Mahe, 

 Seychelles, G. Neville, 1887 (Kew). 



The Balfour specimens (rather mature) both shed the spikelets like 

 Tunis grass from a well formed scar or cicatrice; the other specimens do 

 not usually form a scar. 



Specimens from Madagascar collected by Rev. R. Baron, 2385, Dec. 

 1883 (Kew), 4568, Dec. 1885 (Kew), have the fertile spikelets only 

 5 mm. long, and are a trifle more pubescent. 



A specimen in the Kew herbarium collected by A. Cunningham in 

 Australia before 1862 is doubtfully referred here. 



Most of the material from southeast and south Africa is very similar to 

 the insular plant, but none of it identical. Four forms may be distin- 

 guished, one from German East Africa, a second from about Mount 

 Kilimanjaro, a third from the eastern and southern portion of South 

 Africa, i.e., Natal, Transvaal, the coast of Cape Colony, etc., and a 

 fourth from the drier regions to the westward. 



The material from German East Africa differs in having the pubescence 

 on the spikelets longer and looser. Among the specimens examined are: 

 Tunungus, Ger. E. Africa, Stuhlmann No. 8692 (Berlin); "Werawa, 

 Rukingo, Ger. E. Africa, Stuhlmann No. 6091, Feb. 15, 1894 (Berlin); 

 Usambara, Ger. E. Africa, Dr. J. Buchwald No. 509, Dec. 28, 1896 

 (Berlin); Kavirondo, Br. E. Africa, 6th day from Mumias, ^. Whyte, 

 Dec. 12, 1898 (Kew) ; Ger. E. Africa, Busse No. 140 (Kew) ; Kwa-Wasiri, 

 Uzeguha, Ger. East Africa, W. Busse No. 171 (1900). Busse makes 

 the following notes on No. 171 : found growing about sorghum fields ; 

 natives name "lumbolo." "Diese Pfianze ist im Gabi Sud der culti- 

 vierte sorghum sehr ahnlich, nur durftiger als dass ; vielleicht die durch 

 aberwilderung wiedergewonnen Urform von Andropogon sorghum?" 



Specimens from the vicinity of Mt. Kilimanjaro have decidedly larger 

 spikelets, 6-7 mm. long, but otherwise seem indistinguishable. Such 

 specimens are: Kilimanjaro & Meru near Mebula, Dr. C. Uhlig, No. 

 855, Dec. 18, 1901, 1200 m. alt. with leaves 1 to 2.5 cm. broad and spike- 

 lets mostly fallen, those remaining 5-7 mm. long (Berlin); Kilimanjaro, 

 Moshi, Morderf No. 386, April, 1904 (Berlin), with leaves 3 cm. broad. 



The two following specimens have narrow leaves and small, rather 

 narrow panicles, but the spikelets are very similar: Kilimanjaro, 1100- 

 1200 m. alt. Dr. R. Endlich, No. 41, Nov. 1908 (Berlin). In " Misch- 

 wald und Baumsteppe"; leaves 1.5 cm. broad; Kilimanjaro, Marangu, 

 G. Volkens No. 1477. Dec. 8, 1893 (Berlin). In wet places, scattered; 

 leaves, even the basal ones, narrow, 6-10 mm. wide. 



The plant of the moister portions of Africa south of latitude 28° differs 

 from typical verticilliflorus in having the narrow leaves 1 to 2 cm. broad; 

 a smaller and looser panicle, and usually purple-tinged, slightly larger 

 spikelets, 5-7 mm. long. Here belong the following specimens: without 

 locality, Drege, Hb. Nees No. 4240 (Berlin); Natal, Banks of Tugela, 

 Buchanan, No. 296 (Kew, Berlin); Bet. Shupango & Senna, Dr. J. 

 Kirk, Jan. 1859 (Kew); Kongone mouth of Zambesi, Dr. J. Kirk, Jan. 

 1859 (Kew); Natal, Umlazi River, Krause No. 184 (Kew); Natal, 



