28 Proceedings of the Biological Society of Washington. 



Andropogon halepensis anatherus n. subsp. 



Fertile lemma hyaline, oval, obtuse, 1-nerved, awnless; otherwise 

 similar to A. halepensis. 



Abundantly introduced in the United States, apparently as common as 

 the plant with awns. In the Old World its range seems practically con- 

 terminous with that of A. halepensis. 



Type specimen in U. S. National Herbarium collected at Marco, 

 Florida, August, 1900, A. S. Hitchcock No. 1900. 



Hackel evidently confused under the name A. halepensis muticus two 

 different plants, but his citation of specimens indicates that he had in 

 mind particularly the one which occurs in India, namely, the awnless 

 form of A. halepensis miliformis. 



Andropogon halepensis leiostachyus Hack, in DC. Monogr. Phan. 



6:502. 1889. 



Sessile spikelet awned, glabrous. 

 Known only by specimens from Corsica. 



Andropogon halepensis miliformis (Schultes) n. comb. 



Andropogon miliaceus Roxb. Flora Indica 1:276. 1820. Described from 

 specimens grown at Calcutta from seed collected in the ' ' mountains 

 north of Oude," India. 



Andropogon miliformis Schultes Mantissa 2:448. 1824. Change of name 

 of the above on account of the older A. miliaceus Forsk. 



"Andropogon laxus Linn. Sp. PL ed. Willd." 4:907. 1805. Under the 

 above name Roxburgh, Flora Indica 1:275. 1820, describes a plant 

 which is probably not different from his A. miliaceus. The plant is 

 certainly not A. laxus Willd. as above cited. It has usually been 

 called J., laxus Roxb. 



Andropogon controversus Steud. Syn. PI. Glum. 1:391. 1854. A change 

 of name for the grass described as Andropogon laxus Linn, by Rox- 

 burgh on account of the older Andropogon laxus Willd. 



Rootstocks very abundant, short and thick, making a dense mat; 

 stems tall, slender, 2-3 meters high, with usually 11 nodes; leaf blades 

 5-15 mm. wide; panicle very large and loose, 15-60 cm. long, half as 

 broad, the very slender branches and the tip somewhat drooping; lower 

 and upper glumes coriaceous, becoming dark at maturity; awn 12 mm. 

 long; caryopsis yellow-brown, ellipsoid, compressed, 2-2.5 mm. long. 



Young and starved specimens are not easily distinguished from true 

 halepensis but well grown panicles are always very loose and drooping. 

 Hackel speaks of some of these Indian forms as intermediate between 

 halepensis and effusus, but this statement refers to the panicle and is 

 without due consideration of the rootstock character. 



Stapf (Hooker, Flora Brit. India 3:182. 1897) considers that there are 

 two forms in India, one with tlie sessile spikelets j to ^ inch long 



