94 . TRIANDllIA— DIGYNIA. Cynodon. 



taper-pointed, ribbed, very rough on both sides to the touch, 

 especially at the edges. Sheaths long, smooth in my specimens, 

 though Professor Schrader informs me they are occasionally 

 rough. Panicle 4 — G inches long, alternately lobed, or divided 

 into several large, half-whorled, bundles, of extremely unequal, 

 angular, rough branches ; the lowermost particularly crowded, 

 generally much more so than is represented in Engl. Bot. 

 Calyx-glumes keeled, acute, slightly unequal, rather tumid j 

 their keels either entirely or partially rough Vv'ith little bristles ; 

 their sides smooth ; their edges not more membranous than the 

 other part, in which last character they differ from the vulgaris. 

 Cor. of 2 unequal valves ; the larger ribbed, and occasionally- 

 awned from a little below the summit. Styles very short. 

 Stigmas thick, feathery. The calyx is either greenish white, or 

 brownish purple, but not so constantly, in different individuals, 

 as to mark a durable variety. 



In /3 the stem is more extensively creeping, sometimes floating j 

 but the chief difference consists in the still more dense and 

 tufted lobes of the panicle, as in EugL Bot. and Leers t. 4./. 5. 

 The calyx, besides, is generally rough all over, with little bristly 

 points 5 and I had thought, with the accurate Schrader, that this 

 roughness would afford a specific character ; but it is wanting 

 in the Liverpool specimens, see Engl. Bot. t. 1532. 



y is distinguished by an elongation of the caL 8^ cor., the Jlowers 

 being imperfect, and many of them transformed to leafy buds. 

 The lower branches of the panicle are abortive. 



A. coarctata, Ehrh, Calam. 133, referred by Schrader to the purple 

 variety, or rather state, of our alba, appears extremely different 

 in its narrow involute leaves, trading radical shoots, and slender 

 loose panicle. I have seen nothing like it in England. Ehrhart's 

 specimen in his Calamaria precisely accords with one sent by 

 Dr. Roth as his A. alpina; but it is a real Agrostis, neither A.al- 

 pina, nor A. rupestris of Willdenow • see Trichoclium n. 2 & 3 

 of Schrader; about vv^hich I find various errors, not concerning 

 the British Flora. 



3G. CYNODON. Dogs-tooth -grass. 



" Richard in Pers. Syn. v. 1. 85." Br. Pr. 187. Nutt. Gen. Amer. 



56. Beauv.Agr.37. t.9.f.\. 

 Digitaria. Schreb. Germ. v. 1. 165. 



CaL of 2 nearly equal, lanceolate, acute, keeled, awnless, 

 spreading valves, containing a single floret. Cor. of two 

 unequal, keeled, compressed, awnless valves; the outer^ 

 most much the broadest, clasping the other. Nectary of 

 2 minute scales. Filam. rather longer than the corolla. 

 Anth. cloven at each end. Germ, ovate. Styles distinct. 



