Andropogon."] CLXXIII. GRAMINE.E. (J. D. Hooker.) 205 



Var. filiculmis, Hook, f.,; stem 12-16 in., very slender, leaves fa-fr in. broad, 

 spikes reddish-brown. Upper Burma, Fort Stedman, Herb. Hort. Calcutt. 

 Probably only a depauperate form. 



Var. versicolor, Hack. 1. c. 610 ; leaf-base cordate semi-amplexicaul, spathes 

 dull violet and red, spikelets more or less variegated or the pedicelled violet. 

 A. versicolor, Nees in Wiglit Cat. n. 1705 ; Lisboa in Journ. Bomb. Nat. Hist. 

 Soc. iv. (1889) 120, vi. (1891) 65. A. aromaticus, Sieb. Fl. Maurit. (Herb. 

 Exsicc.) n. 38. A. connatus, Hochst. in A. Rich. Tent. Fl. Abyss, ii. 464. A. 

 excavatus, Hochst. in Flora (1846) 116. A. foliatus & polyneuros, Steud. Syn. 

 Gram. 385, 389. A. nardoides, ft, Nees I.e. A. pachynodes, ,8, Nees PL Meyen. 

 190. A. pruinosus, Nees ex Steud. L c. 388. A. Schoenanthus, Wall. Cat. 8794 

 C.N., 8795, Baker Fl. Maurit. 446 ; Trim. Cat. Ceyl. PL 107; Went. Hort. Gels. 

 t. 89. A. versicolor, Nees ex Steud. L c. ; Thw. Enum. PL Zeyl. 367. Gymnan- 

 thelia connata, Ascliers 8f Schweinf. in Schweinf. Beitr. FL JEthiop. 310. 



Var. ccesius, Hack. 1. c. ; whole plant more or less glaucous, stem slender, leaves 

 very narrow base not dilated, panicle subsimple, gl. I narrowly winged. ?A. 

 bicornis, Forsk. Fl. JEg. Arab. 173. A. csesius, Nees in Hook. Sc Am. Bot. Beech. 

 Vny. 244, in PL Meyen. 189 ; Wight Cat. n. 1700, I700a. A. Martini, Benth. Fl. 

 Hongk. 424. A. Iwarancusa, JBoiss. Fl. Orient, v. 466. Andropogon, Wall. Cat. 

 n. 8796. 



Var. gracillimus, Hook, f . ; glaucous, stem very slender, leaves 2-4 by -1 in., 

 panicle simple, joints of rachis of spikes and pedicels nearly glabrous, gl. I of sessile 

 spikclet truncate dorsal channel deep or shallow. Andropogon, Wall. Cat. n. 8796 

 (Herb. Wight). Trachypogon caesius var. a, Nees in Herb. Wight. 



Subsp. clandestine, Hack. I.e. 612; habit of var. ccesius, but gl. I and II 

 of the pedicelled spikelet pubescent. A. clandestinus, Nees ex Steud. L c. 388. 

 Andropogon, Wall. Cat. n. 8794 I K. Burma, Wallich', Upper Burma, Herb. 

 Hort. Calcutt. 



72. A. Nardus, Linn. Sp. PL 1046 ; stein tall paniculately branched 

 leafy, leaves narrow flat, panicle subsimple to decompound, spikes in 

 densely or loosely" compacted sheathed fascicles or subsolitary, joints and 

 pedicels rather slender tips dilated toothed, sessile spikelets y g- in. oblong 

 or lanceolate, gl. I 2-toothed, dorsally flat or with shallow depressions or 

 concave towards the base, margins winged or narrowly margined and 

 scaberulous above the middle. 



Throughout the hotter parts of INDIA, BURMA, the MALAY PENINSULA and 

 CEYLON, wild or cultivated. DISTRIB. Trop. Asia, Africa, Australia. 



In its common form it is difficult to distinguish A. Nardus from the narrow 

 leaved form of A. Schvenanthus, except by the cleft in the gl. I of the sessile 

 spikelet of the latter ; and not always very clearly by that. Both are cultivated in 

 India and elsewhere, and A. Nardus is (except in foliage never having the broadly 

 cordate leaf-base) quite as variable as its ally. I have been unable satisfactorily to 

 correlate all of the Kew Indian forms with the elaborate classification of varieties 

 established by Hackel. Both his and my conclusions must be regarded as pro- 

 visional and as certain to be disturbed by a study of additional materials, or by a 

 study of all the forms in their native localities by a specialist who has grasped 

 the history of the species wild and cultivated, over a large area of its distribution. 

 One of Hackel's varieties, distans, regarded as a species by Nees, should, I 

 think, be restored to that rank (excluding a Ceylou plant included under it by 

 Hackel) ; it is exclusively a Himalayan form, not found below 4000 ft., and is very 

 constant in habit and foliage. 



The var. genuinus, Hack., is founded on an awnless form of the plant, which I 

 believe (with Eoxburgh) to be a cultivated state, having seen no awnless 

 specimens in the forms known to be wild. The name is therefore inappropriate 

 botanically, whether or no it represents the first described state of the plant, and 

 better suppressed. 



