368 oLxxiii. GRAMiNEj;. (J. D. Hooker.) lAgropyron, 



rachilla usually jointed between the fl. gls. Gl. I and II empty, opposite, 

 (not collateral) narrow, unequal, persistent ; fl. gls. rigid, awned or not ; 

 nerves 5-7, converginor above ; keels of palea ciliate. LodAcwles entire or 

 ciliate. Stamens 3. Ovary hispid at the top, styles short, distant. Grain 

 narrow, grooved in front, adherent to the palea or not. — Species about 20, 

 in all temp, climates. 



The following arrangement and nomenclature of the Indian plants of this most 

 puzzling genus is tentative only, and I cannot render tjiese more worthy of con- 

 fidence without undertaking a critical study of the whole genus. The characters 

 given under the species must not be regarded as absolute, though founded on careful 

 study of very large series of specimens ; for the more of these that are studied, 

 especially if they come from different localities, the more do exceptional characters 

 and connecting forms present themselves. It would hence appear that all the 

 members of the section Eu-agropyrum (perennials with dorsally convex not keeled 

 fl. gls. to which all the Himalayan forms belong) will resolve themselves into a very 

 few widely distributed species. A further difficulty is the distinguishing Agropyrum 

 from Eli/mus, and indeed E. sibiricus from A. longe-aristatum, in which two plants 

 the empty glumes are not placed side by side as in Etymus proper ; added to which 

 the spikelets are often solitary in E. sibiricus, and I think sometimes aggregated 

 in A, longe-aristatum, var. Aitchisoni. A. caninum, Boem. & Sch. ; Duthie Grass. 

 K.W. Ind. 44, has according to Duthie's rass. been found at Simla, alt. 6500 ft., 

 by Gamble; and by himself in Gangotri, alt. lC-11,000 ft. I have seen no 

 specimens. 



* Spikelets awned. 



1. A. longe-aristatum, Boiss. Fl. Orient, i. 660; stems leafy, 

 leaves flat or convolute, spikelets loosely imbricate flattened laterally 

 61-12- fld., I and II long or short 3-5-nerved, flg. gls. spreading linear-lanceo- 

 late 3-5-nerved towards the tip narrowed into a long usually recurved 

 awn 1-1| in. long. Brachypodium longearistatum, Boiss. Diagn. Ser. I. 

 vii. 127. Oriseb. in Goeti. Nachr. (1868) 73. Triticnm longearistatum, 

 Jauh. 8f Sp. 111. Fl. Orient, ii. t. 199 ; Steud. Syn. Gram. 346. T. ely- 

 moides, Hochst. ex A. Rich. Tent. Fl. Abyss, ii. 440; Steud. I. c. 



Temperate Himalaya; from Kashmir to Kuraaon, alt. 5500-13,000 ft. 

 Eastekn and Western Tibet, alt. 12-17,000 ft. — Disteib. Affghan, Persia, 

 Abyssinia. 



Stem 1-3 ft., stout or slender, erect or ascending from a tufted base. Leaves 

 6-10 by J-^ in., rarely convolute, and sheaths smooth ; ligule very short. Spikes 

 3-5 in., erect or nodding, or horizontal from the decurved top of a very slender 

 peduncle, rarely ovoid and very dense-fld. ; rachis slender, smooth ; margins of 

 hollows Bcaberulous or ciliolate. Spikelets sessile, remote or approximate, sometimes 

 crowded and secund, \-% in. long without the awus, green or blue-purple. Ql. I and 

 II most variable in length, J-i the length of the lowest fl. gl., elliptic-ovate to 

 narrowly setaceous, awned or not, tip sometimes notched on one side ; fig. gls. 

 narrowly lanceolate, smooth or scabrid, narrowed into a usually straight but some- 

 times recurved stout or slender scabrid awn, prominently 5-nerved above the middle 

 rarely to the base, tip sometimes toothed unilat«rally or on both sides of the awn ; 

 palea coriaceous, keels ciliolate with erect teeth above the middle. Lodicules hyaline, 

 ciliate. Anthers short. Grain oblong, compressed. — Except by the long awns I 

 am at a loss how to diagnose this variable grass, some states of which approach 

 Elymus sibiricus so closely that I can hardly distinguish them. The following 

 varieties are represented amongst Persian specimens. I do not find that they affect 

 definite localities or elevations in the Himalaya. 



1. Gl. I and II very short oblong-lanceolate acute or acuminate usually strongly 

 3-uerved, fl. gls. glabrous or nearly so except towards the tip. 



