34 Proceedings of the Biological Society of Washington. 



name or its variant gerau is used in lower Egypt for ^. sorghum virgatus. 

 Mr. Hewison writes that he does not know this form as growing wild. 



In the United States this subspecies has been found very valuable as a 

 hay grass, and it is now extensively cultivated in the semi-arid regions 

 under the name Sudan-grass. 



Specimens definitely referable to sudanensis are the following : Aegyptia 

 superior, Seeber (Kew); Cairo, collector unknown (Kew); Senegambia, 

 1837 (Kew); Kondowe-Karonga, Nyasaland, 2000-6000 ft. alt., A. 

 Whyte (Kew). 



A very closely similar plant is represented by the following specimen: 

 Katagum District, Northern Nigeria, Dr. J. M. Dalziel No. 293, "Tall 

 by rivers." " Dawar rafi " (Kew). This specimen differs from Sudan- 

 grass in having the herbage paler and somewhat glaucous, the leaves 

 broader, 15-25 mm. broad, and the fertile spikelets elliptic-lanceolate 

 7-8 mm. long, becoming reddish at maturity and with a transverse de- 

 pression near the base. In some respects it is intermediate toward 

 Andropogon sorghum drummondii. 



Andropogon sorghum vogelianus n. subsp. 



Stout, the culm at base of panicle 5 mm. thick; upper leaf-blade fiat, 

 5 cm. broad, 45 cm. long, green ; panicle pyramidal, very large, 50 cm. 

 long, 30 cm. broad, erect, loose; branches subverticillate in about 10 

 whorls, ascending, mostly naked below the middle, the lowest one-third 

 as long as the panicle, the nodes pilose ; lower glume of fertile spikelet 

 sessile, lanceolate, acuminate, 8-9 mm. long, 2.5-3 mm. broad, firmly 

 indurated, faintly 7-nerved, straw-colored, persisting, smooth on the 

 back, pilose at base and near the margins with ferruginous hairs; awns 

 15-17 mm. long; sterile spikelets lanceolate, strongly nerved, nearly 

 smooth, 8-10 mm. long on shorter pilose pedicels; mature caryopsis not 

 seen. 



Type collected on the banks of the Nun mouth of the Niger River, 

 Vogel No. 11 (Kew). The color of the hairs on the spikelets may be due 

 to age or other causes. A remarkable plant allied to effusus but with 

 stouter panicle and larger longer-haired spikelets. No other specimen 

 seen matches the Vogel plant but the following are very similar. 



Cameroon River, G. Mann No. 210Si, January, 1863 (Kew). Leaves 

 2 cm. broad, pubescence of the spikelets scanty, white. 



Idu, Engermi River, /. H. Holland No. 153, Sept. 23, 1898 (Kew). 

 Perhaps Idume on Ngunie River. A small slender plant with leaves 

 1 cm. or less broad, panicle 8-10 cm. long, and spikelets awnless, other- 

 wise like those of the Vogel plant. 



Korbo, Chari River, near Lake Chad, Chevalier No. 9397, July 30, 

 1903 (Berlin). This agrees with the Vogel plant, except that the spike- 

 lets are 10 mm. long and the pubescence is white. 



Waly Fluss, German East Africa, R. Bohm No. 101, March 2, 1882 

 (Berlin). Panicle 45 cm. long, 30 cm. broad, loose and open with stout 

 spreading branches. Awns 14 mm. long. Spikelets 7-10 mm. Agrees 

 well with the Chevalier plant. 



