304 CLxxiii. gramineji. (J. D. Hooker.) [Fhragmitei: 



P. pamila, Oriseb. in Ooett. Nachr. (1868) 76 {an Willk. ?). P. vnleraris, 

 Trin. Fund. Agrost. 134. Arundo graeca, Link in Linnsea, ix. (1834) 136. 

 A. maxima, Forsk. Fl. ^g. Arab. 24 A. Phragmites, Linn. 8p. PI. 81 ; 

 Most Oram. Avstr. iv. 28, t. 39 ; Bnql. Bot. t. 401 ; Knapp Oram. Britt. 

 t. 95 ; Ledeb. Fl. Ross. iv. 393 ; Aitchis. Gat. Panjah PL 266. ? A. tecta, 

 Blanc. Fl. Philipp. 48. A. vulgaris, Lamk. Fl. Franc, iii. 615. Czernya 

 arnndinacea, Presl Cyp. ^ Gram. Sicil. 22. 



Kashmik Like, nit. 5-6000 ft., Jacqvemonf, &e. Garwhal, alt. 4-5000 ft., 

 Dvihie. Western Tibet, alt. 9-10,000 it., Thomson, &c. — Disteib. N. and 

 S. temp, regions. 



Stem erect from a stont creeping rootstoek, leafy up to the panicle, fistalar. 

 Leaves distichous, linear to broadly lanceolate, tip subulate, base rounded, not 

 amplezicaul ; lignle a ridge of short hairs. Panicle usually brownish purple ; 

 branches very slender, smooth, hairy ; peduncle often silkily hairy at the very base 

 of the panicle, sometimes densely. SpiTcelets 4-i in.; gls. spreading in frt. and 

 exposing the long silky hairs of the riicliilla, I and II longer than the flg. — Very 

 variable ; in dry situations in Tibet and Scind the stems are very short, straggling, 

 with ovate-subulate concave rigid pungent leaves 1-2 in. long, and a reduced panicle. 

 This starved state is the Cenchrus frutescens, Sieb. in Bot. Zeit. ix. (1822) 14. 

 F. pumila, Griseb., is another reduced condition of P. communis, from Tibet, iu which 

 gl. Ill is as short as in P. Karka. Grisebach in Goett. Nachr. (1868) 76, has also 

 var. isiaca (P. isiaca, IteicAb.) as a native of Tibet. I have not seen his specimens, 

 but neither Dr. Stapf nor I can find a valid character for this variety in European 

 specimens, 



2. P. XLarka, Trin. ex Stevd. Nom. Ed. ii. 324 ; stems very tall stont 

 and leaves perennial, panicle ^-2 ft. erect, branches widely spreading-, gl. Ill 

 rarely more than \ in. P. Roxburghii, Sleud. I. c. ; Nees in Nov. Act. Nat. 

 Gnr. xix., Suppl. i. (1843) 178; Thwaites Enum. PI. Zeyl. 370; Trim. Cat. 

 Ceyl. PI. 109 ; Benth. Fl. Eongk. 428 ; DutAie Grass. N. W. Ind. i. 35, Fodd. 

 Grass. N. Lnd. 60 ; Franch. Sf Sav. Enum. PI. Jap. ii. 1 70 ; lAshoa in Journ. 

 Bomb. Nat. Hist. Sac. vii. (1893) 378. P. bifaria, Wight Berb. n. 3:<10. 

 P. nepalensis, Nees ex Steud. Syn. Oram. 196. Arundo Karka, Retz. Ohs. 

 iv. 21 ; Boxb. Fl. Ind. i. 348. A. Roxburghii, Kunth Bevis. Gram. i. 79, 

 Enum. PL i. 248; Griff. NotuL iii. 47, Ic. PL AsiaL t. 1S9, f. 243. A. 

 Eoxburghiana, Kunth ex Steud. Nom. I. c. A. tibialis, Boxb. Ic. Pict. 

 iined.), t. 854, Sc in Wall. Gat n. 5017 a. A.Donax, Serb. Bam. ex WalL 

 I.e. B. A. Corea, Herb. Bottl. Oxyanthe japonica, Bterid. I.e. 197. 

 Sericura japonica, Steud. in Flora, xxix. (1846) 20. Trichoon Karka, Both. 

 Catalect. ii. 2.— Arundo, Wall. Cat. n. 5017, 5018 P., 5019.— Arundo, 

 Griff. I. c. 48. 



Throughout India ; from the Panjab to Burma, and southward to Ceylon, ascend- 

 ing the Himalaya in hot valleys to 3000 ft. — DisiKlB. Affghan., Japan, Trop. Asia, 

 Africa, and Australia. 



Except by its greater size, larger more spreading panicle with rather smaller 

 spikelets, more spreading glumes, and shorter gl. Ill, I can point out no character 

 whereby herbarium specimens of P. Karka can be distinguished from P. communis, 

 and none of those are constant. In both dwarf or very slender states occur with 

 almost filiform leaves and grcMtly reduced panicles. Dr. itapf, who has carefully 

 revised my separation of the great pile of Indian specimens into the two species, is 

 equally at a loss to point out auy other characters than those given above whereby to 

 distinguish them, except that the rachilla appears to him to be rather longer and 

 more slender in P. Karka and the glumes hence more widely apart ; and tlii" glumes 

 also are rather narrower. The perennial duration of the stems and leaves of P. Karka 

 requires verification. These organs are undoubtedly annual in the northern form of 



